Estudio de La Primera
Estudio de La Primera
fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
1
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
2
IFFT 1 DAC 1
... B1
...
...
Transmitter
...
...
...
...
...
...
...
... P/SMt
...
RFMt AAMt
An array of two or more antenna elements can yield a ...
IFFT Mt DAC Mt
... BMt
directional radiation pattern. The width and the direction of the
...
...
...
...
...
...
...
...
fc
radiation pattern, or the beam, depend upon the geometry of the
RECEIVER RF1 AA1
array pattern, the amplitude and phase of the signal transmitted ...
S/P1
...
by or received from each element of the array. The far-field FFT1 ADC 1
...
...
B1
...
...
...
...
...
Receiver
...
radiation pattern of an array antenna at an angle for a beam
fc
MIMO
...
...
... S/PMr
...
RFMr AAMr
pointing towards k can be represented as [35] ...
...
N A 1 FFTMr ADC Mr
...
...
BMr
...
...
...
...
...
n0
Fig. 2. Transmitter and receiver with beamforming antennas.
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
3
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
4
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
5
HO Interruption Time
8. Data Transfer
Detach from serving cell and
the above frame structure of Fig. 4, the DL beam measurement synchronize to new cell
HO Delay
period is 10 ms and 20 ms when the number of BS beams is 8 9. RACH procedure
and 16, respectively. When even more beams are used at the BS, 10. UL allocation + UL Timing Adjustment
11. Handover Complete
the beam measurement period will increase, or more slots must
Packet data Packet data
be allocated as the BM Slot.
The random access channel (RACH) Slot is dedicated to the Fig. 5. BS initiated HO procedure.
UL RACH operation and the UL synchronization. In the RACH This means that hundreds of multiple access users can be
Slot, the MS transmits the RACH signal with a fixed beam, and supported in each subframe. Various multi antenna techniques
the BS receives the signal with the beam switched to more than of the 3GPP LTE [31] can be supported in the frame structure
one direction. The MS may choose candidate transmit beam(s) when multiple antenna arrays are used [22]-[25]. In particular,
for the RACH Slot from the DL beam measurement. An IFFT digital precoding in addition to analog beamforming is possible.
signal block without the CP is repeatedly transmitted during the Reference signal structure corresponding to each antenna array
time period allocated as the RACH Slot, and the IFFT size of or signal layer for the multi antenna techniques can be defined
8192, 6.66 s was used in the testbed to support the RACH for the Data Slot as in the 3GPP LTE [31].
coverage of up to 1 km. The BS has to switch the receive beam
to receive the RACH signal from more than one direction. This IV. HANDOVER ALGORITHM
is necessary because the RACH Slot can be transmitted from In order to better support user mobility in multi-cell
any location within the cell area, but the best BS beams to environment, we considered two types of handover (HO)
receive the signal are usually unknown to the BS at that time. procedures: the BS initiated HO and the MS initiated HO.
The received signal power can change quite dramatically as the
receive beam is switched. As a result, automatic gain control A. BS Initiated HO
(AGC) was implemented at each beam switching in the testbed. The BS initiated HO procedure is based on and modified
The Control Slot is dedicated to short bursts of information from the 3GPP Intra-LTE HO procedure [33] and is shown in
such as the physical control channel in the DL and the UL. In Fig. 5. In the BS initiated HO, the MS sends measurement
the Control Slot, the BS can switch its beam at the start of each reports to the source BS, which is the serving cell before HO,
OFDM symbol to the direction of the MS to transmit or receive and the source BS makes HO decision based on the reports
the signal. Each OFDM symbol has reference signals at from the MS and sends HO request message to the target BS via
predefined subcarrier positions, and channel estimation at the the gateway. The target BS performs admission control and
receiver is done at each symbol. Transmission of broadcast sends HO response message to the source BS via the gateway.
information can be supported by repeating transmission of the If the HO response is affirmative, the source BS issues HO
same broadcast information with different beams. command to the MS which then performs random access
The Data Slot is dedicated to a large burst of data such as the procedure with the target BS and establishes UL
user data in the DL and the UL. In the Data Slot, the BS beam synchronization. When the UL synchronization is successfully
can be switched only at the start of the slot and is fixed within established, the MS sends HO complete message to the target
the slot, so the radio channel is static in the Data Slot, and the BS. The target BS sends MS context release message to the
resource block (RB) and reference signals similar to those source BS via the gateway, confirming a successful handover
defined in the 3GPP LTE [31] can be designed. By scheduling and enabling the source BS resources to be released.
