NIST Technical Note 1552

Measurements to Support Public Safety

Communications: Attenuation and
Variability of 750 MHz Radio Wave Signals
in Four Large Building Structures

William F. Young
Kate A. Remley
John Ladbury
Christopher L. Holloway
Chriss Grosvenor
Galen Koepke
Dennis Camell
Sander Floris
Wouter Numan
Andrea Garuti
NIST Technical Note 1552

Measurements to Support Public Safety

Communications: Attenuation and Variability of
750 MHz Radio Wave Signals in Four Large
Building Structures
Christopher L. Holloway
William F. Young
Kate A. Remley
John Ladbury
Christopher L. Holloway
Chriss Grosvenor
Galen Koepke
Dennis Camell
Sander Floris
Wouter Numan
Andrea Garuti

Electromagnetics Division
National Institute of Standards and Technology
325 Broadway
Boulder, CO 80305

U.S. Department of Commerce

Gary Locke, Secretary

National Institute of Standards and Technology

Patrick D. Gallagher, Deputy Director
 Certain commercial entities, equipment, or materials may be
identified in this document in order to describe an experimental
procedure or concept adequately. Such identification is not
intended to imply recommendation or endorsement by the
National Institute of Standards and Technology, nor is it
intended to imply that the entities, materials, or equipment
are necessarily the best available for the purpose.

National Institute of Standards and Technology Technical Note 1552

Natl. Inst. Stand. Technol. Tech. Note 1552, 133 pages (August 2009)
CODEN: NTNOEF

U.S. Government Printing Office

Washington: 2005

For sale by the Superintendent of Documents, U.S. Government Printing Office

Internet bookstore: gpo.gov Phone: 202-512-1800 Fax: 202-512-2250
Mail: Stop SSOP, Washington, DC 20402-0001
 Contents

Executive Summary ...................................................................................................................... v

1. Introduction ............................................................................................................................... 1
2. Signal Measurement Methods ................................................................................................ 2
2.1 Radio Mapping Using a Spectrum Analyzer .................................................................... 2
2.1.1 Transmitters ................................................................................................................... 3
2.1.2 Receiving Antenna and Measurement System................................................................ 3
2.1.3 Spectrum Analyzer Data Processing ............................................................................. 3
2.2 Radio Mapping Using a Narrowband Communications Receiver ................................. 5
2.3 Wideband Excess-Path-Loss Measurements and Time-Delay Spread .......................... 6
2.3.1 VNA-based measurement system ................................................................................... 6
2.3.2 Excess path loss ............................................................................................................. 8
2.3.3 RMS delay spread .......................................................................................................... 8
2.3.4 Measurement set-up ....................................................................................................... 9
3. Building Structure Descriptions and Experimental Set-up .................................................. 9
3.1 Colorado Convention Center, Denver, CO....................................................................... 9
3.2 Republic Plaza, Denver CO ............................................................................................. 10
3.3 Apartment Building, Boulder, CO .................................................................................. 10
3.4 NIST Laboratory, Boulder, CO....................................................................................... 10
4. Experimental Results .............................................................................................................. 11
4.1 Colorado Convention Center Results.............................................................................. 12
4.1.1. Radio Mapping............................................................................................................ 12
4.1.2. Synthetic Pulse System ................................................................................................ 12
4.2 Republic Plaza Results ..................................................................................................... 13
4.2.1. Radio Mapping............................................................................................................ 13
4.2.2. Synthetic Pulse System ................................................................................................ 14
4.3 Boulder, Colorado Apartment Building Results ............................................................ 14
4.3.1. Radio Mapping............................................................................................................ 14
4.3.2. Synthetic Pulse System ................................................................................................ 15
4.4 NIST Boulder Laboratory Results Results ..................................................................... 15
4.4.1. Radio Mapping............................................................................................................ 16
4.4.2. Synthetic Pulse System ................................................................................................ 17
5. Summary of Results and Conclusion .................................................................................... 18
6. References ................................................................................................................................ 24
Appendix I: Experiment Setups and Locations ....................................................................... 26

iii
Executive Summary

This is the sixth in a series of NIST technical notes (TN) on propagation and detection of radio
signals in large building structures (apartment complex, hotel, office buildings, sports stadium,
shopping mall, etc). The first, second, and third NIST Tech Notes (NIST TN 1540, NIST TN
1541, and NIST TN 1542) described experiments related to radio propagation in a structure
before, during, and after implosion. The next two Tech Notes (NIST TN 1545 and NIST TN
1546) focused exclusively on RF propagation into large buildings, with no implosion results.
These reports are intended to give first responders and system designers a better understanding of
what to expect from the radio-propagation environment in disaster situations. The overall goal of
this project is to create a large, public-domain data set describing the attenuation and variability
of radio signals in various building types in the public safety frequency bands.

Because the Federal Communications Commission (FCC) is in the process of auctioning

spectrum between 764 MHz and 776 MHz for a new public-safety band, additional
measurements were carried out in the 750 MHz frequency band in four different large building
structures: a convention center, a high-rise office building, an apartment building, and a
laboratory/office building. Three different types of signal measurements included (1) spectrum
analyzer radio mapping, (2) narrowband communications receiver radio mapping, and (3)
broadband synthetic-pulse measurements.

The first two measurement techniques using radio mapping provide the received signal strength
at a fixed location outside the structure from a transmitter that is carried through the structure
while emitting an unmodulated, 750 MHz radio signal within the structure. The synthetic pulse
system provides stepped-frequency measurements of the received signal across a wide frequency
band (100 MHz to 18 GHz for the measurements reported here). The phase of each received
frequency component is synchronized with the excitation by the use of an optical fiber cable.
Thus, we are able to reconstruct a short-pulse, time-domain waveform in post processing. The
short pulse enables the study of the multipath in a given environment. Figures of merit such as
the RMS delay spread may be calculated and used to quantify the time it takes for multipath
reflections to decay below a given threshold level.

Similar median and standard deviation values are observed across most of the measurements in
both the spectrum analyzer and receiver systems. For the spectrum analyzer, the median values
for all four buildings calculated on data normalized to a direct line-of-sight path ranged from
-25.1 dB to -98.5 dB, and the corresponding standard deviation values ranged from 6.8 dB to
30.1 dB. With the narrowband receiver, the median values ranged from -27.9 dB to -93.0 dB, and
the standard deviation ranged from 9.0 dB to 29.1 dB. From the synthetic pulse measurements,
the RMS delay spread results ranged from 15 ns to 450 ns at various test locations within the
four buildings. In these measurements, the calculation of the RMS delay spread was not
significantly affected by either (a) the directionality of antennas used in acquiring the data, or (b)
the frequency bandwidth used in signal processing. The RMS delay values for measurements
made in large open floor plan buildings were typically two to five times that of measurements in
buildings with relatively narrow corridors.

The measured results presented here provide key parameters that describe the wireless
propagation environment in representative responder environments. Measurement uncertainties

v
are not reported; however, end users of these data are interested primarily in large-scale behavior
on the order of tens of decibels. Including instrumentation repeatability on the order of ±1 dB,
expected for these measurements, is of little value in this case. We anticipate that improved
channel descriptions provided by these measurements will be useful for assessing current and
future wireless technology in emergency scenarios, for standards development, and for
qualifying wireless equipment in environments such as those studied here.

vi
 Measurements to Support Public Safety Communications: Attenuation and
Variability of 750 MHz Radio Wave Signals in Four Large Building Structures

William Young, Kate A. Remley, John Ladbury, Christopher L. Holloway,

Galen Koepke, Dennis Camell, Sander Floris, Wouter Numan, and Andrea Garuti

Electromagnetics Division
National Institute of Standards and Technology
325 Broadway, Boulder, CO 80305

In this report, we investigate radio communication problems faced by emergency responders

(firefighters, police, and emergency medical personnel) in disaster situations. A fundamental
challenge to radio communications into and out of large buildings is the strong attenuation of
radio signals caused by losses and scattering in the building materials and structure, as well as
the large amount of additional signal variability due to multipath that occurs throughout these
large structures. We conducted measurements in four large building structures in an effort to
quantify radio-signal attenuation and variability in scenarios encountered by emergency
response organizations. We performed three different types of measurements. For the first two
types of measurements, we carried RF transmitters throughout the structures and placed two
different types of receiving systems outside the structures. One receiver type was based on a
commercial, off-the-shelf spectrum analyzer, and the other on a NIST-developed narrowband
communications receiver system having a high dynamic range. The transmitters were tuned to
approximately 750 MHz. The third type of measurement tested the time-domain response of the
channel at particular locations within the buildings using a synthetic-pulse measurement system.
These measurements utilized the vector network analyzer with its output port tethered to the
receive antenna by a fiber-optic cable to allow for reconstruction of the time-domain response of
the propagation channel. This report summarizes the experiments performed in four large
building structures. We describe the experiments, detail the measurement systems, show results
from the data we collected, and discuss some of the propagation effects we observed.

Key words: attenuation; building penetration; building shielding; emergency responders;

multipath; excess path loss; public-safety; radio communications; radio propagation
experiments; RMS delay spread; signal variability; weak-signal detection; wireless system.

1. Introduction
When emergency responders enter large structures (apartment and office buildings, sports
stadiums, stores, malls, hotels, convention centers, warehouses, etc.) radio communication to
individuals on the outside is often impaired. Mobile-radio and cell-phone signal strength is
reduced due to attenuation caused by propagation through the building materials and scattering
by the building structural members [1-8]. Also, a large amount of signal variability may be
encountered due to multipath reflections throughout the structures, which can cause signal
degradation in communication systems.

1
Here, we report on a project conducted by the National Institute of Standards and Technology
(NIST) to investigate the communications problems faced by emergency responders (firefighters,
police, and emergency medical personnel) in disaster situations involving large building
structures. The project was sponsored by the Department of Justice Community Oriented
Policing Services (COPS) program. As part of this work, we are investigating the propagation
and coupling of radio waves into large building structures in the 750 MHz frequency band. This
band is currently of interest because the FCC is in the process of allocating spectrum between
764 MHz and 776 MHz for a nationwide, next-generation, interoperable broadband network for
use by the public-safety community.

The experiments reported here were performed in four different large building structures and are
the results of measurements of the reduction and variability of radio signal strength caused by
propagation through the structures. These structures include a convention center, an apartment
building, a 57-story office building, and an office corridor. This study includes data gathered
using three different measurement techniques, each of which will be discussed below.

2. Signal Measurement Methods

Three different measurement techniques were used to capture the behavior of signals penetrating
into structures in the 750 MHz frequency band from both time- and frequency-domain
viewpoints. Use of various types of measurement provides more comprehensive insight into
signal behavior in a given environment because some effects, such as the time it takes for
multipath reflections to die out, are more pronounced in the time domain than in the frequency
domain. While measurement uncertainties are not reported, end users of these data are interested
primarily in large-scale behavior on the order of tens of decibels. Including uncertainties on the
order of ±1 dB, expected for these measurements, is of little value when we are studying
quantities that can vary by tens of decibels when changing by a fraction of a wavelength.

2.1. Radio Mapping Using a Spectrum Analyzer

The experiments performed here are referred to as “radio mapping.” These experiments involved
carrying transmitters (or radios) tuned to approximately 750 MHz throughout the four structures
while recording the received signal at a fixed site located outside the building. A reference level
was used to normalize the received power levels for these data. It consisted of a direct,
unobstructed line-of-sight signal-strength measurement with the transmitters external to the
different structures and in front of the receiving antennas. The separation distance between the
transmitter and receiver antennas varied depending on the receive site location and the building,
and these differences are reflected in the different reference levels. The purpose of the radio-
mapping measurements was to investigate how signals at 750 MHz couple into the structures,
and to determine the field-strength variability throughout the structures. A detailed description of
the transmitters we used is provided in Section 2.1.1 Transmitters, and a description of the
measurement system and antennas are given in Section 2.1.2 Receiving Antenna and
Measurement System. The main procedures and components of the radio-mapping receive
systems are shown in Figure 1. (See Appendix I)

2
2.1.1 Transmitters

The design requirements for the transmitters used in the experiments discussed here were that
they should (1) transmit at 750 MHz, (2) operate continuously for several hours, and (3) be
portable. To accomplish this, a commercially available off-the-shelf radio system was used. The
radios could transmit continuously for 12 to 18 hours. Permission was granted to transmit from
licensees at the frequency of interest, and the National Telecommunications and Information
Administration (NTIA) frequency coordinator was also made aware of our transmissions.

2.1.2 Receiving Antenna and Measurement System

Each receive site used an omnidirectional discone antenna at a height between 2 and 3 meters. In
addition, the receive sites contained a spectrum analyzer, a narrowband receiver, computers, and
associated cabling. Not shown are the power generator and uninterruptible power supply (UPS)
for powering the receive site.

As shown in Figure 1(c), the measurement system consisted of a portable spectrum analyzer and
a laptop computer. The data collection process was automated by use of a graphical
programming language. This software was designed to control the analyzer, and collect, process,
and save data at the maximum throughput of the equipment. The software controlled the
spectrum analyzer via an IEEE-488 interface bus.

The software was written to maximize throughput of the data collection process and to run for an
undefined time interval. This was achieved by running parallel processes of collecting,
processing, and saving the data for post-collection processing. The data were continuously read
from the spectrum analyzer and stored in data buffers. These buffers were read and processed for
each signal and displayed for operator viewing. The processed data were then stored in
additional buffers to be re-sorted and saved to a file on disk.

The sampling rate of the complete measurement sequence was the major factor in how much
spatial resolution we had during radio-mapping experiments (we also had some flexibility in our
walking speed) and the time resolution for recording the signals. Software and parameter settings
allowed experiment sampling rates of between 3 to 5 samples per second. The resolution
bandwidth (RBW) is the key parameter in balancing the dynamic range versus the time between
samples. These experiments used an RBW setting of approximately 500 Hz, which corresponded
to an average spectrum analyzer noise floor in the range of -101 to -103 dBm. The data from the
spectrum analyzer trace were saved in binary format to disk; they were then processed to extract
the signals at the desired frequency. These results were put into a spreadsheet file and saved
with other measurement information.

2.1.3 Spectrum Analyzer Data Processing

The spectrum analyzers collected the raw received power P rec , but the subsequent processing
was performed on the normalized received power P norm . Only the samples collected during the
actual walk-through were used; i.e., the N samples were collected while the prescribed path was
covered. The normalized power was calculated as

3
 P norm (dB)  P rec (dBm)  P ref (dBm) , (1)

where P ref is the average of several samples obtained from the reference line-of-sight
measurement.

The mean and standard deviation were found from

1 N

N
P
i 1
i
norm
(mW) , (2)

and

s
1 N norm

 Pi (dB)   (dB) (dB) ,
N  1 i 1
 (3)

respectively. The median is the value that lies in the middle of the measured values. The
calculation was as follows. First, the collected power samples were ordered by magnitude

Ordered power samples  P1norm  P2norm .....  Pmnorm  Pmnorm

1  Pm  2 .....  PN 1  PN
norm norm norm
, (4)

where N is the total number of samples, and m is the middle value if N is odd or m and m+1 are
the middle two values if N is even. Then, the median value is

M  Pmnorm , if N is odd, and M  [ Pmnorm  Pmnorm

1 ]  2 if N is even. (5)

The collected data were plotted as normalized power level versus the sample number. Note that
the reference level was not always the strongest measured signal level. Also, in a few cases, a
line-of-sight reference was not available, and an average of near-maximum values was utilized as
the reference value instead. This difference in reference level does not affect the standard
deviation of the received-signal level, but does impact the median and the mean.