the Data Slots to MSs, time division multiplexing/multiple
B. MS Initiated HO
access (TDM/TDMA) is supported, which means at least 10
users can be supported by TDM/TDMA in a subframe. Also, by In the MS initiated HO, the source BS configures the HO
scheduling RBs of a Data Slot, frequency division multiplexing event conditions, but the MS makes HO decision and selects a
target BS, as illustrated in Fig. 6. Some of the previous works
/multiple access (FDM/FDMA) can be supported for the MSs
on the MS initiated HO procedure can be found in [40]-[41].
for which the best BS beams are the same. For example, when
When the HO decision is made, the MS goes to the
the signal bandwidth is 800 MHz with 2640 subcarriers, if the non-connected state with the source BS, and starts the HO
RB size is 60 subcarriers in the frequency domain, then up to 44 RACH procedure with the target BS. A group of RACH
(=2640/60) FDM/FDMA users can be supported in a Data Slot.
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
6
MS
UE Source BS Target BS Gateway
1. Measurement Configuration
Packet data Packet data
Measurement
2. HO Decision (HO_d)
Detach from serving cell
and synchronize to new cell
HO RACH (MSG1)
HO Interruption Time
MSG2
MSG3
4. MS-Reg-Request
5. MS-Reg-Response
MSG4
6. MS-Connection-Release
7. RLC-State-Notify-Request
8. NotifyGW-Request
9. NotifyGW-Response
10. RLC-State-Notify-ACK (HO Complete)
Packet data Packet data
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
7
MS
assigned with unique reference signals different from those of 0
5 10 15 20 25 30
other antenna arrays. This is equivalent to the case in which one
BS with 4 antenna arrays transmits 4 data streams to two MSs. 1
Antenna
BestMS
1000
[Mbps]
of Fig. 10. In the test of Fig. 10, each BS-MS path achieved the 500
0
5 10 15 20 25 30
data rate of 3.77 Gbps by 2x2 MIMO, supporting two data Time[s]
streams with 64 QAM and 3/4 code rate with the block error
Fig.11. Adaptive beamforming test at 110 km/h in a single cell.
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
8
until the end of the test. The test was logged at every 100 ms by
the MS. The beam measurement period was 10 ms, and the MS
determined the best DL beam pair and fed back the best beam
information to the BS. Also, the MS reported the channel
quality information (CQI) to the BS at a regular interval. Based
on the feedback information, the BS updated the transmit beam
and the modulation and coding scheme (MCS).
Fig. 11 shows plots of the beam IDs of the BS and the MS,
the best antenna of the MS and the throughput during the test.
The best BS beam ID changed from 1 to 7 one by one as the
vehicle moved. The best beam and antenna of the MS also
changed similarly during the test, except at about 12 s. This
sudden change at 12 s was due to a guard post that blocked the
LOS path between the BS and the MS. The testbed achieved the
DL data rate of over 1.2 Gbps with 16 QAM and code rate of
3/4 during the test when the LOS path was available, but the Fig.12. Handover field test scenario.
data rate fell to about 500 Mbps with QPSK and code rate of 3/4 after the HO decision (HO_d). The average HO delay was
when the LOS path was blocked. In this test, MIMO was not about 12 ms, and the average HO interruption time was about
supported, and only the best antenna was selected at the BS and 21 ms, satisfying the HO interruption time requirement of 27.5
the MS to deliver single data stream. In the testbed, the DL ms of the IMT-Advanced [28]. UHD video ran smoothly
transmit power was fixed, whereas the UL transmit power was without interruption at 98% of the HO locations.