Histograms were created by use of a bin width of 1 dB, based on normalized received-signal
power levels. The empirical cumulative density function (CDF) was calculated from the same 1
dB bin widths in the following manner:
n

N i
CDF (n)  i 1
, (6)
N

where Ni is the sample count in bin i, n is the current bin, and N is the total number of samples.
The normalized received-signal power levels were left in units of dB for the purposes of
obtaining bin counts for the histograms.

4
Median, mean, and standard deviation statistics were all calculated based on the normalized
values of the measured data. In the present work, the mean values were calculated based on the
linear representation of the signal-power (i.e., milliwatts as opposed to dBm) and then converted
to a logarithmic value, while the standard deviation was calculated directly on the logarithmic
quantities. This approach is commonly used as evident in the literature; for example [9-15]. In
addition, use of the median values rather than the mean is illustrated in [10], [14], and [15]. Use
of the median values helps prevent skewing of the centrality measure by outlying results. All
resulting quantities are reported in decibels. Note that the median decibel value is the same
regardless of whether it was calculated on the linear values and then converted to a decibel
representation, or calculated directly on the decibel values.

2.2. Radio Mapping Using a Narrowband Communications Receiver

We also collected radio-mapping data using a narrowband communications receiver. This

instrument, when combined with NIST-developed post-processing techniques, provides a high-
dynamic-range measurement system that is affordable for most public safety organizations. Part
of our intent was to demonstrate a user-friendly system that could be utilized by public-safety
organizations to assess their own unique propagation environments. These data were collected at
the same time the spectrum analyzer measurements described above were conducted. We carried
the radio transmitter throughout the structures while recording the received signal at a fixed
receive site, as illustrated in Figure 1(a)

The use and calibration of the communications receiver is described in more detail in other
publications, (for example [16]), but is briefly summarized here for convenience. The system,
shown in block diagram form in Figure 2(a), is based on a commercially available
communications receiver and a personal computer (PC) sound card. The receiver is set to its
“upper sideband” mode and is tuned to a frequency slightly below the carrier. In this way, the
receiver acts as a frequency downconverter (Figure 2(b)), transforming the RF signal to baseband
(audio) frequencies. The baseband signal is digitized by use of the sound card and is then post-
processed and graphically displayed, letting the operator know whether a radio signal is present
and what the level of that signal is.

We may observe the upper and lower sidebands of the down-converted signal by setting the
receiver’s center frequency to approximately the middle of its intermediate-frequency (IF)
passband. For example, a signal with a 100 MHz carrier frequency may be measured by a
receiver with a 3 kHz passband by tuning the receiver to 99.9985 MHz. In this case, the receiver
will display the 100 MHz signal at 1.5 kHz (see Figure 2(c)).

The communications receiver has an automatic-gain-control (AGC) circuit whose function is to

control the receiver gain to produce a constant-output-level signal regardless of the input power.
The AGC circuit ensures the receiver circuitry operates in its optimal range. For the receiver we
used, the AGC is active only for signals above a certain power threshold, and does not modify
weak signals (on the order of received powers less than -90 dBm).

5
The purpose of the AGC circuit is to ensure that the receiver will demodulate at as high of a
signal level as possible without overdriving the front end. It does this by altering the received-
signal level to fit the best range for the demodulator. However, for received-power
measurements the signal-level information is exactly what we are trying to determine. Hence, we
need to undo the effects of the AGC modification of the signal level. To extract the signal-level
information from the AGC-modified received signal, we monitor the DC voltage level that
corresponds to the feedback of the AGC circuit. This voltage is directly related to the received-
signal power and may be used to compensate for the AGC. We measure the DC voltage at the
AGC jack on the back panel of the receiver with a digital multimeter having a recording feature.
By synchronizing the recorded signal with the recorded AGC levels, we may determine the
received signal’s level during post-processing.

Once the post-processing steps are carried out, the average received-signal power for each
location in a building may be measured. Note that the received power depends on the antenna
and cabling used with the receiver. However, these effects may be calibrated out to display
system-independent electric field level, enabling easy comparison of measurements from
different measurement systems. In the measurements described here, we used an
omnidirectional, discone antenna for frequencies below 1 GHz.

The communications receiver allows us to determine the range of received-power values that can
be expected for a given structure. In the following section, we instead report on the received-
signal level relative to a reference line-of-sight measurement, as described in the previous
section. Similar to the spectrum analyzer data collected in earlier NIST Technical Notes [5, 7,
17, 18], these data are then processed in terms of mean, median, and standard deviation to
provide channel models for network simulations. Because data are collected continuously, signal
levels can easily be associated with features in the propagation environment. One advantage of
the communications receiver over the spectrum analyzer is its increased dynamic range; that is,
we can detect weaker signals with this system. Also, our receiver system has a lower cost
compared to that of a spectrum analyzer.

2.3. Wideband Excess-Path-Loss Measurements and Time-Delay Spread

2.3.1 VNA-based measurement system

In addition to single-frequency radio-mapping measurements, we studied excess path loss over a

wide frequency band at selected locations within each structure with the use of a vector network
analyzer (VNA). Wideband excess path loss measurements provide measurements of received
signal strength relative to the expected direct-path, free-space signal for a wide frequency band
of 25 MHz to 18 GHz. Our wideband measurements provide a channel transfer function H(f),
and our excess-path-loss is then H(f)2/Hr(f)2, where Hr(f) is the free-space reference. While this
ratio provides the typical notion of excess-path-loss at a particular frequency, the VNA
measurements provide a much richer data set; that is, they include both magnitude and phase
information.

For the measurements reported here, we also present a subset of these measurements covering
frequencies from 700 MHz to 800 MHz. The excess-path-loss measurements complement the
narrowband, continuously recorded received power measured with the spectrum analyzer and
6
communications receiver discussed above by providing the received power over a wide range of
frequencies but at only a few locations. Time-delay spread was calculated from the excess path
loss data in post processing. Root-mean-square (RMS) delay spread is a figure of merit that gives
an indication of the level of multipath interference encountered during the signal transmission.

For the wideband measurements, we used a synthetic-pulse, ultrawideband system based on a

VNA [19]. Figure 3 is a diagram of the measurement system. Here the system is shown
collecting a line-of-sight reference measurement. In practice, the transmitting and receiving
antennas may be separated by significant distances, although they must remain tethered together
by the fiber-optic link. Even though directional horn antennas are shown in Figure 3,
omnidirectional antennas were also used in our measurements, offering insight into antenna
systems most often used in public-safety applications.

In the synthetic-pulse system, the VNA acts as both transmitter and receiver. The transmitting
section of the VNA steps over a wide range of frequencies a single frequency at a time. The
transmitted signal is amplified and fed to a transmitting antenna, as shown in Figure 3. The
received signal is picked up over the air by the receiving antenna and sent back to the VNA via a
fiber-optic cable. The fiber-optic cable minimizes the loss and phase change that would be
associated with RF coaxial cables between the receiving antenna and the transmitting antenna.
We can then reconstruct the time-domain characteristics of the received signal in post-
processing. Because the wideband transmitted signal corresponds to a short-duration pulse in the
time domain, this system lets us measure the transmitted signal, modified by the propagation
path, including losses and multipath reflections that the short-duration pulse experiences as it
travels from the transmitter to the receiver. One advantage of this system is that it provides a
high dynamic range relative to true time-domain-based measurement instruments.

To make the measurement, the vector network analyzer is first calibrated by use of standard
techniques, where known impedance standards are measured. The calibration corrects for the
systematic errors in the response of the fiber-optic system, amplifiers, and any other electronics
used in the measurement. Then, a reference measurement is conducted where the transmitting
and receiving antennas are placed to minimize reflections from the surrounding environment.
This reference measurement allows us to verify that the system is operating correctly and allows
us to determine the far-field response of the antennas and the cables that connect them to the
measurement system. We compare this reference with the frequency response of the antennas
measured separately in the laboratory environment at NIST. The antenna response is
deconvolved in a post-processing step.

Once the measurements have been made, an additional post-processing step is carried out on the
raw VNA measurements to provide clean frequency-domain and time-domain representations.
Our optical links add a large amplitude oscillation to the measured signal, most likely due to the
reflection off a fiber face. Because this oscillation occurs at a low frequency, we are able to
suppress it by applying a high-pass filter, as shown in Figure 4.

Note that we use the phrase “excess path loss” in the context of vector-network-analyzer (VNA)-
based measurements. Technically, we are measuring received signal relative to the transmitted
signal, not path loss. Graphs of path loss would have positive ordinates and increase with

7
distance. However, the phrase “excess path loss” has a specific meaning in the measurement
community and will be used throughout this report.

2.3.2 Excess path loss

Our wideband measurements provide a channel transfer function H(f), where H(f) typically is
derived from the measured transmission parameter S21(f). To find the frequency-dependent path
loss between the transmit and receive antennas, we first compute |H(f)|2/|Hr(f)|2, where Hr(f) is a
free-space reference made at a known distance dr from the transmit antenna. The use of a ratio to
find the path loss enables us to calibrate out the antenna response of the system.

Excess path loss is typically understood to be the loss seen to exceed that measured in a free-
space environment [20]. To find the excess path loss, we reduce the total path loss by the
expected free-space path loss over the overall separation distance d between the transmitting and
receiving antennas. To do this, we divide the measurement of |H(f)|2 by (4πd/λ)2. Equivalently,
we can multiply |H(f)|2/|Hr(f)|2 by (dR/d)2. The distance d may be measured or estimated from
maps, depending on the environment. As stated earlier, this provides the loss in excess of that
which would be measured at the same distance in free space. We note that communication
engineers typically think of excess path loss as a single frequency or narrowband measurement.
However, the VNA measurements provide a much richer data set because they include both
magnitude and phase information over a broad frequency range.

As an example, Figure 5 shows the time-domain response for a reference measurement using a
pair of dual-ridged-guide antennas separated by 3 m. In Figure 5(a), the reference measurement,
transformed to the time domain, is shown with all environmental effects. The reference
measurement is gated (windowed) from 20 ns to 32 ns to isolate the antenna response, which was
determined previously in a separate measurement. The corresponding frequency-domain
response in Figure 5(b) shows a noisier trace when environmental effects are included, compared
with a smoother trace for isolated antennas. The gated response is what we would see if the
antenna were measured in a free-space environment, free from environmental reflections.

2.3.3 RMS delay spread

Root-mean-square (RMS) delay spread is calculated from the power-delay profile of a measured
signal [21-23]. Figure 6 shows the power-delay profile for a typical building propagation
measurement. The peak level usually occurs when the signal arrives at the receiving antenna,
although sometimes we see the signal build up gradually to the peak value and then fall off (the
latter behavior is indicative of a reverberation environment). A common rule of thumb is to
calculate the RMS delay spread by use of signals at least 10 dB above the noise floor of the
measurement. For typical measurements, we define the maximum dynamic range to be about
40 dB below the peak value. However, for the measurement shown in Figure 6, we extended the
window down to 70 dB below the peak value, because the RMS delay spread does not change
appreciably due to the almost constant slope of the power decay curve. Note that the dynamic
range value may change for low signal levels. The following equation is used to define the RMS
delay spread,   :

8
 RMS     
2
  
2
. (5)

In (2),  is defined as the average of the power-delay profile in the defined dynamic range, and
2 is the variance of the power-delay profile.

2.3.4 Measurement set-up

For the measurements reported in this document, the VNA-based synthetic pulse system was set
up with the following parameters. The initial output power was set to -15 dBm to -13 dBm,
across all frequencies, but do not compensate for the frequency dependence of components such
as cables and antennas. The gain of the amplifier and the optical link and the system losses
ensured that the power level at the receiving port was not more than 0 dBm. An intermediate-
frequency (IF) averaging bandwidth of around 1 kHz was used to average the received signal.
We typically used 6401 or 16001 points per frequency band, depending on the frequency band.
For these measurements, the high-band measurements (750 MHz to 18 GHz, using directional
dual ridge guide (DRG) antennas) were taken by measuring 48003 points for a total of three
bands. Only one band was required for the low-frequency measurements (25 MHz to 1.4 GHz
with omnidirectional dicone antennas and 6401 points). The dwell time was approximately 25 μs
per point.

3. Building Structure Descriptions and Experimental Set-up

This section briefly describes the four different large building structures characterized in these
experiments and details the experimental set-up. These structures include a convention center, an
apartment building, a laboratory building with multiple office corridors, and a 57-story office
building. This study includes data gathered with the three different measurement techniques
described above.

3.1 Colorado Convention Center, Denver, CO

The first structure was the Colorado Convention Center. This massive three-level structure is
constructed of reinforced concrete, steel, and standard interior finish materials. The exterior of
the building is a combination of glass, metal, and concrete. Figure 7 and Figure 8 show both
internal and external photos of the convention center. As shown in Figure 9, the convention has a
basement and two above-ground levels. An auditorium was located on the Speer Boulevard side
of the second level. The third level consists of a large open space that can be subdivided by
moveable walls. During the experiment, the space was open and nearly empty.

Our receive sites were located at the street level, which is level two. The three receive sites are
depicted on Figure 9, denoted as RX1, RX2, and RX3, respectively. All three locations were
located between 10 m to 15 m from the building at the closest point. Placement of these receive
sites was intended to simulate the location of emergency response vehicles during a response
scenario.

9
3.2 Republic Plaza, Denver, CO

The Republic Plaza is a 57-story office building in downtown Denver. The construction
materials are a typical combination of concrete and steel. The interior building materials are a
combination of metal framing, drywall, and trim, with stone finishes in the lobby. The exterior is
a combination of glass and metal. Figure 10 and Figure 11 illustrate the exterior and interior of
the building, respectively. In Figure 12, floor plans for several levels are shown.

In this experiment, the three receive sites, depicted in Figure 13 and Figure 14, varied
substantially in distance from the building. Receive site 1 was located on the 17th Street side,
approximately 10 m from the building; receive site 2 was located on the 16th Street side,
approximately 25 m from the building; and finally, receive site 3 was located on the roof of a
parking garage approximately 215 m southwest on Tremont Plaza. These locations were
intended to simulate the locations of command vehicles in an emergency response scenario.

Light blue numbers enclosed by squares in Figure 15 show the locations within the building used
in the synthetic pulse tests. Those locations that correspond to a radio-mapping test location
shown in Figure 12 are indicated as well.

3.3 Apartment Building, Boulder, CO

This building was the 11-story Horizon West apartment building in Boulder, CO (Figure 16).
The building is constructed of reinforced concrete, steel, and brick with standard interior finish
materials. The building was fully furnished and occupied during the experiments. Measurements
were performed during daytime hours and, as a result, people were moving throughout the
building during the experiments. Figure 17 displays some internal photos of the apartment
complex building.

For the radio-mapping experiments, two fixed receive sites (see Figure 18(a)) were assembled on
the east side and north side of the apartment building, approximately 60 m and 80 m from the
apartment building. During this experiment, the transmitters were carried throughout the
building. Measurements were performed with the receive antennas polarized in the vertical
direction. As the received signal was recorded, the location of the transmitters in the apartment
building was also recorded. The locations of the synthetic pulse measurements are shown in
Figure 18(b). These measurements were acquired approximately every 5 m, as indicated in the
figure, on floors two and seven of the building.

3.4 NIST Laboratory, Boulder, CO

This building is the main building (referred to as the Radio Building) at the NIST laboratories in
Boulder, CO. The building is constructed of reinforced concrete and is basically a four-story
building. However, the building is built on a hillside, and consequently, some locations in the
building are below ground level. Measurements were made on the 3rd floor hallway called “Wing
4”, with the radio-mapping continuing in to “Wing 3” on the 3rd floor, and “Wing 5” on the

10
fourth floor. The measurements were performed during daytime hours and, as a result, people
were moving throughout the building during the experiments.

For the radio-mapping experiments, two fixed receiving sites were assembled on the south side
of the laboratory building (see Figure 19 and Figure 20). The receive site at Wing 4 was located
on the loading dock, while the receive site at Wing 6 was approximately 10 m from the building.
During these experiments, the transmitters were carried throughout the laboratory. Measurements
were performed with the receiving antennas polarized in the vertical direction. As the received
signals were recorded, the location of the transmitters in the buildings was also recorded.

Note that in Figure 20 there are two separate paths that were covered during the radio-mapping
data collection process. Path one is reference location→A→B→C→D→C→B→A→reference
location and path 2 is reference location→A→B→C2→B→A→reference location. The section
between the reference location and position B is covered twice in both paths. Figure 21 shows
the locations of the synthetic pulse measurements.

4. Experimental Results

In this section, the measured data (see Appendix II) are presented from several perspectives.
Each subsection discusses a different set of building structure measurements. Plots of the radio-
mapping data, normalized to a line-of-site (LOS) reference level, are provided for the frequency
of 750 MHz, and the appropriate figures containing the data are indicated in the subsection
discussions. Plots of the radio-mapping data statistics are also included. Discrete locations,
where the synthetic-pulse measurements were collected, are marked on the appropriate figures.
We provide excess path loss data covering the frequencies from 700 MHz to 800 MHz as well as
the RMS delay spread associated with each location.

Descriptions of the experimental setups for the various sites were provided earlier in Section 2.
Any unique features at a site that appear to impact the collected data are pointed out in the
appropriate subsection. The common acronyms and abbreviations used in the labeling of the
radio-mapping figures and statistical plots are provided in Table 1 below.

Table 1. Abbreviations used in Statistic Tables and Data Figures.

Acronym or Abbreviation Meaning

Apmt. apartment
Approx. approximate
DRG dual ridged guide horn
Fl., FL, fl. floor
Freq. frequency
LOS line-of-sight
Lx level x or floor x, where x is a number
N.E. northeast
N.W. northwest
Ref. reference signal level
RMS root mean square

11
 Std. dev. standard deviation
S.E. southeast
S.W. southwest
Vert. vertical

4.1 Colorado Convention Center: Measurement Results

The first set of results is from the Colorado Convention Center. All three types of measurements
were performed. Due to high attenuation of the received signals caused by this structure, both the
spectrum analyzer data and the synthetic pulse data were affected. The statistics from the
spectrum analyzer data were skewed by data points below the noise floor of the instrument and
only a few locations in the synthetic pulse tests provided enough signal strength to compute a
meaningful RMS delay spread.

4.1.1. Radio Mapping

The spectrum analyzer radio mapping results for the Colorado Convention Center are shown in
Figure 22 through Figure 24, and the narrowband receiver radio mapping results are shown in
Figure 26 through Figure 28. In all six plots, the line-of-sight reference signal is clearly seen by
the distinct largest peaks in each plot. The features of the results from the two different
measurement systems track well at each of the receive sites. The narrowband receiver results
cover a broader range of signal levels. We see that spectrum analyzer results are at the noise
floor across much of the collected data. However, the narrowband receiver data provides useful
results down to approximately 15 dB below the noise floor of the spectrum analyzer system.

Figure 25 and Figure 29 show histograms and empirical CDFs from the Colorado Convention
Center for data from the spectrum analyzer and narrowband receiver systems, respectively. One
of the most notable features is the step function behavior of the spectrum analyzer CDFs due to
the amount of time the signal was below the noise floor level for that measurement system. The
results from the narrowband receiver are probably a more accurate representation of the
propagation effects due to the increased dynamic range. The high bin count on two of three
narrowband receiver histograms is due to the hard noise floor limit imposed during the data
processing. In other words, the received values below a certain level are all given the same value.
This affects the median value but has little effect on the mean. Note that we need to include these
points because they represent deep fades.

4.1.2 Synthetic Pulse System

The VNA measurements were carried out at the positions indicated in Figure 9, with the VNA
located at receive site 2 (RX2). The VNA results are limited in number due to the significant
signal attenuation experienced while moving into the center of the building. Two different
antenna configurations were used for the measurements; a directional dual ridged guide horn
(DRG) to DRG and a DRG to omnidirectional “tophat” antenna. We also considered two
different frequency bands in data processing: 700 MHz to 800 MHz and 700 MHz to 18 GHz.
The excess path loss plots for these four antenna and frequency band combinations are shown in
Figures 30 through 33.

12
We computed the RMS delay spread from the 700 MHz to 800 MHz excess path loss data. When
the dynamic range falls below a certain threshold, the accuracy of the computed RMS delay
spread value decreases. (See Figure 6 for an explanation of the dynamic range.) The RMS delay
spread is nearly the same across the four combinations of measurements.

In Figure 34, we see the RMS delay spread plotted for each test position, where the positions
illustrated in Figure 9 are located progressively deeper into the Convention Center. Past position
six, the dynamic range was insufficient to allow an accurate calculation of the RMS delay
spread. Positions one through four were measurements made in a large, open, two-story lobby of
the convention center. Here, we see RMS delay spreads of between 100 ns and 150 ns. These
values increase sharply (up to a maximum of 200 ns) once the receiver turns the corner and
proceeds down a large hallway.

4.2 Republic Plaza Measurement Results

The second set of measurement results is from the 57-story Republic Plaza Building in Denver.
All three types of measurements were performed. Because there was less attenuation, we were
able to calculate the RMS delay spread at more positions than we could in the Colorado
Convention Center.

4.2.1. Radio Mapping

The spectrum analyzer radio mapping results for the Republic Plaza Building are shown in
Figure 35 through Figure 37, and the narrowband receiver radio mapping results are shown in
Figure 39 through Figure 41. In the plots for receive sites two and three, the line-of-sight is
clearly seen by the distinct largest peaks in each plot. The basic structure of the two types of
measurements track well at each of the receive sites, and the additional free space path loss for
receive site three is evident in the approximately 18 dB reduction in the reference value as
compared to values obtained at receive sites one and two. As in the Convention Center
measurements, the narrowband receiver results cover a broader signal range; however, the most
notable difference in the signal range covered by spectrum analyzer versus the narrowband
receiver occurs at site three. This is due to the fact that site three was located much further from
the building than either site one or two.

Figure 38 and Figure 42 show the histograms and empirical CDFs of the radio mapping
measurements from the Republic Plaza for the spectrum analyzer and narrowband receiver
systems, respectively. The basic shapes of the empirical CDFs are quite similar for the first two
sites, (see Figure 38 (a), (b) and Figure 42 (a), (b)). The spectrum analyzer empirical CDF for
site three exhibits a much more rapid rise due to the high number of data points at the noise floor
of the measurement system. The histograms for the narrowband receiver all demonstrate a hard
noise floor by the high count in the last bin. As discussed in the Colorado Convention Center
section, hard noise floor does not significantly affect the mean value, but the data are important
because they represent very deep fades in the signal level.

13
4.2.2 Synthetic Pulse System

Synthetic pulse measurements were collected at positions marked in Figure 15, with the VNA
located at receive site 1 (RX1) in Figure 14. We collected excess path loss measurements over
the frequency bands 200 MHz to 180 MHz and 700 MHz to 18 GHz. For the lower-frequency
measurements we used omnidirectional discone antennas, and for the higher-frequency
measurements we used DRG antennas.

We also calculated the RMS delay spread from the excess path loss data. Twenty-one positions
within the building were tested, and eighteen of those positions had sufficient dynamic range to
provide useful results. Figures 43 to Figure 54 show the excess path loss, and Figure 55 shows
the calculated RMS delay spread. In cases where the dynamic range fell below 20 dB, no RMS
delay spread was calculated.

In some cases, the 700 MHz to 800 MHz frequency band data did not exhibit a dynamic range
greater than 20 dB. To estimate the RMS delay spread over this frequency band, we calculated
the RMS delay spread for frequency bands where we did have more than 20 dB of dynamic
range and compared these to what we calculated between 700 MHz and 800 MHz. These
frequency bands were 200 MHz to 1.8 GHz, and 200 MHz to 600 MHz. We also included the
RMS delay spread in the 900 MHz ISM band, with calculation limits between 865 MHz and
965 MHz. As shown in Figure 55, the computed RMS delay spread values for the four frequency
bands track quite well across the eighteen positions, and suggests that results for the other two
frequency bands provide good estimates of the RMS delay spread for the 700 MHz to 800 MHz
band.

Figure 55 shows that the RMS delay spread is lowest at the landing of each floor, where a
window was located. The delay spread does moderately increase as the height increases (from
around 50 ns on floor 1 to around 150 ns on the highest floor). However, when the measurements
were made deeper within the building on floors 5 and 10 (positions 9, 10, and 16), the RMS
delay spread increased significantly (to between 300 and 450 ns, depending on the location and
the frequency band used to calculate the RMS delay spread).

4.3 Boulder, CO Apartment Building Results

The third set of results is from the 11-story Horizon West apartment building in Boulder,
Colorado. This building has been used in other experiments, with results for radio mapping in
other frequency bands found in [18] and synthetic-pulse measurements found in [8].

4.3.1. Radio Mapping

The radio-mapping experiment included sixty-three locations throughout the building, with a
brief description of these locations provided in Table 4. The spectrum analyzer radio mapping
results for the apartment building are shown in Figure 56 and Figure 57, and the narrowband
receiver radio mapping results are shown in Figure 59 and Figure 60. Both sets of figures
illustrate a similar structure, with peaks at the same positions. Note that some difference in
locations is attributable to the manual process by which the positions are recorded during the

14
experiment. In both cases the signal is well above the noise floor, which is evident in the
histograms shown in Figure 58 and Figure 61. The empirical CDFs are quite similar and are
consistent with a Gaussian shaped probability density function.

The overall spread, minimum to maximum, of the measured signals is approximately 60 dB for
both receive sites and both measurement systems. Also, the median and mean values are at least
15 dB closer than the two other building measurements, primarily because the received signals
are well above the noise floor. Even when the transmitters were in the elevator, (locations 61 to
62 on the radio-mapping plots), the received signal did not experience a deep fade.

The difference in relative proximity to the building for receive site one and two at Horizon West
is evident by the difference in the reference levels of approximately 14 dB for both measurement
systems. Generally, the Horizon West apartment building represented a fairly benign propagation
environment, as the measured signals were typically well above the noise floor.

4.3.2 Synthetic Pulse System

Synthetic pulse measurements for this building were carried out previously, and the results,
reproduced from TN 1546 [8], can be found in Figures 62 through 65. The RMS delay spread is
summarized in Figure 66. While these results are not calculated specifically across the 700 MHz
to 800 MHz band, as demonstrated by the earlier results, the expected behavior should be quite
similar even when considering a much wider bandwidth, especially because the signal energy is
greatly reduced at the higher frequencies.

The RMS delay spread values at the apartment building are all in a range below 50 ns. This is in
contrast to those from the other two buildings, which were more massive and had longer
propagation delays between reflections.

4.4 NIST Boulder Laboratory Results

The fourth set of data is from the NIST Laboratory in Boulder, CO. In the radio mapping
experiments, the transmitters were carried over two different paths, shown in Figure 20. Path I
started at the reference location, and proceeded to position D before returning to the reference
location. Location C2 was not included in this path. Path II again started at the reference
location, proceeded to position C2, and then returned to the reference location. Locations C and
D were not covered in this second path. Unlike the other three structures, data were collected for
a repeated walk over each path. Thus, the combination of two collection sites, two walking paths,
and a repeat measurement for each path created four independent data sets for both the spectrum
analyzer and narrowband receiver measurement setups. The repeated walk provides an indication
of the repeatability. The statistics are computed on the data collected between the location A
points marked in the figures. These points are clearly marked by the on/off transition in the
radio-mapping plots. The reference value is taken as the maximum value across all the collected
data, so for site one, this maximum occurs near location A, and for site 2, the maximum occurs at
the reference location indicated in Figure 20. In both cases, the transmitter has line-of-sight
communications with the receive site, which is consistent with the experimental setup and data-
processing approach used for the other three buildings.

15
4.4.1. Radio Mapping

The spectrum analyzer results collected at the Wing 4 (site 1) and Wing 6 (site 2) receive
locations are shown in Figure 67 and Figure 68, respectively, with the repeated Path I walk
results shown for sites 1 and 2 shown in Figure 70 and Figure 71 respectively. The histogram and
empirical CDF plots corresponding to these two walks are shown in Figure 69 and Figure 72,
respectively. The power profile results in Figure 67 and 70 are quite similar, as are the profile
results in Figure 68 and Figure 71. In addition, the shape of the histograms and empirical CDFs
in Figure 69(a) and (b) compare well to those in Figure 72(a) and (b). However, the calculated
statistics indicate differences for the receive site one case, with the first walk-through resulting in
a median value that is 11.8 dB lower than the repeated walk-through.

The difference between these repeat walk-throughs is not due to the repeatability of the
instrumentation: it occurs because of small-scale fading, a phenomenon caused by constructive
and destructive interfering signal reflections from environmental features such as walls, floors,
and metallic objects. Small-signal fading can introduce signal level variations of several decibels
even for measurements made nominally along the same path because it is very position
dependent. However, the large-scale path loss dependence of the path (if the small-scale fading
was smoothed out) represented by the statistics presented here remains consistent. The mean and
the standard deviation are 3.1 dB and 1.3 dB lower for the first walk-through, respectively. Note
that the reference level is within 0.1 dB for the two walks, so the statistics are not being skewed
by normalization with respect to the reference level. The site two cases exhibit a median
difference of 1.7 dB between the two walks of the first path, and difference of 0.9 dB and 0.2 dB
for the mean and standard deviation, respectively. The reference level differs by 1.2 dB.