updated based on the DL received power without the control of The MS kept track of the received power of the serving cell
the BS. and that of the best target cell for the MS initiated HO
procedure. HO decision was made at the MS when the average
C. Handover Test in Three-Cell Environment at 20 km/h target cell received power was consistently higher than the
Extensive multi-cell handover tests for 5G testbed were average serving cell power by a predefined threshold. Fig. 13
conducted at Samsung Electronics, Suwon, Korea facility. shows plots of the received powers of the serving cell and the
Example HO scenario used for field test is shown in Fig. 12. best target cell, the serving cell ID, the best beam ID of the
Three BSs were installed at different locations having cell IDs 0, serving cell, the best beam ID of the best MS antenna and the
2 and 4, respectively. Each BS generated 16 beams in total; 8 timing advance (TA) values during 10 seconds of MS travel
horizontal beams and 2 vertical beams. The coverage for each from location 1a to 1b (marked in Fig. 12). Similarly, Fig. 14
cell is illustrated in the Fig. 12. These three BSs were connected
to a common gateway with S1 interface. An MS was installed
in a minivan and the MS RF and antenna unit were mounted on
top of the vehicle. The MS had two antenna arrays that
provided the coverage of about 360o in horizontal direction;
each array generating 8 beams with the coverage of 180o. The
path taken by the vehicle for these tests is illustrated in the
figure by colored directed arrows. The colors of the arrows are
chosen to match those of the serving cells in the test route.
Throughout the tests, the vehicle moved at about 20 km/h.
Initially, HO tests were performed using the BS initiated HO
procedure of Fig. 5. Handover failure rates using this procedure
were very high, especially at location 2 (star-marked in Fig. 12).
This high HO failure rate was mainly attributed to the UL radio
link failure with the serving BS before completing the HO
procedure. Even for successful HO cases, around 30% of trials
incurred more than 500 ms of HO delay. This was mainly due
to the large delay incurred in steps 7 and 8 (Fig. 5) caused by
high interference from target cell and high DL and UL packet
losses. To address the high HO failure rate and large HO delay
issues of the BS initiated HO procedure, the MS initiated HO
procedure of Fig. 6 was implemented and tested.
HO tests were repeated using the MS initiated HO procedure
in the same test environment as before. The HO delay in Fig. 6
is defined as the time taken for the MS to successfully receive
MSG4 from the target BS after the HO decision (HO_d). The
HO interruption time is defined as the time taken for the MS to Fig.13. Received powers of cells, Serving cell ID, Best serving cell beam,
Best MS beam, and Timing advance values at HO location 1.
successfully receive the DL data packets from the target BS
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
9
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
10
[7] M. Elkashlan, T. Q. Duong, H. -H. Chen, “Millimeter-wave [27] Draft Standard for Information Technology–Telecommunications and
communications for 5G: fundamentals: Part I [Guest Editorial],” IEEE Information Exchange Between Systems–Local and Metropolitan Area
Commun. Mag., vol. 52, no. 9, pp. 52–54, 2014. Networks–Specific Requirements–Part 11: Wireless LAN Medium
[8] M. Elkashlan, T. Q. Duong, H. -H. Chen, “Millimeter-wave Access Control (MAC) and Physical Layer (PHY) Specifications–
communications for 5G–Part 2: applications [Guest Editorial],” IEEE Amendment 4: Enhancements for Very High Throughput in the 60 Ghz
Commun. Mag., vol. 53, no. 1, pp. 166–167, 2015. Band, IEEE P802.11ad/D9.0, Oct. 2012.
[9] T. S. Rappaport, S. Sun, R. Mayzus, H. Zhao, Y. Azar, K. Wang, G. N. [28] “Requirements related to technical performance for IMT-Advanced radio
Wong, J. K. Schulz, M. Samimi, and F. Gutierrez, "Millimeter wave interface(s),” Report ITU-R M.2134, 2008-11, [Online]. Available:
mobile communications for 5G cellular: It will work!," IEEE Access, vol. http://www.itu.int/publ/R-REP-M.2134-2008/en
1, pp. 335–349, 2013. [29] “Guidelines for evaluation of radio interface technologies for
[10] L. Wei, R. Hu, Y. Qian, and G. Wu, “Key elements to enable millimeter IMT-Advanced,” Report ITU-R M.2135, 2008-11, [Online]. Available:
wave communications for 5G wireless systems,” IEEE Wireless http://www.itu.int/publ/R-REP-M.2135-2008/en
Commun., Issue 6, pp. 136-143, Dec. 2014. [30] 3rd Generation Partnership Project, 3GPP TS 36.201.
[11] S. Rangan, T. S. Rappaport, and E. Erkip, "Millimeter-Wave Cellular [31] 3rd Generation Partnership Project, 3GPP TS 36.211.