Path II spectrum analyzer power profiles for the first walk for sites one and two are given in
Figure 73 and Figure 74, respectively, and the repeat walk results are shown in Figure 76 and
Figure 77 for sites one and two, respectively. The corresponding histogram and empirical PDF
plots for the two walks are shown in Figure 75 and Figure 78, respectively. The site one profiles
(Figure 73 and Figure 76), indicate some small differences at the first pass though location B,
and the histogram and empirical CDF plots (Figure 75(a) and Figure 78(a)) show some 20
differences as well. The mean value is 2.4 dB lower for the first walk, but the median is 1.8 dB
lower for the second walk. The standard deviation is 1.7 dB less for the first walk, and the
reference level is 3.1 dB lower for the second walk. The site two profiles for the two walks
(Figure 76 and Figure 77) are quite similar, and the histogram and empirical CDF plots (Figure
75(b) and Figure 78(b)) exhibit similar shapes. The difference in median, mean, standard
deviation, and reference level values is 0.8 dB, 2.5 dB, 1.5 dB and 0.2 dB respectively. As with
the Path I results, the site 2 results track closely between repeated walks.

Figure 79 and Figure 80 show the received power profile for the narrowband receiver at sites one
and two, respectively, for the first path (Path I) and the first walk. Figure 82 and Figure 83 show
the narrowband receiver results at sites one and two, respectively, for the repeated walk of Path I.
The corresponding histogram and empirical CDF plots for the two walks are shown in Figure 81
and Figure 84. All the corresponding power profile, histogram, and empirical CDF plots for the
two walks of Path I are quite similar. A noticeable difference occurs in the statistics for site one,
where the median, mean and standard deviation differ by 7.4 dB, 7.1 dB, and 1.0 dB,

16
respectively (see Figure 81(a) and Figure 84(a)). However, the reference level at site 1 is almost
identical for the two walks of Path I. This behavior is consistent with the spectrum analyzer Path
I results discussed earlier in this section. Site 2 statistics indicate a difference of 0.8 dB, 2.5 dB,
0.2 dB, and 1.5 dB in the median, mean, standard deviation, and reference levels (see Figure
81(b) and Figure 84(b)). These relatively small differences are consistent with the spectrum
analyzer results discussed previously.

Path II narrowband profiles of data collected at sites one and two for the first walk are given in
Figure 85 and Figure 86, respectively, and the profiles for the second walk of Path II are given in
Figure 88 and Figure 89, respectively. The corresponding histogram and empirical CDF plots for
the two walks of Path II are shown in Figure 87 and Figure 90. All the corresponding
narrowband received power profiles match well between repeated walks, (for example, compare
Figures 85 to 88, and Figures 86 to 89). An important feature in all four of these plots is that the
signal level is typically above the noise floor. This is illustrated clearly in the histogram and
empirical CDF plots of Figure 87 and Figure 90. While the general shape of the corresponding
histogram and empirical CDF results in Figure 87(a) and Figure 90(a) are quite similar, the mean
and standard deviation values differ by 5.9 dB and 3.0 dB, respectively. The median and
reference levels differ by 1.9 dB and 0.4 dB, respectively, in Figure 87(a) and Figure 90(a).
Histogram and empirical CDF results for receive site two (Figure 87(b) and Figure 90(b)) are
similar in shape, and the differences in median, mean, standard deviation, and reference levels
are 3.4 dB, 1.2 dB, 1.5 dB, and 0 dB, respectively.

4.4.2 Synthetic Pulse System

We also carried out measurements in the Wing 4 corridor with the VNA-based synthetic-pulse
measurement system described in Section 2.3. Measurements were made at points approximately
every 15.25 meters (50 feet) along the corridor as shown in Figure 21. The transmitting antenna
was located either at site one at the end of Wing 4 or at site two, outside the building by the end
of Wing 6, as shown earlier in Figure 19 and Figure 20. The receiving antenna was moved along
the corridor to the locations marked in Figure 21. In this report, we discuss the building
penetration results; i.e., the transmitting antenna is located at site two. Additional results with the
transmitting antenna at site one are found in [8].

We collected data covering two frequency bands. For the lower-frequency band (25 MHz to
1.3 GHz), we used omnidirectional antennas, as would be used by most emergency response
companies. For the higher-frequency band (750 MHz to 18 GHz) we used both omnidirectional
antennas and a set of directional horn antennas. Since this report is focused on 750 MHz, we
present the complete lower frequency band here (Figure 91 through Figure 97), while the upper
frequency band data are available in [8]. However, we include a plot that compares the
omnidirectional results for both the lower and upper bands (Figure 98).

In Figure 91, the top plot shows the excess path loss with the receiver located at position 1 in
Figure 21. The received signal is greater than the system noise across the frequency band, and
thus allows computation of a RMS delay spread. In contrast, note that by positions 9 and 10,
(Figure 95), the received signal is less than the system noise across much of the frequency band.

17
Thus, only the first eight positions exhibit sufficient power across the frequency band to
calculate the RMS delay spread.

The RMS delay spread calculated from the VNA measurements with the transmitter at site two is
summarized in Figure 98. Two different measurement configurations are shown covering
different frequency ranges, both with omnidirectional antennas. The lower frequency band
indicates a range of 43 ns - 85 ns, while the upper frequency band shows a range of 31 ns - 79 ns.
The greatest difference between the two frequency bands (approximately 25 ns) occurs at
position 4.

5. Summary of Results and Conclusion

Here we summarize the results from all four buildings to provide a straightforward comparison.
Table 2 lists the computed statistics for the spectrum analyzer and narrowband receiver data.
Figures 99 through 101 provide graphical representations of these data. The index column in
Table 2 is used in the figures that follow, where several aspects of the data are compared. Note
that index number pairs of 9 and 10, 11 and 12, 13 and 14, and 15 and 16 represent repeated
walks over the same path. In addition, these four pairs are all from the NIST laboratory building.

18
 Table 2. Aggregate statistics from the radio-mapping experiments for the spectrum
analyzer (SA#) and narrowband radio receiver (R#) tests.
Median Mean Std. Dev. Ref.
Experiment Index (dB) (dB) (dB) (dBm)
Spectrum Analyzer Tests
Colorado Convention Center SA1 1 -52.2 -22.3 9.0 -30.9
Colorado Convention Center SA2 2 -57.4 -23.2 10.7 -26.5
Colorado Convention Center SA3 3 -44.2 -23.4 6.8 -39.4
Republic Plaza SA1 4 -65.2 -23.2 20.0 -33.7
Republic Plaza SA2 5 -63.1 -20.3 20.0 -32.3
Republic Plaza SA3 6 -46.5 -18.2 11.8 -50.2
Horizon West SA1 7 -29.5 -17.7 11.0 -34.5
Horizon West SA2 8 -25.1 -18.3 9.9 -49.0
NIST Laboratory SA1, Path I, Walk 1 9 -98.5 -19.8 28.8 -11.5
NIST Laboratory SA1, Path I, Walk 2 10 -86.7 -16.7 30.1 -11.6
NIST Laboratory SA2, Path I, Walk 1 11 -61.5 -39.6 9.5 -40.7
NIST Laboratory SA2, Path I, Walk 2 12 -63.2 -38.2 11.3 -41.9
NIST Laboratory SA1, Path II, Walk 1 13 -62.8 -18.4 25.2 -11.6
NIST Laboratory SA1, Path II, Walk 2 14 -64.6 -16.0 26.9 -14.7
NIST Laboratory SA2, Path II, Walk 1 15 -61.1 -39.8 10.9 -41.6
NIST Laboratory SA2, Path II, Walk 2 16 -61.9 -37.3 12.4 -41.8
Narrowband Receiver Tests
Colorado Convention Center R1 1 -75.0 -22.6 21.0 -26.0
Colorado Convention Center R2 2 -77.7 -25.1 19.8 -22.1
Colorado Convention Center R3 3 -68.2 -26.1 20.2 -37.2
Republic Plaza R1 4 -68.3 -26.0 20.3 -29.3
Republic Plaza R2 5 -60.0 -18.9 21.5 -32.7
Republic Plaza R3 6 -49.3 -19.9 17.8 -49.8
Horizon West R1 7 -33.2 -20.6 10.8 -31.1
Horizon West R2 8 -27.9 -19.3 9.0 -45.2
NIST Laboratory R1, Path I, Walk 1 9 -93.0 -17.4 29.1 -19.4
NIST Laboratory R1, Path I, Walk 2 10 -85.6 -24.5 28.1 -19.4
NIST Laboratory R2, Path I, Walk 1 11 -71.5 -39.9 16.5 -34.8
NIST Laboratory R2, Path I, Walk 2 12 -70.7 -42.4 16.3 -33.3
NIST Laboratory R1, Path II, Walk 1 13 -64.0 -21.8 23.9 -19.8
NIST Laboratory R1, Path II, Walk 2 14 -65.9 -15.9 26.9 -19.4
NIST Laboratory R2, Path II, Walk 1 15 -62.2 -39.1 15.2 -33.0
NIST Laboratory R2, Path II, Walk 2 16 -65.6 -37.8 16.7 -33.0

Figure 99 plots the standard deviation for the two radio-mapping systems versus all the receive
sites. Excluding the Colorado Convention Center results, the standard deviation values are within
7 dB for the spectrum analyzer and narrowband measurements when compared on a site-by-site
basis. In addition, excluding the results from receive site 3 at Republic Plaza and the first results
for Path I, receive site 2 at the NIST Laboratory, leaves a maximum difference of 5 dB in the
standard deviation between the two systems. Perhaps more importantly, in Figure 100, we see a
large difference between the median and mean values. The Horizon West median and mean are

19
within 12.6 dB for both sites. However, the Colorado Convention Center and the Republic Plaza
results show a difference between the mean and the median ranging from 20.8 dB to 52.6 dB.
The NIST Boulder Laboratory shows differences in the median and mean ranging from 21.3 dB
to 78.7 dB. These wide ranges illustrate the difference between calculating statistics using
logarithmic quantities (the median was selected as the middle value in dB) and the mean (which
was calculated as the mean of the normalized linear power and then converted to dB). The
former gives a better indication of the differences in average received power for the different
sites and different receiver types. Note that the median does not change whether computed on the
logarithmic or linear power quantities.

The mean received signal strengths shown in Figure 100 indicate similar values between
measurement systems and across all sites, with the only significant difference occurring at the
NIST building, site 2. This is due the dominance of the values near 0 dB (i.e., the normalized
reference level) on the calculation of the mean using linear values. A few values near 0 dB can
skew the mean due to the weight of those values when the mean is computed. Excluding the
NIST building, the maximum site-by-site difference for the spectrum analyzer is 2.9 dB (Horizon
West, site 1). The mean values of the measurements taken at the NIST building, site 2, are
approximately 20 dB lower than the mean results for the NIST building, site one. In addition, the
mean values of the measurements taken at NIST, site 2, are 15 dB to 20 dB lower than the mean
of all other receive sites for the three other buildings. This is due to the fact that the processed
data for the NIST building, site 2, only included data that were approximately 20 dB below the
reference value (normalized to 0 dB). From a comparison standpoint, the NIST building, site 1,
results (i.e., experiment index numbers 9, 10, 13, and 14) demonstrate mean behavior similar to
that of the other three buildings.

Figure 101 plots the normalized median value added to the reference value that was used to
normalize the data for each site and building combination (The reference value is measured when
the transmitter is located at the reference location.) The results shown in Figure 101approximate
the average power we would expect to receive when using a system with similar performance
characteristics (receiver sensitivity and antenna systems) and location at the respective receive
sites; for example, an emergency response vehicle located at the receive sites. All the statistics
are computed by use of data normalized by the corresponding reference value. The dynamic
range of the spectrum analyzer measurement system was extended for cases 4 through 16 by
narrowing the search bandwidth. Due to the similarity in dynamic ranges, the difference between
the two measurement systems is now less than 6 dB for cases 4 through 16.

The results presented in this report support a few observations. First, in comparing the two
measurement techniques used in the radio mapping, we observe similar statistical results when
the measured signal is significantly above the noise floor. For example, the statistics are least
similar at the Colorado Convention Center and most similar for Horizon West. At the Colorado
Convention Center, we see the spectrum analyzer results influenced by the increased number of
samples in the noise as compared to the radio receiver. With respect to the uniqueness of the
structures, the mean value does not indicate significant differences between buildings, or even
the receive site locations at a particular building. The median values suggest greater difference in
building behavior, and between the receive sites at a particular building. However, the median
for the narrowband receive data may be skewed by the calibration issue described in Section

20
4.1.1., and by the hard noise floor. A smaller standard deviation value indicates less signal
variability.

Aggregate results for the synthetic-pulse are provided in Table 3 below and Figure 102. Table 3
and Figure 102 shows that the calculation of the RMS delay spread is not significantly affected
by either (a) the pair of antennas used in capturing the data, or (b) the frequency band used in the
processing. Both of these factors allow more flexibility in the measurement process.

As mentioned in Section 4, we can directly trace the changes in the RMS delay spread to physical
features in the structures. An interesting overall observation is that the RMS delay spread was two to
five times larger in the convention center and the skyscraper, as compared to the apartment and the
NIST laboratory/office buildings. The skyscraper and the convention center are both physically
larger buildings than the apartment and NIST buildings. Also, for both the apartment and the NIST
laboratory/office building, most of the measurements were made in a hallway or corridor, while
many of the measurements for the skyscraper and the convention center were made in open areas.

21
Table 3. Aggregate RMS delay spread values and statistics; all values are in nanoseconds
(ns).
Colorado Convention
Building Republic Plaza Horizon West NIST Lab
Center
DRG-Tophat DRG-DRG
Comments Flr. 2 Flr. 7
Antennas Antennas
Index for
Figure 1 2 3 4 5 6 7 8 9 10 11 12
plot →

Frequency bands used in RMS calculation

1000-18000 MHz

1000-18000 MHz

750 - 18000 MHz

700-18000 MHz

700-18000 MHz

200-18000 MHz

25 -1300 MHz
700-800 MHz

700-800 MHz

200-600 MHz

700-800 MHz

865-965 MHz
Position
Index
(Location
in
building)
0 24.1 32.4
1 70.4 94.7 72.3 88.3 62.0 65.2 64.7 70.1 21.4 23.7 44.2 47.9
2 67.8 86.6 90.2 100.1 56.8 63.5 56.7 60.6 16.7 24.3 44.2 32.7
3 104.5 113.3 123.4 94.6 66.9 76.9 72.2 103.4 15.3 14.5 53.7 64.2
4 147.2 170.4 145.0 158.7 89.8 134.9 96.9 136.0 24.9 35.9 51.0 76.2
5 194.0 89.6 108.3 97.0 113.8 48.2 44.2 73.6 76.9
6 74.6 110.4 74.6 107.7 30.7 37.4 75.3 78.7
7 73.9 140.3 79.6 148.1 36.3 40.8 71.4 64.7
8 96.1 143.5 94.4 165.6 31.2 38.8 84.8 63.6
9 326.0 416.7 344.7 446.5 30.4 30.5
10 384.1 235.9 376.6 405.6 31.7 31.0
11 148.5 184.0 146.4 206.9 39.4 34.8
12 153.6 244.4 140.6 232.6 38.0 36.8
13 131.6 209.8 128.1 222.3 20.8
14 172.3 235.2 182.8 246.3
15 181.7 292.6 171.3 277.0
16 303.6 409.4 285.4 350.6
17 198.5 288.7 198.7 303.0
18 158.2 215.2 174.8 254.9

Mean 116.8 116.3 107.7 110.4 153.8 198.6 154.8 213.9 29.2 32.7 62.3 63.1
Median 104.5 104.0 106.8 97.3 140.0 196.9 134.4 214.6 30.6 34.8 62.6 64.5
Minimum 67.8 86.6 72.3 88.3 56.8 63.5 56.7 60.6 15.3 14.5 44.2 32.7
Maximum 194.0 170.4 145.0 158.7 384.1 416.7 376.6 446.5 48.2 44.2 84.8 78.7

—————————

22
Disclaimer: Mention of any company names serves only for identification, and does not
constitute or imply endorsement of such a company or of its products.

—————————

This work was sponsored by the U.S. Department of Justice through the Public Safety
Communications Research Lab of NIST. We thank members of the technical staff of the
Electromagnetics Division 818, who collected the measurements, and Dennis Friday, Mike
Kelley, Perry Wilson, and Dereck Orr for programmatic support.

23
6. References

[1] Statement of Requirements: Background on Public Safety Wireless Communications,

The SAFECOM Program, Department of Homeland Security, Vol. 1, March 10, 2004.

[2] M. Worrell and A. MacFarlane, Phoenix Fire Department Radio System Safety Project,
Phoenix Fire Dept. Final Report, Oct. 8, 2004,
http://www.ci.phoenix.az.us/FIRE/radioreport.pdf

[3] 9/11 Commission Report, National Commission on Terrorist Attacks Upon the United
States, 2004.

[4] Final Report for September 11, 2001 New York World Trade Center terrorist attack,
Wireless Emergency Response Team (WERT), Oct. 2001.

[5] C.L. Holloway, G. Koepke, D. Camell, K.A. Remley, D.F. Williams, S. Schima, S.
Canales, and D.T. Tamura, “Propagation and Detection of Radio Signals Before, During
and After the Implosion of a Thirteen Story Apartment Building,” NIST Technical Note
1540, Boulder, CO, May 2005.

[6] C.L. Holloway, G. Koepke, D. Camell, K.A. Remley, and D.F. Williams, “Radio
Propagation Measurements During a Building Collapse: Applications for First
Responders,” Proc. Intl. Symp. Advanced Radio Tech., Boulder, CO, March 2005, pp.
61-63.

[7] C.L. Holloway, G. Koepke, D. Camell, K.A. Remley, D.F. Williams, S. Schima, and
D.T. Tamura, “Propagation and Detection of Radio Signals Before, During and After the
Implosion of a Large Sports Stadium (Veterans' Stadium in Philadelphia),” NIST
Technical Note 1541, Boulder, CO, October 2005.

[8 ] K. A. Remley, G. Koepke, C.L. Holloway, C, Grosvenor, D. Camell, J. Ladbury, R. T.

Johnk, D. Novotny, W. F. Young, G Hough, M. C. McKinley Y. Becquet, J Korsnes,
“Measurements to Support Broadband Modulated-Signal Radio Transmissions for the
Public-Safety Sector, “ NIST Technical Note 1546, Boulder CO, April 2008.

[9] L.P. Rice, “Radio transmission into buildings at 35 and 150 mc,” Bell Syst. Tech. J., pp.
197-210, Jan. 1959.

[10] E. Walker, “Penetration of radio signals into building in the cellular radio environment,”
Bell Syst. Tech. J., 62(9), Nov. 1983.

[11] D. Molkdar, “Review on radio propagation into and within buildings,” IEE Proceeding-
H, 38(1): 61-73; Feb. 1991.

24
[12] W.J. Tanis and G.J. Pilato, “Building penetration characteristics of 880 MHz and 1922
MHz radio waves,” Proc. 43th IEEE Veh. Technol. Conf., Secaucus, NJ, 18-20 May
1993, pp. 206-209.

[13] L.H. Loew, Y. Lo, M.G. Lafin, and E.E. Pol, “Building penetration measurements from
low-height base stations at 912, 1920, and 5990 MHz,” NTIA Report 95-325, National
Telecommunications and Information Administration, Sept. 1995.

[14] A. Davidson and C. Hill, “Measurement of building penetration into medium buildings at
900 and 1500 MHz,” IEEE Trans. Veh. Technol., 46(1): 161-168; Feb. 1997.

[15] E. F. T. Martijn and M. H. A. J. Herben, “Characterization of Radio Wave Propagation

Into Buildings at 1800 MHz,” IEEE Ant. and Wireless Prop. Letters, Vol. 2, 2003.

[16] M. Rütschlin, K. A. Remley, R. T. Johnk, D. F. Williams, G. Koepke, C. Holloway, A.

MacFarlane, and M. Worrell, “Measurement of weak signals using a communications
receiver system,” Proc. Intl. Symp. Advanced Radio Tech., Boulder, CO, March 2005,
pp. 199-204.

[17] C.L. Holloway, G. Koepke, D. Camell, K.A. Remley, D.F. Williams, S. Schima, M.
McKinely, and R.T. Johnk, “Propagation and Detection of Radio Signals Before, During
and After the Implosion of a Large Convention Center,'” NIST Technical Note 1542,
Boulder, CO, June 2006.

[18] C.L. Holloway, W. Young, G. Koepke, D. Camell, Y. Becquet, K.A. Remley,

“Attenuation, Coupling, and Variability of Radio Wave Signals Into and Throughout
Twelve Large Building Structures,'' NIST Technical Note 1545, Boulder, CO, Aug 2008.

[19] B. Davis, C. Grosvenor, R.T. Johnk, D. Novotny, J. Baker-Jarvis, M. Janezic,

“Complex permittivity of planar building materials measured with an ultra-wideband
free-field antenna measurement system,” Natl. Inst. Stand. Technol. J. Res., 112(1):67-
73, Jan.-Feb., 2007.

[20] M. Riback, J. Medbo, J. Berg, F. Harryson, H. Asplund, “Carrier frequency effects on

path loss,” Proc. 63rd IEEE Vehic. Technol. Conf., Vol. 6, pp. 2717-2721, 2006.

[21] J.C.-I. Chuang, “The effects of time delay spread on portable radio communications
channels with digital modulation,” IEEE J. Selected Areas in Comm., SAC-5(5): 879-
889, June 1987.

[22] Y. Oda, R. Tsuchihashi, K. Tsuenekawa, M. Hata, “Measured path loss and multipath
propagation characteristics in UHF and microwave frequency bands for urban mobile
communications,” Proc. 53rd IEEE Vehic. Technol. Conf., Vol. 1, pp. 337-341, May
2001.

[23] J.A. Wepman, J.R. Hoffman, L.H. Loew, "Impulse Response Measurements in the 1850-
1990 MHz Band in Large Outdoor Cells", NTIA Report 94-309, June 1994.

25
Appendix I: Experiment Setups and Locations

x x x x
x x
receiver x x x
x x x x
x x x
x x
x x x x x x

transmitter path
(a)

Discone Antenna
50 MHz to 1 GHz

(b)

Spectrum Analyzer
(c)

Figure 1. Illustration of the radio-mapping measurements. (a) A radio transmitter was carried
throughout the building on a continuous path, and the signal strength was recorded by the
receiver. (b) For one set of radio mapping experiments, the receiver consisted of a commercially
available communications receiver, followed by post-processing steps to determine the received-
signal level. (c) A spectrum analyzer and laptop made up the other data collection method. A
separate receive antenna was used for the two data collection setups.

26
 receiver

(a)

(b) (c)

Figure 2. (a) Set-up used to enable a communications receiver to display received signal strength.
The receiver down-converts the signal to baseband frequencies, as shown in (b), and the sound
card digitizes the baseband signal. A DC voltmeter monitors the automatic-gain-control setting
on the receiver. The AGC data are used in post-processing to determine the signal’s actual level.
Graphic (c) shows a recorded, down-converted signal with time on the x-axis and frequency on
the y-axis. The stripe at the center corresponds to the carrier frequency, and the bands on either
side correspond to the FM sidebands. In this case the received signal is in Morse code.

27
 Receiving Transmitting
antenna antenna

tripods
62 inches
VNA

Port 1 Port 4

Ground Plane

Fiber Optic Fiber Optic

Transmitter Receiver
RF Optical
200 m
RF Optical
Optical
Fiber

Figure 3. Synthetic-pulse measurement system based on a vector network analyzer. Frequency-

domain measurements, synchronized by the optical fiber link, are transformed to the time
domain in post-processing. This enables determination of excess path loss, time-delay spread,
and other figures of merit important in characterizing broadband modulated-signal transmissions.

-4
x 10 Signal distored by optical link
2

1
power

-1

-2
0 1 2 3 4 5
time [ns] 4
x 10
-4
x 10 Distortion removed by lowpass filter
1

0.5
power

-0.5

-1
0 1 2 3 4 5
time [ns] 4
x 10

Figure 4. Low-frequency oscillations introduced by the optical link (top) are removed by use of a
high-pass filter in post-processing (bottom). This enables us to clearly discern the measured data;
here an initial pulse is received at approximately 5 ns, followed by a reflected pulse at
approximately 30 ns.

28
 1.2

Antenna response
0.8

0.4
Ground bounce

S21 amplitude (V)

0

Spurious
environmental
-0.4 effects

-0.8

-1.2

20 30 40 50
Time (ns)

(a)
20

10
S21 Magnitude (dB)

-10

DRG (3 m reference)
Ungated response
Gated Response
-20

0 4000 8000 12000 16000 20000

Frequency (MHz)

(b)

Figure 5. (a) Time-domain waveform for a dual ridge guide (DRG) horn antenna 3 m reference
measurement. The waveform shows the antenna response, the ground-bounce response and the
spurious environmental effects. (b) The frequency-domain response for both the ungated
response (noisy black trace), which includes all environmental effects, and the gated response
(smoother red trace), which includes only the antenna response.

29
 0

Peak level

-50 Maximum
Dynamic Range

Power Delay Profile (dB)

-100

-150

Mean
Delay Spread
-200

RMS
Delay Spread
-250

0 400 800 1200

Time (ns)

Figure 6. Power-delay profile for a building propagation measurement. The important parameters
for a measured propagation signal are the peak level, the maximum dynamic range, the mean
delay spread, and the RMS delay spread.

30
Figure 7. Photographs of the Colorado Convention Center.

31
Figure 8. Photographs of the Colorado Convention Center (cont.).

32
 1 6
2
5
3
4

Figure 9. Layout of the Colorado Convention Center including the radio-mapping path.
Numbers in boxes represent synthetic pulse measurement locations; the non-enclosed
numbers denote positions indicated on the spectrum analyzer and narrowband receiver
measurements.

33
Figure 10. Photographs of the Republic Plaza Building.

34
Figure 11. Photographs inside the Republic Plaza Building.

35
Figure 12. Layout of the Republic Plaza36
Building and radio-mapping path.
Figure 13. Photographs of the three receive sites at the Republic Plaza Building.

37
 North

RX1

RX2

RX3

Figure 14. Location of receive sites at the Republic Plaza Building.

38
 17

21
18 16 20

19

Figure 15. Synthetic pulse test points in Republic Plaza. The test locations are the blue
numbers in the black boxes; the unboxed numbers are radio-mapping locations as in
Figure 12.

39
Figure 16. Photographs of the Horizon West apartment building in Boulder, CO.

40
Figure 17. Photographs inside the Horizon West apartment building in Boulder, CO.

41
 (a)

0 5 6 7 8 9 10 11 12 13

2
Elevators
3

Receive
Site 1,
VNA
(b)

Figure 18. Receiver locations for the Horizon West apartment building. (a) Locations for radio-
mapping tests. (b) Locations for synthetic pulse system measurements.

42
Figure 19. The receiver setups for the NIST’s laboratories in Boulder, CO.

43
 Reference Location
Wing 4, Receive Site 1
(used as reference for
receive site 2)
A
Note: Location A used as
reference for receive site 1

B
Wing 6, Receive Site 2

C2

Figure 20. Receiver location and layout for the Boulder, CO laboratories.

44
Figure 21. Synthetic pulse building penetration measurements were carried out at specific test
locations indicated on the outline. The Wing 4 corridor is lined with offices having windows to
the outside, thus indoor-to-outdoor coupling can be expected.

45
Appendix II: Measurement Results

position 1 (LOS 1)

12000
position 4
position 3
position 28
position 27

11000
position 26

position 25

10000
position 24
position 23

9000
position 21
position 20

8000
position 19

Approx. noise fl. = -53 dB

position 18

Sample Number
position 17

Ref. = -30.9 dBm

Freq. = 749 MHz
position 2

7000
position 16

position 15
position 14

6000
position 13

position 12 (LOS 2)

position 11

5000
position 10

position 9
position 8
4000

position 7
3000

position 5
position 4
position 3
2000

position 2

position 1 (LOS 1)
20

10

-10

-20

-30

-40

-50

-60

Power (dB)

Figure 22. Colorado Convention Center, receive site 1. Normalized received-signal

data from the spectrum analyzer as the 749 MHz transmitter is carried through the
building.

46
 3000
position 1 (LOS 1)
position 4
position 3

position 28
position 27
position 26

2500
position 25

position 24
position 23

position 21

2000
position 20
position 19

Approx. noise fl. = -59 dB

Sample Number
Ref. = -26.5 dBm
Freq. = 749 MHz
position 17
position 2

position 16

1500
position 15

position 14
position 13

position 12 (LOS 2)

position 7
position 10
1000
position 9
position 8

position 7

position 6 (LOS 3)

position 5
500

position 4
position 3
position 2
20

10

-10

-20

-30

-40

-50

-60

-70

Power (dB)

Figure 23. Colorado Convention Center, receive site 2. Normalized received-signal data
from the spectrum analyzer as the 749 MHz transmitter is carried through the building.
47
 position 1 (LOS 1)
position 4
position 3

12000
position 28
position 27
position 26

position 25

position 24
position 23

10000
position 21
position 20
position 19

Approx. noise fl. = -46 dB

position 17

8000
Sample Number
position 2 position 2

Freq. = 749 MHz

Ref. = -39.4 dBm
position 16

position 15
position 14
position 13

6000
position 12 (LOS 2)

position 11
position 10

position 9
position 8
4000

position 7

position 6 (LOS 3)
position 5
position 4
position 3
2000

position 2

position 1 (LOS 1)
20

10

-10

-20

-30

-40

-50

-60

Power (dB)

Figure 24. Colorado Convention Center, receive site 3. Normalized received-signal data
from the spectrum analyzer as the 749 MHz transmitter is carried through the building.