Wireless Networks: Potentials and Challenges," Proceedings of the IEEE, [32] 3rd Generation Partnership Project, 3GPP TS 36.214.
vol. 102, no. 3, Mar. 2014.
[33] 3rd Generation Partnership Project, 3GPP TS 36.300.
[12] Y. Niu, Y. Li , D. Jin, L. Su, and A. V. Vasilakos, "A survey of millimeter
wave communications (mmWave) for 5G: opportunities and challenges," [34] "NGMN 5G WHITE PAPER," Next Generation Mobile Networks
Wireless Networks, vol. 21, Issue 8, pp. 2657-2676, Nov. 2015, DOI (NGMN) Alliance, Feb. 2015. [Online]. Available: https://www.ngmn.org
10.1007/s11276-015-0942-z. /uploads/media/NGMN_5G_White_Paper_V1_0.pdf
[13] A. Ghosh, T. A. Thomas, M. C. Cudak, R. Ratasuk, P. Moorut, F. W. [35] Robert J. Mailloux, "Chapter 1. Phased Arrays in Radar and
Vook, T. S. Rappaport, G. R. MacCartney, S. Sun, and S. Nie, Communication Systems, " in Phased Array Antenna Handbook, 2nd ed.,
Artech House, 2005.
“Millimeter-Wave Enhanced Local Area Systems: A High-Data-Rate
Approach for Future Wireless Networks,” IEEE J. Sel. Areas Commun., [36] T. S. Rappaport, G. R. MacCartney, Jr., M. K. Samimi, S. Sun,
vol. 32, no. 6, pp. 1152–1163, Jun. 2014. “Wideband Millimeter-Wave Propagation Measurements and Channel
Models for Future Wireless Communication System Design,” IEEE
[14] R. Baldemair, T. Irnich, K. Balachandran, E. Dahlman, G. Mildh, Y. Seln,
Trans. Commun., vol. 63, no. 9, pp. 3029-3056, Sep. 2015.
S. Parkvall, M. Meyer, and A. Osseiran, “Ultra-dense networks in
millimeter-wave frequencies,” IEEE Commun. Mag., vol. 53, no. 1, pp. [37] J. Ning, T.-S. Kim, S. V. Krishnamurthy, and C. Cordeiro, “Directional
202–208, Jan. 2015. neighbor discovery in 60 GHz indoor wireless networks,” in Proc. ACM
MSWiM ’09 (Tenerife, Spain), Oct. 2009, pp. 365–373.
[15] H. Mehrpouyan, M. Matthaiou, R. Wang, G. K. Karagiannidis, and Y.
Hua, “Hybrid millimeter-wave systems: a novel paradigm for HetNets,” [38] M. Kim, Y. S. Kim, and W. Lee, “Analysis of directional neighbour
IEEE Commun. Mag., vol. 53, no. 1, pp. 216–221, Jan. 2015. discovery process in millimetre wave wireless personal area networks,”
IET networks, vol. 2, no. 2, pp. 92–101, 2013.
[16] K. Sakaguchi, G. K. Tran, H. Shimodaira, S. Nanba, T. Sakurai, K.
Takinami, I. Siaud, E. C. Strinati, A. Capone, I Karls, R. Arefi, and T. [39] H. Park, Y. Kim, T. Song, S. Pack, “Multi-band directional neighbor
Haustein, "Millimeter-Wave Evolution for 5G Cellular Networks," IEICE discovery in self-Organized mmWave ad-hoc networks,” IEEE Trans.
Trans. Commun., vol. E98-B, no. 3, pp.388-402, Mar. 2015. Veh. Technol., vol. 64, no. 3, pp. 1143-1155, Mar. 2015.
[17] P. Wang, Y. Li, L. Song, and B. Vucetic, "Multi-Gigabit Millimeter Wave [40] M. Niswar, S. Kashihara, Y. Taenaka, K. Tsukamoto, Y. Kadobayashi,
Wireless Communications for 5G: From Fixed Access to Cellular and S. Yamaguchi, “MS-initiated handover decision criteria for VoIP
Networks," IEEE Commun. Mag., pp.168-178, Jan. 2015 over IEEE 802.16e,” in Proc. IEEE Pacific Rim Conf. Commun.,
Computers and Signal Processing (PACRIM'09), 2009, pp. 497-502.