48
 2100 1
1680 median = -52.2 dB 0.8
Histogram mean = -22.3 dB
1260 std. dev. = 9.0 dB 0.6
Empirical CDF freq. = 749 MHz
840 ref. = -30.9 dBm 0.4
420 0.2
0 0
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20
(a)
525 1
number of values [N]

420 median = -57.4 dB 0.8

empircal F(x)
mean = -23.2 dB
315 Histogram std. dev. = 10.7 dB 0.6
Empirical CDF freq. = 749 MHz
210 ref. = -26.5 dBm 0.4
105 0.2
0 0
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20
(b)
2100 1
1680 Histogram median = -44.2 dB 0.8
mean = -23.4 dB
1260 Empirical CDF std. dev. = 6.8 dB 0.6
freq. = 749 MHz
840 ref. = -39.4 dBm 0.4
420 0.2
0 0
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20
Power [dB]
(c)

Figure 25. Histogram and empirical CDF of the received signal power from the spectrum
analyzer at the Colorado Convention Center for (a) receive site one, (b) receive site two, and (c)
receive site three. Note that histogram count scale is different for receive site two, plot (b).

49
 location 1
location 4
location 3

1500
location 28
location 27
location 26

location 25

Approx. noise fl. = -105 dB

location 24

location 23
location 22

location 21

1000
location 20
location 19

location 18
location 17

Ref. = -26.0 dBm

Time (s)
location 2
location 16
location 15
location 14
location 13

location 12

location 11

500
location 10
Freq. = 749 MHz
location 9
location 8

location 7

location 6

location 5
location 4
location 3
location 2

location 1
0
20

-20

-40

-60

-80

-100

Power (dB)

Figure 26. Colorado Convention Center narrowband receiver results for receive site 1.

50
 location 1
location 4

1500
location 3

location 28
location 27
location 26

Approx. noise fl. = -108 dB

location 25

location 24
location 23

location 22

location 21

1000
location 20
location 19

location 18

Ref. = -22.1 dBm

location 2

Time (s)
location 16

location 15

location 14
location 13

location 12

location 11

500
location 10

Freq. = 749 MHz

location 9
location 8

location 7

location 6

location 5
location 4
location 3

location 2

location 1
0
20

-20

-40

-60

-80

-100

Power (dB)

Figure 27. Colorado Convention Center narrowband receiver results for receive site 2.

51
 location 1
location 4

1500
location 3

location 28
location 27
location 26

Approx. noise fl. = -93 dB

location 25

location 24
location 23

location 22

location 21

1000
location 20

location 17

Ref. = -37.2 dBm

location 2

Time (s)
location 16

location 15
location 14
location 13

location 12

location 11

500
location 10

Freq. = 749 MHz

location 9
location 8

location 7

location 6

location 5
location 4
location 3
location 2

location 1
0
20

-20

-40

-60

-80

-100

Power (dB)

Figure 28. Colorado Convention Center narrowband receiver results for receive site 3.

52
 11000 1
8800 median = -75.0 dB 0.8
mean = -22.6 dB
6600 Histogram std. dev. = 21.0 dB 0.6
Empirical CDF freq. = 749 MHz
4400 ref. = -26.0 dBm 0.4
2200 0.2
0 0
-100 -80 -60 -40 -20 0 20
(a)
5500 1
number of values [N]

4400 median = -77.7 dB 0.8

empircal F(x)
mean = -25.1 dB
3300 std. dev. = 19.8 dB 0.6
Histogram freq. = 749 MHz
2200 Empirical CDF ref. = -22.1 dBm 0.4
1100 0.2
0 0
-100 -80 -60 -40 -20 0 20
(b)
11000 1
8800 median = -68.2 dB 0.8
mean = -26.1 dB
6600 std. dev. = 20.2 dB 0.6
Histogram freq. = 749 MHz
4400 Empirical CDF ref. = -37.2 dBm 0.4
2200 0.2
0 0
-100 -80 -60 -40 -20 0 20
Power [dB]
(c)

Figure 29. Histogram and empirical CDF of the narrowband receiver signal power at the
Colorado Convention Center for (a) receive site one, (b) receive site two, and (c) receive site
three. Note that histogram count scale is different for the receive site two, plot (b).

53
 Convention Center, Directional TX and RX Antennas Convention Center, Directional TX and RX Antennas
20 20
Position 1 Position 2
0 Noise Pos 4 0 Noise Pos 4
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Convention Center, Directional TX and RX Antennas Convention Center, Directional TX and RX Antennas
20 20
Position 3 Position 4
0 Noise Pos 4 0 Noise Pos 4
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Convention Center, Directional TX and RX Antennas Convention Center, Directional TX and RX Antennas
20 20
Position 5 Position 6
0 Noise Pos 4 0 Noise Pos 4
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Figure 30. Excess path loss measurements at the Colorado Convention Center for the 700
MHz to 800 MHz frequency band. Directional transmit and receive antennas were used.

54
 Convention Center, Directional TX and RX Antennas Convention Center, Directional TX and RX Antennas
20 20
Position 1 Position 2
0 Noise Pos 4 0 Noise Pos 4
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Convention Center, Directional TX and RX Antennas Convention Center, Directional TX and RX Antennas
20 20
Position 3 Position 4
0 Noise Pos 4 0 Noise Pos 4
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Convention Center, Directional TX and RX Antennas Convention Center, Directional TX and RX Antennas
20 20
Position 5 Position 6
0 Noise Pos 4 0 Noise Pos 4
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Figure 31. Excess path loss measurements at the Colorado Convention Center for the 700
MHz to 18 GHz frequency band. Directional transmit and receive antennas were used.

55
 Convention Center, Directional TX and Omnidirectional RX Antennas Convention Center, Directional TX and Omnidirectional RX Antennas
20 20
Position 1 Position 2
0 Noise Pos 5 0 Noise Pos 5
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Convention Center, Directional TX and Omnidirectional RX Antennas Convention Center, Directional TX and Omnidirectional RX Antennas
20 20
Position 3 Position 4
0 Noise Pos 5 0 Noise Pos 5
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Convention Center, Directional TX and Omnidirectional RX Antennas Convention Center, Directional TX and Omnidirectional RX Antennas
20 20
Position 5 Position 6
0 Noise Pos 5 0 Noise Pos 5
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Figure 32. Excess path loss measurements at the Colorado Convention Center for the 700 to 800
MHz frequency band. Directional transmit antenna and omnidirectional receive antennas were
used.

56
 Convention Center, Directional TX and Omnidirectional RX Antennas Convention Center, Directional TX and Omnidirectional RX Antennas
20 20
Position 1 Position 2
0 Noise Pos 5 0 Noise Pos 5
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Convention Center, Directional TX and Omnidirectional RX Antennas Convention Center, Directional TX and Omnidirectional RX Antennas
20 20
Position 3 Position 4
0 Noise Pos 5 0 Noise Pos 5
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Convention Center, Directional TX and Omnidirectional RX Antennas Convention Center, Directional TX and Omnidirectional RX Antennas
20 20
Position 5 Position 6
0 Noise Pos 5 0 Noise Pos 5
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Figure 33. Excess path loss measurements at the Colorado Convention Center for the 700 MHz
to 18 GHz frequency band. Directional transmit antenna and omnidirectional receive antennas
were used.

57
 250

200
RMS delay spread [ns]

150 DRG-Tophat 700-800MHz

DRG-Tophat 700-18000MHz
DRG-DRG 700-800MHz
100 DRG-DRG 700-18000MHz

50

0
0 1 2 3 4 5 6
Position

Figure 34. Colorado Convention Center RMS delay spread versus position for two antenna
configurations and two different frequency bands. “DRG” refers to dual ridge guide horn, and is
a directional antenna. “Tophat” refers to an omnidirectional receive antenna.

58
 location 39

location 38
location 37

7000
location 36
location 35
location 34
location 33

location 32

6000
location 31
location 30
location 29
location 28
location 27
location 26
location 25
location 24

5000
location 23
location 22

Approx. noise fl. = -89 dB

Sample Number
Ref. = -33.7 dBm
Freq. = 749 MHz
location 21
location 20
location 19

4000
location 18
location 17

location 16
location 15
location 14

location 13

3000
location 12
location 11
location 10
location 9
location 8
location 7
2000

location 6

location 5

location 4

location 3
location 2
1000

location 1
20

-20

-40

-60

-80

-100

Power (dB)

Figure 35. Republic Plaza, receive site 1. Normalized received signal power from the spectrum
analyzer as the 749 MHz transmitter is carried through the building.

59
 location 39

location 38
location 37

7000
location 36
location 35
location 34
location 33
location 32

location 31

6000
location 30
location 29
location 28 location 27
location 26
location 25
location 24

5000
location 23
location 22

Approx. noise fl. = -90 dB

location 21

Sample Number
Ref. = -32.3 dBm
Freq. = 749 MHz
location 20
location 19
location 18

location 17

4000
location 16

location 14

location 13
location 12
location 11
3000
location 10

location 9
2000

location 4

location 3
location 2

location 1
1000
20

-20

-40

-60

-80

-100

Power (dB)

Figure 36. Republic Plaza, receive site 2. Normalized received signal power from the
spectrum analyzer as the 749 MHz transmitter is carried through the building.

60
 location 39

location 38
location 37

6000
location 36
location 35
location 34

location 32

5000
location 31

location 30
location 29
location 28
location 27
location 26
location 25
location 24
location 23

4000
Approx. noise fl. = -50 dB
location 22

Sample Number
location 21

Ref. = -50.2 dBm

Freq. = 749 MHz
location 20
location 19
location 18
location 17
location 16

3000
location 15
location 14

location 13
location 12
location 11
location 10
location 9 2000
location 8
location 8
location 7
location 6

location 5

location 4
1000

location 3
location 2
location 1
location 1
20

10

-10

-20

-30

-40

-50

-60

-70

Power (dB)

Figure 37. Republic Plaza, receive site 3. Normalized received signal power from the
spectrum analyzer as the 749 MHz transmitter is carried through the building.

61
 200 1
160 median = -65.2 dB 0.8
mean = -23.2 dB
120 Histogram std. dev. = 20.0 dB 0.6
Empirical CDF freq. = 749 MHz
80 ref. = -33.7 dBm 0.4
40 0.2
0 0
-100 -80 -60 -40 -20 0 20
(a)
200 1
number of values [N]

160 median = -63.1 dB 0.8

empircal F(x)
Histogram mean = -20.3 dB
120 std. dev. = 20.0 dB 0.6
Empirical CDF freq. = 749 MHz
80 ref. = -32.3 dBm 0.4
40 0.2
0 0
-100 -80 -60 -40 -20 0 20
(b)
800 1
640 Histogram median = -46.5 dB 0.8
mean = -18.2 dB
480 Empirical CDF std. dev. = 11.8 dB 0.6
freq. = 749 MHz
320 ref. = -50.2 dBm 0.4
160 0.2
0 0
-100 -80 -40-60 -20 0 20
Power [dB]
(c)
Figure 38. Histogram and empirical CDF of the received spectrum analyzer signal power at
Republic Plaza for (a) receive site one, (b) receive site two, and (c) receive site three. Note that
histogram count and the power scales are different for the receive site three, plot (c).

62
 location 39

2000
location 38
location 37

1800
location 36
location 35
location 34

Approx. noise fl. = -101 dB

location 33
location 32

1600
location 31

location 30
location 29
location 28

1400
location 27
location 26
location 25

location 24

1200
location 23
location 22
location 21

Ref. = -29.3 dBm

location 20

Time (s)
location 19

1000
location 18

location 17
location 16
location 15
location 14

800
location 13
location 12

Freq. = 749 MHz

location 11

600
location 10

location 9
location 8
location 7

location 6 400

location 5

location 4
200

location 3
location 2

location 1
0
20

-20

-40

-60

-80

-100

Power (dB)

Figure 39. Republic Plaza, receive site 1. Normalized narrowband receiver signal power
as 749 MHz transmitter is carried through the building.

63
 location 39

2000
location 38
location 37

1800
location 36
location 35

Approx. noise fl. = -98 dB

location 34
location 33
location 32

1600
location 31

location 30
location 29
location 28
location 27

1400
location 26
location 25
location 24
location 23

1200
location 22
location 21

Ref. = -32.7 dBm

Time (s)
location 20
location 19

1000
location 18
location 17
location 16
location 15
location 14

800
location 13
location 12
location 11
location 10
Freq. = 749 MHz

600
location 9
location 8
400

location 4
200

location 3
location 2

location 1
0
20

-20

-40

-60

-80

-100

Power (dB)

Figure 40. Republic Plaza, receive site 2. Normalized narrowband receiver signal power as
749 MHz transmitter is carried through the building.

64
 location 39

2000
location 38
location 37

1800
location 36
location 35

Approx. noise fl. = -81 dB

location 34
location 33
location 32

1600
location 31

location 30
location 29
location 28
location 27

1400
location 26
location 25
location 24
location 23

1200
location 22
location 21

Ref. = -49.8 dBm

location 20

Time (s)
location 19
location 18

1000
location 17
location 16
location 15
location 14

800
location 13
location 12
location 11
location 10

Freq. = 749 MHz

600
location 9
location 8
location 7
location 6 400

location 5

location 4
200

location 3
location 2

location 1
0
20

10

-10

-20

-30

-40

-50

-60

-70

-80

-90

Power (dB)

Figure 41. Republic Plaza, receive site 3. Normalized narrowband receiver signal power as
749 MHz transmitter is carried through the building.

65
 6000 1
4800 Histogram median = -68.3 dB 0.8
Empirical CDF mean = -26.0 dB
3600 std. dev. = 20.3 dB 0.6
freq. = 749 MHz
2400 ref. = -29.3 dBm 0.4
1200 0.2
0 0
-100 -80 -60 -40 -20 0 20
(a)
6000 1
num ber of values [N ]

4800 Histogram median = -60.0 dB 0.8

em pircal F(x)
Empirical CDF mean = -18.9 dB
3600 std. dev. = 21.5 dB 0.6
freq. = 749 MHz
2400 ref. = -32.7 dBm 0.4
1200 0.2
0 0
-100 -80 -60 -40 -20 0 20
(b)
6000 1
Histogram
4800 Empirical CDF median = -49.3 dB 0.8
mean = -19.9 dB
3600 std. dev. = 17.8 dB 0.6
freq. = 749 MHz
2400 ref. = -49.8 dBm 0.4
1200 0.2
0 0
-100 -80 -60 -40 -20 0 20
Power [dB]
(c)

Figure 42. Histogram and empirical CDF of the narrowband receiver signal power at Republic
Plaza for (a) receive site one, (b) receive site two, and (c) receive site three. Note that the
discontinuities near the centers of the histograms are not due to the environment, but rather an
artifact of the testing equipment.

66
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 1 Position 2
0 Noise Pos 1 0 Noise Pos 1
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 3 Position 4
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 5 Position 6
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Figure 43. Republic Plaza 200 MHz to 1800 MHz VNA excess path loss measurements for
positions 1 through 6.

67
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 7 Position 8
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 9 Position 10
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 11 Position 12
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Figure 44. Republic Plaza 200 MHz to 1800 MHz VNA excess path loss measurements for
positions 7 through 12.