[18] R.J. Weiler, M. Peter, W. Keusgen, E. Calvanese-Strinati, A. De
Domenico, I. Filippini, A. Capone, I. Siaud, A.-M. Ulmer-Moll, A. [41] “LTE Mobility Enhancements,” Qualcomm Inc., Feb., 2010. [Online].
Maltsev, T. Haustein, and K. Sakaguchi, "Enabling 5G backhaul and Available:
access with millimeter-waves," European Conf. Networks and Commun. https://www.qualcomm.com/documents/lte-mobility-enhancements
(EuCNC), Jun. 2014, pp. 1-5. [42] Y. H. Cho and J. J. Kim, "Line-of-Sight MIMO Channel in
[19] C. Dehos, J. L. González, A. De Domenico, D. Kténas, and L. Dussopt, Millimeter-Wave Beamforming System: Modeling and Prototype
"Millimeter-Wave Access and Backhauling: The Solution to the Results," in Proc. 81st Veh. Technol. Conf., May 2015, pp.1-5.
Exponential Data Traffic Increase in 5G Mobile Communications [43] Y. Kim, H. Y. Lee, J. Oh, J. Lee, W. Roh and K. Cheun, "Feasibility of
Systems?," IEEE Commun. Mag., pp. 88-95, Sep. 2014. Mobile Cellular Communications at Millimeter Wave Frequency," in
[20] R. Taori and A. Sridharan,“Point to Multipoint In-Band mm-Wave Proc. IEEE Global Commun. Conf. (GLOBECOM), Dec. 2015.
Backhaul for 5G Networks,” IEEE Commun. Mag., pp. 195 - 201, Jan.
2015.
[21] K. Zheng, L. Zhao, J. Mei, M. Dohler, W. Xiang, and Y. Peng, "10 Gb/s Yungsoo Kim is a principal engineer at the
HetSNets with Millimeter-Wave Communications: Access and Communications Research Team,
Networking Challenges and Protocols," IEEE Commun. Mag., pp.
222-231, Jan. 2015. Samsung Electronics, Suwon, Korea. He
[22] S. Sun, T. S. Rappaport, R. W. Heath, Jr., A. Nix, and S. Rangan, "MIMO received the B.S. degree from Yonsei
for Millimeter-Wave Wireless Communications: Beamforming, Spatial University, Korea in 1987, the M.S. degree
Multiplexing, or Both?," IEEE Commun. Mag., pp.110-121, Dec. 2014. from the University of Wisconsin,
[23] O. El Ayach, S. Rajagopal, S. Abu-Surra, Z. Pi, and R. W. Heath, Jr., Madison in 1989, and the Ph.D. degree
“Spatially Sparse Precoding in Millimeter Wave MIMO Systems,” IEEE
Trans. Wireless Commun., vol.13, no.3, pp. 1499-1513, Mar. 2014.
from the Korea Advanced Institute of
[24] A. Alkhateeb, O. El Ayach, G. Leus, R. W. Heath Jr, “Channel Estimation
Science and Technology (KAIST) in 2000. He joined Samsung
and Hybrid Precoding for Millimeter Wave Cellular Systems,” IEEE J. Advanced Institute of Technology (SAIT), Korea in 1991, and
Sel. Topics in Signal Processing, vol.8, no.5, pp.831-846, Oct. 2014. then transferred to DMC R&D Center at Samsung Electronics,
[25] J. Singh, S. Ramakrishna, “On the Feasibility of Codebook-Based Suwon, Korea in 2006. He contributed to the research and
Beamforming in Millimeter Wave Systems With Multiple Antenna development of 4G physical layer and MAC layer algorithms as
Arrays,” IEEE Trans. Wireless Commun., vol. 14, no. 5, pp. 2670-2683,
May 2015. a project leader at SAIT and Samsung Electronics. He also
[26] IEEE 802.15 WPAN Millimeter Wave Alternative PHY Task Group 3c contributed to the design and implementation of 5G millimeter
(TG3c). [Online]. Available: http://www.ieee802. org/15/pub/TG3c.html. wave communications system. His current research activities
are in 5G mobile communications system.
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSTSP.2016.2520901, IEEE Journal
of Selected Topics in Signal Processing
11
1932-4553 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.