68
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 13 Position 14
0 Noise Pos 15 0 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 15 Position 16
0 Noise Pos 15 0 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
20 20
Position 17 Position 18
0 Noise Pos 15 0 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
200 400 600 800 1000 1200 1400 1600 1800 200 400 600 800 1000 1200 1400 1600 1800
Frequency (MHz) Frequency (MHz)

Figure 45. Republic Plaza 200 MHz to 1800 MHz VNA excess path loss measurements for
positions 13 through 18.

69
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 1 Position 2
-20 Noise Pos 1 -20 Noise Pos 1
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 3 Position 4
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 5 Position 6
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Figure 46. Republic Plaza 700 to 800 MHz VNA excess path loss measurements for
positions 1 through 6.

70
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 7 Position 8
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 9 Position 10
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 11 Position 12
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Figure 47. Republic Plaza 700 to 800 MHz VNA excess path loss measurements for
positions 7 through 12.

71
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 13 Position 14
-20 Noise Pos 15 -20 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 15 Position 16
-20 Noise Pos 15 -20 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 17 Position 18
-20 Noise Pos 15 -20 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
700 710 720 730 740 750 760 770 780 790 800 700 710 720 730 740 750 760 770 780 790 800
Frequency (MHz) Frequency (MHz)

Figure 48. Republic Plaza 700 to 800 MHz VNA excess path loss measurements for
positions 13 through 18.

72
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 1 Position 2
-20 Noise Pos 1 -20 Noise Pos 1
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 3 Position 4
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 5 Position 6
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Figure 49. Republic Plaza 200 to 600 MHz VNA excess path loss measurements for
positions 1 through 6.

73
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 7 Position 8
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 9 Position 10
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 11 Position 12
-20 Noise Pos 6 -20 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Figure 50. Republic Plaza 200 to 600 MHz VNA excess path loss measurements for
positions 7 through 12.

74
 Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 13 Position 14
-20 Noise Pos 15 -20 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 15 Position 16
-20 Noise Pos 15 -20 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Republic Plaza, Directional TX and RX antennas Republic Plaza, Directional TX and RX antennas
0 0
Position 17 Position 18
-20 Noise Pos 15 -20 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-40 -40

-60 -60

-80 -80

-100 -100
200 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600
Frequency (MHz) Frequency (MHz)

Figure 51. Republic Plaza 200 to 600 MHz VNA excess path loss measurements for
positions 13 through 18.

75
 Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 1 Position 2
0 Noise Pos 1 0 Noise Pos 1
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 3 Position 4
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 5 Position 6
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Figure 52. Republic Plaza 700 MHz to 18 GHz VNA excess path loss measurements for
positions 1 through 6.

76
 Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 7 Position 8
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 9 Position 10
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 11 Position 12
0 Noise Pos 6 0 Noise Pos 6
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Figure 53. Republic Plaza 700 MHz to 18 GHz VNA excess path loss measurements for
positions 7 through 12.

77
 Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 13 Position 14
0 Noise Pos 15 0 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 15 Position 16
0 Noise Pos 15 0 Noise Pos 15
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Republic Plaza, Directional TX and Omnidirectional RX Antennas Republic Plaza, Directional TX and Omnidirectional RX Antennas
20 20
Position 17 Position 18
0 Noise Pos 15 0 Noise Pos 19
Excess Path Loss (dB)

Excess Path Loss (dB)

-20 -20

-40 -40

-60 -60

-80 -80

-100 -100
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
Frequency (GHz) Frequency (GHz)

Figure 54. Republic Plaza 700 MHz to 18 GHz VNA excess path loss measurements for
positions 13 through 18.

78
 500

450

400

350
RMS delay spread [ns]

300
200-600 MHz
700-800 MHz
250
200-1800 MHz
865-965 MHz
200

150

100

50

0
0 5 10 15 20
Position number

Figure 55. Republic Plaza RMS delay spread versus position based on four different frequency
bands.

79
Table 4. Legend for locations in the Horizon West walk-through plots shown in Figures 56, 57,
59 and 60 below.

Index Description Index Description

1 Building entrance 33 7th floor location 12
2 1st floor south 34 7th floor location 11
3 2nd floor south 35 7th floor location 10
4 2nd floor location 3 36 7th floor location 9
5 2nd floor location 2 37 7th floor location 8
6 2nd floor location 1 38 7th floor room 707
7 2nd floor location 0 39 7th floor location 7
8 2nd floor location 4 40 7th floor location 6
9 2nd floor location 5 41 7th floor location 5
10 2nd floor location 6 42 7th floor location 0
11 2nd floor location 7 43 7th floor location 4
12 2nd floor location 8 44 7th floor location 0
13 2nd floor location 9 45 7th floor location 1
14 2nd floor location 10 46 7th floor location 2
15 2nd floor location 11 47 7th floor location 3
16 2nd floor location 12 48 7th floor south
17 2nd floor location 13 49 8th floor south
18 2nd floor north 50 8th floor elevator
19 3rd floor north 51 8th floor north
20 3rd elevator 52 9th floor north
21 3rd floor south 53 9th floor elevator
22 4th floor south 54 10th floor south
23 4th floor elevator 55 10th floor elevator
24 4th floor north 56 10th floor north
25 5th floor north 57 11th floor north
26 5th floor elevator 58 11th floor elevator
27 5th floor south 59 11th floor south
28 6th floor south 60 11th floor wait for elevator
29 6th floor elevator 61 11th floor elevator in
30 6th floor north 62 1st floor elevator out
31 7th floor north 63 outside building south
32 7th floor location 13 64 Reference

80
 20

1
2
3
10
19
20
21
22
23
24
25
26
27
28
29
30
36
39
48
49
50
52
53
54
55
57
58
59
61
62
63

-20

-40

81
Power (dB)
-60

Freq. = 749.9 MHz

Ref. = -34.5 dBm
-80
Approx. noise fl. = -87 dB

All numbered locations are indicated by dashed vertical lines, where the order is sequential. (Some numbers are omitted to keep the figure readable.)

spectrum analyzer as the 750 MHz transmitter is carried through the building.
-100
1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500
Sample Number

Figure 56. Horizon West, receive site 1. Normalized received signal power from the
 20

1
2
3
10
19
20
21
22
23
24
25
26
27
28
29
30
37
39
48
49
50
52
53
54
55
57
58
59
61
62
63

10

-10

-20

-30

82
Power (dB)
-40

-50

-60

Freq. = 749.9 MHz

-70 Ref. = -49.0 dBm

Approx. noise fl. = -74 dB

-80

spectrum analyzer as the 750 MHz transmitter is carried through the building.
All numbered locations are indicated by dashed vertical lines, where the order is sequential. (Some numbers are omitted to keep the figure readable.)

500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Figure 57. Horizon West, receive site 2. Normalized received signal power from the
Sample Number
 300 1

240 Histogram median = -29.5 dB 0.8

number of values [N]

Empirical CDF mean = -17.7 dB

empircal F(x)
180 std. dev. = 11.0 dB 0.6
freq. = 749.9 MHz
120 ref. = -34.5 dBm 0.4

60 0.2

0 0
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20
(a)

300 1

240 Histogram median = -25.1 dB 0.8

number of values [N]

Empirical CDF
mean = -18.2 dB

empircal F(x)
180 std. dev. = 9.9 dB 0.6
freq. = 749.9 MHz
120 ref. = -49.0 dBm 0.4

60 0.2

0 0
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20
Power [dB]
(b)

Figure 58. Histogram and empirical CDF of the received spectrum analyzer signal
power at Horizon West for (a) receive site one and (b) receive site two.

83
 63 64

62
61

1400
60

Approx. noise fl. = -99 dB

1200
1000
50

40

Ref. = -31.1 dBm

800
Time (s)
30

600
Freq. = 749.9 MHz

400
20
200

10

3
2
1
0
20

-20

-40

-60

-80

-100

Power (dB)

Figure 59. Horizon West, receive site 1. Normalized narrowband receiver signal power as the
750 MHz transmitter is carried through the building.

84
 64
63

1400
62

61
60

1200
Approx. noise fl. = -85 dB

1000
50

40

800
Ref. = -45.2 dBm

Time (s)
30

600
Freq. = 749.9 MHz

20 400
200

10

3
2
1
0
20

10

-10

-20

-30

-40

-50

-60

-70

-80

-90

Power (dB)

Figure 60. Horizon West, receive site 2. Normalized narrowband receiver signal power as the
750 MHz transmitter is carried through the building.

85
 3200 1

2560 Histogram median = -33.2 dB 0.8

num ber of values [N ]

Empirical CDF mean = -20.6 dB

em pircal F (x)
1920 std. dev. = 10.8 dB 0.6
freq. = 749.9 MHz
1280 ref. = -31.1 dBm 0.4

640 0.2

0 0
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20
(a)

3200 1

Histogram
2560 median = -27.9 dB 0.8
num ber of values [N ]

Empirical CDF
mean = -19.3 dB

em pircal F (x)
1920 std. dev. = 9.0 dB 0.6
freq. = 749.9 MHz
1280 ref. = -45.2 dBm 0.4

640 0.2

0 0
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20
Power [dB]
(b)

Figure 61. Histogram and empirical CDF of the narrowband receiver signal power at Horizon
West for (a) receive site one and (b) receive site two.

86
 0 0
Position 0 (200 MHz BW) Position 1 (200 MHz BW)
second floor second floor
seventh floor seventh floor
-10 noise floor -10 noise floor

-20 -20
Penetration (dB)

Penetration (dB)
-30 -30

-40 -40

-50 -50

-60 -60
2000 4000 6000 8000 10000 12000 14000 16000 18000 2000 4000 6000 8000 10000 12000 14000 16000 18000
Frequency (MHz) Frequency (MHz)
0 0
Position 2 (200 MHz BW)
second floor
seventh floor
-10 noise floor -10

-20 -20
Penetration (dB)

Penetration (dB)

Position 3 (200 MHz BW)

-30 -30 second floor
seventh floor
noise floor

-40 -40

-50 -50

-60 -60
2000 4000 6000 8000 10000 12000 14000 16000 18000 2000 4000 6000 8000 10000 12000 14000 16000 18000
Frequency (MHz) Frequency (MHz)

Figure 62. Excess path loss data for positions 0 through 3 at the Horizon West apartment building.
Black: floor two. Red: floor seven. Blue: Noise floor.

87
 0
0
Position 4 (200 MHz BW)
second floor
-10 seventh floor
Position 5 (200 MHz BW)
second floor
noise floor -10 seventh floor
noise floor

-20
Penetration (dB)

-20

Penetration (dB)
-30
-30

-40
-40

-50
-50

-60
2000 4000 6000 8000 10000 12000 14000 16000 18000 -60
Frequency (MHz) 2000 4000 6000 8000 10000 12000 14000 16000 18000
Frequency (MHz)
0
0
Position 6 (200 MHz BW)
second floor Position 7 (200 MHz BW)
seventh floor second floor
-10 noise floor seventh floor
-10 noise floor

-20
-20
Penetration (dB)

Penetration (dB)
-30
-30

-40
-40

-50
-50

-60
-60
2000 4000 6000 8000 10000 12000 14000 16000 18000
2000 4000 6000 8000 10000 12000 14000 16000 18000
Frequency (MHz)
Frequency (MHz)

Figure 63.Excess path loss data for positions 4 through 7 at the Horizon West apartment building.
Black: floor two. Red: floor seven. Blue: Noise floor.

88
 0 0
Position 8 (200 MHz BW) Position 9 (200 MHz BW)
second floor second floor
seventh floor seventh floor
-10 noise floor -10 noise floor

-20 -20
Penetration (dB)

Penetration (dB)
-30 -30

-40 -40

-50 -50

-60 -60
2000 4000 6000 8000 10000 12000 14000 16000 18000 2000 4000 6000 8000 10000 12000 14000 16000 18000
Frequency (MHz) Frequency (MHz)

0 0
Position 10 (200 MHz BW) Position 11 (200 MHz BW)
second floor second floor
seventh floor seventh floor
-10 noise floor -10 noise floor

-20 -20
Penetration (dB)

Penetration (dB)
-30 -30

-40 -40

-50 -50

-60 -60
2000 4000 6000 8000 10000 12000 14000 16000 18000 2000 4000 6000 8000 10000 12000 14000 16000 18000
Frequency (MHz) Frequency (MHz)

Figure 64. Excess path loss data for positions 8 through 11 at the Horizon West apartment building.
Black: floor two. Red: floor seven. Blue: Noise floor.

89
 0
0
Position 12 (200 MHz BW)
second floor Position 12 (200 MHz BW)
seventh floor second floor
-10 noise floor seventh floor
-10 noise floor

-20
-20
Penetration (dB)

Penetration (dB)
-30
-30

-40
-40

-50
-50

-60
-60
2000 4000 6000 8000 10000 12000 14000 16000 18000
2000 4000 6000 8000 10000 12000 14000 16000 18000
Frequency (MHz)
Frequency (MHz)

Figure 65. Excess path loss data for positions 12 and 13 at the Horizon West apartment building.
Black: floor two. Red: floor seven. Blue: Noise floor.

90
 50
Floor 2
RMS Delay Spread (ns) 45 Floor 7
40

35

30

25

20

15

10
0 5 10 15
Position

Figure 66. Summary of RMS delay spread values calculated at the Horizon West apartment building. Data were
calculated using the frequency band from 1 GHz to 18 GHz.

91
 5500
Ref. Location
A

Approx. noise fl. = -104 dB

5000
B

4500
C

4000
D

Ref. = -11.5 dBm

Sample Number
3500
C

3000
B

Freq. = 749.9 MHz

2500
A

Ref. Location
2000
1500
20

-20

-40

-60

-80

-100

-120

Power (dB)

Figure 67. NIST Boulder laboratory, receive site 1. Normalized received signal power from the
spectrum analyzer as the 750 MHz transmitter is carried through the building. Path I, walk 1.

92
 Ref. Location

4500
A

Approx. noise fl. = -63 dB

B

4000
C

3500
D

Sample Number
Ref. = -40.7 dBm

3000
C

2500
B
Freq. = 749.9 MHz

2000

A
1500

Ref. Location
20

10

-10

-20

-30

-40

-50

-60

-70

Power (dB)

Figure 68. NIST Boulder laboratory, receive site 2. Normalized received signal power
from the spectrum analyzer as the 750 MHz transmitter is carried through the building.
Path I, walk 1.

93
 500 1
Histogram
400 Empirical CDF median = -98.5 dB 0.8
number of values [N]

mean = -19.8 dB

empircal F(x)
300 std. dev. = 28.8 dB 0.6
freq. = 749.9 MHz
200 ref. = -11.5 dBm 0.4

100 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

500 1

400 median = -61.5 dB 0.8

number of values [N]

Histogram mean = -39.6 dB

empircal F(x)
300 Empirical CDF std. dev. = 9.5 dB 0.6
freq. = 749.9 MHz
200 ref. = -40.7 dBm 0.4

100 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)

Figure 69. Histogram and empirical CDF of the spectrum analyzer signal power at
NIST Boulder laboratory for (a) receive site one and (b) receive site two. Path I,
walk 1.

94
 Ref. Location
A

3500
Approx. noise fl. = -104 dB
B

3000
C

2500
Freq. = 749.9 MHz

Sample Number
D

2000
C

1500
B
Ref. = -11.6 dBm

1000

Ref. Location
500
20

-20

-40

-60

-80

-100

Power (dB)

Figure 70. NIST Boulder laboratory, receive site 1. Normalized received signal power from
the spectrum analyzer as the 750 MHz transmitter is carried through the building. Path I, walk
2.

95
 Ref. Location

2000
Approx. noise fl. = -66 dB
B

1500
Ref. = -41.9 dBm

Sample Number
D

1000
C

Freq. = 749.9 MHz

B

500

Ref. Location
0
20

10

-10

-20

-30

-40

-50

-60

-70

Power (dB)

Figure 71. NIST Boulder laboratory, receive site 2. Normalized received signal power from the
spectrum analyzer as the 750 MHz transmitter is carried through the building. Path I, walk 2.

96
 400 1
Histogram
320 median = -86.7 dB 0.8
number of values [N]

Empirical CDF
mean = -16.7 dB

empircal F(x)
240 std. dev. = 30.1 dB 0.6
freq. = 749.9 MHz
160 ref. = -11.6 dBm 0.4

80 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

400 1

320 median = -63.2 dB 0.8

number of values [N]

Histogram
Empirical CDF mean = -38.7 dB

empircal F(x)
240 std. dev. = 11.3 dB 0.6
freq. = 749.9 MHz
160 ref. = -41.9 dBm 0.4

80 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)

Figure 72. Histogram and empirical CDF of the spectrum analyzer signal power at NIST Boulder
laboratory for (a) receive site one and (b) receive site two. Path I, walk 2.

97
 Ref. Location

2500
A

Approx. noise fl. = -105 dB

2000
B

1500
C2

Sample Number
Ref. = -11.6 dBm
B

1000
Freq. = 749.9 MHz

500

Ref. Location
20

-20

-40

-60

-80

-100

Power (dB)

Figure 73. NIST Boulder laboratory, receive site 1. Normalized received signal power from
the spectrum analyzer as the 750 MHz transmitter is carried through the building. Path II,
walk 1.

98
 2000
Ref. Location

Approx. noise fl. = -66 dB

1800
1600
B

1400
Sample Number
Ref. = -41.6 dBm
C2

10001200
B

Freq. = 749.9 MHz

800
600

Ref. Location
400
20

10

-10

-20

-30

-40

-50

-60

-70

Power (dB)

Figure 74. NIST Boulder laboratory, receive site 2. Normalized received signal power from the
spectrum analyzer as the 750 MHz transmitter is carried through the building. Path II, walk 1.

99
 100 1
Histogram
80 median = -62.8 dB 0.8
number of values [N]

Empirical CDF
mean = -18.4 dB

empircal F(x)
60 std. dev. = 25.2 dB 0.6
freq. = 749.9 MHz
40 ref. = -11.6 dBm 0.4

20 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

200 1
Histogram
160 median = -61.1 dB 0.8
number of values [N]

Empirical CDF
mean = -39.8 dB

empircal F(x)
120 std. dev. = 10.9 dB 0.6
freq. = 749.9 MHz
80 ref. = -41.6 dBm 0.4

40 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)

Figure 75. Histogram and empirical CDF of the spectrum analyzer signal power at NIST Boulder
laboratory for (a) receive site one and (b) receive site two. Path II, walk 1.

100
 3000
Ref. Location

Approx. noise fl. = -101 dB

2500
B

Ref. = -14.7 dBm

2000
Sample Number
C2

1500
B Freq. = 749.9 MHz

A
1000

Ref. Location
20

-20

-40

-60

-80

-100

Power (dB)

Figure 76. NIST Boulder laboratory, receive site 1. Normalized received signal power from the
spectrum analyzer as the 750 MHz transmitter is carried through the building. Path II, walk 2.

101
 Ref. Location

2000
A

Approx. noise fl. = -66 dB

1800
B

1600 1400
Sample Number
Ref. = -41.8 dBm
C2

1200
Freq. = 749.9 MHz

1000
B
800

Ref. Location
600
20

10

-10

-20

-30

-40

-50

-60

-70

Power (dB)

Figure 77. NIST Boulder laboratory, receive site 2. Normalized received signal power from the
spectrum analyzer as the 750 MHz transmitter is carried through the building. Path II, walk 2.

102
 50 1
Histogram freq. = 749.9 MHz
ref. = -14.7 dBm median = -64.6 dB
40 Empirical CDF mean = -16.0 dB 0.8
number of values [N]

std. dev. = 26.9 dB

empircal F(x)
30 0.6

20 0.4

10 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

200 1
Histogram
160 Empirical CDF median = -61.9 dB 0.8
number of values [N]

mean = -37.3 dB

empircal F(x)
120 std. dev. = 12.4 dB 0.6
freq. = 749.9 MHz
80 ref. = -41.8 dBm 0.4

40 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)

Figure 78. Histogram and empirical CDF of the spectrum analyzer signal power at NIST Boulder
laboratory for (a) receive site one and (b) receive site two. Path II, walk 2.

103
 1300
Ref. Location

Approx. noise fl. = -111 dB

1200
B

1100
C

1000
D

Ref. = -19.4 dBm

Time (s)
900
800
C

B Freq. = 749.9 MHz

700

A
600

Ref. Location
500
20

-20

-40

-60

-80

-100

-120

Power (dB)

Figure 79. NIST Boulder laboratory, receive site 1. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path I, walk 1.

104
 Ref. Location
A

1400
Approx. noise fl. = -96 dB
B

1300
C

1200
D

1100
Time (s)
Ref. = -34.8 dBm

1000
C

900
B
Freq. = 749.9 MHz

800

Ref. Location
700
20

-20

-40

-60

-80

-100

Power (dB)

Figure 80. NIST Boulder laboratory, receive site 2. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path I, walk 1.

105
 2000 1

Histogram
1600 median = -93.0 dB 0.8
number of values [N]

Empirical CDF
mean = -17.4 dB

empircal F(x)
1200 std. dev. = 29.1 dB 0.6
freq. = 749.9 MHz
800 ref. = -19.4 dBm 0.4

400 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

2000 1
Histogram
1600 Empirical CDF median = -71.5 dB 0.8
number of values [N]

mean = -39.9 dB

empircal F(x)
1200 std. dev. = 16.5 dB 0.6
freq. = 749.9 MHz
800 ref. = -34.8 dBm 0.4

400 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)
Figure 81. Histogram and empirical CDF of the narrowband receiver signal power at NIST Boulder
laboratory for (a) receive site one and (b) receive site two. Path I, walk 1.

106
 Ref. Location

700
Approx. noise fl. = -111 dB
B

600
C

500
Time (s)
D

Ref. = -19.4 dBm

400
C

300
B Freq. = 749.9 MHz

200

A
100

Ref. Location
20

-20

-40

-60

-80

-100

-120

Power (dB)

Figure 82. NIST Boulder laboratory, receive site 1. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path I, walk 2.

107
 Ref. Location

700
Approx. noise fl. = -97 dB
B

600
C

500
Time (s)
D

Ref. = -33.3 dBm

400
C

300
B Freq. = 749.9 MHz

200

Ref. Location
100
20

-20

-40

-60

-80

-100

Power (dB)

Figure 83. NIST Boulder laboratory, receive site 2. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path I, walk 2.

108
 2000 1

Histogram
1600 median = -85.6 dB 0.8
number of values [N]

Empirical CDF
mean = -24.5 dB

empircal F(x)
1200 std. dev. = 28.1 dB 0.6
freq. = 749.9 MHz
800 ref. = -19.4 dBm 0.4

400 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

2000 1
Histogram
1600 Empirical CDF median = -70.7 dB 0.8
number of values [N]

mean = -42.4 dB

empircal F(x)
1200 std. dev. = 16.3 dB 0.6
freq. = 749.9 MHz
800 ref. = -33.3 dBm 0.4

400 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)
Figure 84. Histogram and empirical CDF of the narrowband receiver signal power at NIST Boulder
laboratory for (a) receive site one and (b) receive site two. Path I, walk 2.

109
 700
Ref. Location

Approx. noise fl. = -111 dB

650
A

600
B

550
500
Ref. = -19.8 dBm
C2

450
Time (s)
400
B

350
Freq. = 749.9 MHz

300
250

Ref. Location
200
20

-20

-40

-60

-80

-100

-120

Power (dB)

Figure 85. NIST Boulder laboratory, receive site 1. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path II, walk 1.

110
 700
Ref. Location

Approx. noise fl. = -98 dB

A

650
600
B

550
500
Ref. = -33.0 dBm

Time (s)
C2

450
400
B

350
Freq. = 749.9 MHz

300
250

A
200

Ref. Location
20

-20

-40

-60

-80

-100

Power (dB)

Figure 86. NIST Boulder laboratory, receive site 2. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path II, walk 1.

111
 2000 1
Histogram
1600 Empirical CDF median = -64.0 dB 0.8
number of values [N]

mean = -21.8 dB

empircal F(x)
1200 std. dev. = 23.9 dB 0.6
freq. = 749.9 MHz
800 ref. = -19.8 dBm 0.4

400 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

2000 1
Histogram
1600 Empirical CDF median = -62.2 dB 0.8
number of values [N]

mean = -39.1 dB

empircal F(x)
1200 std. dev. = 15.2 dB 0.6
freq. = 749.9 MHz
800 ref. = -33.0 dBm 0.4

400 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)

Figure 87. Histogram and empirical CDF of the narrowband receiver signal power at NIST Boulder
laboratory for (a) receive site one and (b) receive site two. Path II, walk 1.

112
 Ref. Location

500
Approx. noise fl. = -111 dB
A

450
400
B

350 300
Ref. = -19.4 dBm

Time (s)
C2

250
200
B
Freq. = 749.9 MHz

150
100

Ref. Location
50
20

-20

-40

-60

-80

-100

-120

Power (dB)

Figure 88. NIST Boulder laboratory, receive site 1. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path II, walk 2.

113
 650
Ref. Location

600
A

Approx. noise fl. = -98 dB

550
500
B

450
Ref. = -33.0 dBm

Time (s)
400
C2

350
300
B
Freq. = 749.9 MHz

250

A
200

Ref. Location
150
20

-20

-40

-60

-80

-100

Power (dB)

Figure 89. NIST Boulder laboratory, receive site 2. Normalized received signal power from the
narrowband receiver as the 750 MHz transmitter is carried through the building. Path II, walk 2.

114
 750 1
freq. = 749.9 MHz
600 ref. = -19.4 dBm median = -65.9 dB 0.8
number of values [N]

mean = -15.9 dB

empircal F(x)
450 std. dev. = 26.9 dB 0.6
Histogram
300 Empirical CDF 0.4

150 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
(a)

1500 1
Histogram
1200 median = -65.6 dB 0.8
number of values [N]

Empirical CDF
mean = -37.8 dB

empircal F(x)
900 std. dev. = 16.7 dB 0.6
freq. = 749.9 MHz
600 ref. = -33.0 dBm 0.4

300 0.2

0 0
-120 -100 -80 -60 -40 -20 0 20
Power [dB]
(b)

Figure 90. Histogram and empirical CDF of the narrowband receiver signal power at NIST
Boulder laboratory for (a) receive site one and (b) receive site two. Path II, walk 2.

115
Figure 91. Excess path loss data from 25 MHz to 1.3 GHz for a path that includes building
penetration obtained with data for the Wing 4 hallway. Omnidirectional antennas were
used. Distance down the corridor is D = 2.40 m (top) and D = 17.65 m (bottom).

116
Figure 92. Excess path loss data from 25 MHz to 1.3 GHz for a path that includes building
penetration obtained with data for the Wing 4 hallway. Omnidirectional antennas were used.
Distance down the corridor is D = 32.90 m (top) and D = 48.15 m (bottom).

117
Figure 93. Excess path loss data from 25 MHz to 1.3 GHz for a path that includes building
penetration obtained with data for the Wing 4 hallway. Omnidirectional antennas were
used. Distance down the corridor is D = 63.40 m (top) and D = 78.65 m (bottom).

118
Figure 94. Excess path loss data from 25 MHz to 1.3 GHz for a path that includes building
penetration obtained with data for the Wing 4 hallway. Omnidirectional antennas were used.
Distance down the corridor is D = 93.90 m (top) and D = 109.15 m (bottom).

119
Figure 95. Excess path loss data from 25 MHz to 1.3 GHz for a path that includes
building penetration obtained with data for the Wing 4 hallway. Omnidirectional antennas
were used. Distance down the corridor is D = 124.4 m (top) and D = 139.65 m (bottom).
120
Figure 96. Excess path loss data from 25 MHz to 1.3 GHz for a path that includes building
penetration obtained with data for the Wing 4 hallway. Omnidirectional antennas were used.
Distance down the corridor is D = 154.9 m (top) and D = 170.15 m (bottom).

121
Figure 97. Excess path loss data from 25 MHz to 1.3 GHz for a path that includes building
penetration obtained with data for the Wing 4 hallway. Omnidirectional antennas were used.
Distance down the corridor is D = 185.40 m (top) and D = 200.65 m (bottom).
122
 90

80

70
RMS delay spread (ns)

60

50
25 -1300 MHz
750 - 18000 MHz
40

30

20

10

0
1 2 3 4 5 6 7 8 9 10
Position

Figure 98. NIST Boulder Laboratory RMS delay spread versus position based on two
different frequency bands.

123
 Experiment Index
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
35.0
Path 1 Path 2
Colorado
30.0
Convention Horizon
Center West
25.0

20.0
dB

15.0

10.0
Republic
Plaza NIST Laboratory Building
5.0

0.0

Spectrum Analyzer Narrowband Receiver

Figure 99. Comparison of the standard deviation for the normalized spectrum analyzer and
narrowband receiver data across all the test locations.

Experiment Index
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
0.0
NIST Laboratory Building

‐20.0

‐40.0
dB

‐60.0

Republic
‐80.0 Plaza Horizon
Colorado
Convention West Path 1 Path 2
Center
‐100.0
Median of Spectrum Analyzer Mean of Spectrum Analyzer
Median of Narrowband Receiver Mean of Narrowband Receiver

Figure 100. Comparison of median and mean values from the normalized spectrum analyzer and
narrowband receiver data across all the test locations.

124
 Experiment Index
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
‐60.0

Colorado NIST Laboratory Building

‐70.0 Convention
Center Republic
Plaza
‐80.0
dB

‐90.0

‐100.0
Horizon
West
‐110.0
Path 1 Path 2
‐120.0

Spectrum Analyzer Narrowband Receiver

Figure 101. Comparison with the reference value added to the median values from the
normalized spectrum analyzer and narrowband receiver data across all the test locations.

500.0

450.0
Colorado
Horizon
400.0 Convention
West
Center
350.0

300.0 Republic
Plaza
ns

250.0
NIST
200.0

150.0

100.0

50.0

0.0
1 2 3 4 5 6 7 8 9 10 11 12
Mea n Medi a n Mi ni mum Ma xi mum

Figure 102. RMS delay spread statistics for the four buildings, indexed on the horizontal axis by
the various frequency bands provided in Table 3.

125