0 ratings 0% found this document useful (0 votes) 7 views 45 pages Hao Ticket
The document discusses various concepts related to probability, random processes, and noise in communication systems. It covers topics such as joint distributions, error correction, and the impact of noise on signal quality. Additionally, it highlights the importance of source coding and the effects of different types of noise on communication systems.
AI-enhanced title and description
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content,
claim it here .
Available Formats
Download as PDF or read online on Scribd
Go to previous items Go to next items
Save Hao ticket For Later mcg
»
“hen the possible ous comes of an e
x
‘ are easly lisely Sot seedy erate aw
- sapiely auch
y| asdcal paobubsitry
gis)
a _aendom “proces with Hime uveruged
eu +o engemble averdges TS called uf
Ergodic
8) | Two independent aerdom proces wih hav
Ahelx CSS conve ton seg of
individu) mewh’
7) sero product
4) | Auto comrelotion of lag zero of amy -parocegs
ts SCee ome 7
| | variance oF Process.
sy | oise eS
Jingle) eave)
[Match hw he sume power spectral density 2
eee VWinize noise |
DtAnen x uml J} ove shanspealiy iad ependen
[hen TT Ug is P
So|
odof* 4 ¥
¥
e's
gy || Wok foo™ coding sechnig' man - j
con struck ny phe Gousc &
cedundaney
— | mM inimuch
code
we presents
oa consi stS
Ser rush
3) Which mong Ae Forrowin
[ame code ||_ sus
a
yey) rhe Cupactty eelukomsbip 18 gj
Pea
gl c= Winga C +S ee,
| N
12) | \Ahidh of 5
almvabatt of ne Follo
| OF speech eas Se AN pe
\A uve Foam collec eel aan
me cod
ors,@
v E
= yWelle a note on!
. : Uigtribution of +o van varie) eg_|
Joint Ne aman] Hon £ wo ondom Na jb) eg.
J \imit theorem
cenrre
v Wd Rundo ry vumchl 4
of +wo Tamdory
ace olbcr tot V4 behavior
4 ny Pegether
it 1 ord
‘ \ue
_ v r L orp\ 4
meas. Q yy x un Hen
oF ee ‘ ne & ens, Onan
wo dimensional spare The joinF parobevbt Hit
On " t
Py j
Czy) doodMHon of 480 render vari! eg,
+heorem »
censral \ieit
Joint pistrbution of TTwo Rundom vuriah! 4
—the ean Uetri bution of +wo Fondo
variwbles descnrbeg ane probate ity behavior
c€ born vusirblep © cura y yegtther .
z For oiscreke Romdom variables
=the joint paobobiity mags Fanon CPIM)
FC ag ee paroberbrl! ty
Or keep tne. Wulace ree ee
| awe values 4d
SO ap ee ecm ap ees
ss)
: for Continuouty Romdom vuriedel ay»
=}
che Joint pavbobi Ripe den Sip Funchion
POF) f Ges) dive Ache Lemsity Over
2 two dimensional _Spaives « othe Joine parcbecht tity
=
a
— Over a _reainn A 1S 5
eee
PCCK Sa cara Tian Cad® derive
e com ~
° faom the joint Yarn ween t = ~
+ [marginal dtsrei buttons ——— 4
IvprosDaliitia paar den seit a
[Nasi wble Gndependemy r
= enlivional ais TeoHne
Usriburion of one varie) &
Defim Hon -
+ | pesertber, the poobuliliry
Lvurtobieg occurring together.
ee aera
Discrete cae!
wl. Jens, pes
fie Worse Niels)
mom ey PCx =x) Yay hed
continuong coe!
© | Joint POF.
fa@re|))
We Fiefeoeh Ceig) i doc, dy = "s
Fron Joint DiS+aib UHR@
conrre] Limit “Theorem CCtT)
- the cemirel Limit theorem 36 a fundumented
’ wesudt In probability sheen] umd shatistiey,
sah ement
Kae 2Xn ore indepundemt end
RUE colly Narviuured C14 crendorn _Vurieubl
worn mean WU ewnd Finite varl\ omte
thenth ne
inal diataribution
ynibuion of Yee
Sample eneeun
distrivaton of the
+e Mcrae j read vse of the
2, AnrNSEHiG uae 4cal Inference
\ ump de 8 al
. Gon Fidemce, Inteyvuls
WAP oxne sis seeingnieriwihel Pepi meee ag
Natiance Oo > —— ae
e sumple mewn we ss
Xen Gupte)
Es eee
stundurdized Form :
z= X—-r — Nod)
Cae ENED
\mpov tence *
Wore) approximation of Sample meunsg
[works mequrdless of origina) diadaibutton
lor lorqeo 7
Applicettons -
con @idente Ynteweale
wyromesis Aes n
forecustiny od quality como)|
. Me
Mises be ps : i ees
[rhe mondo verriahre sna} — eae men) y
Goalies Wastlen aie Ver bio eat Hp Y
pede Odi erfepoh iG. Squared — ~ value fo 7
Salk -alalaehes ee aE box,
> |
——_ =
— eee “3p Pa
1 a -rervran-dem variable % ey uni fea (o
ib B® creendom protege defined vu
|vCe\= ex
Ta
* Since V4) = 6e% depends only on & “end
Be | ot on ty ny 6 effectively a tendom
~ [variable crother then o ime = varying
b. | process Sane problem edKS For:
Meon oF v(t)
- “Mean squared value - i
F ews oseae ee |
Ven" or Vice yeice=@
50
fierexa =), ie maleate “2 1
(ers Dea
2
2)
Mewn Squared value of VC)
We Compose |
7
Eel Wet Rhee Fal Gbierk Die dari aa, Gul Sage)
—— A
eee ae a a ee ae
le je Saipan)
S90:
ARTES ena) ERD)
q
“Fined ans -
Peet eee eee
L'
© 7
sinple explanalion = =
Boi Carl ere
DBE sch ST arp) vay
i + Te sw wromdem varied e pun for die
bebveen 0 ord 2. pt
i fl ‘on)
: ie on eee as depends 2?)
CN net Secon basically just a
i condom vurio-ble.
B ae Memnrof vA)
- we wont +e Find the average Cmeen ) vedue of
* Pe Ere
|
| r
_fet A niform digitibution between o§ 2,
{Mle of V Ct) = 6. average of ex “From
| oO +> a
| <6) ft eee
| as = es) CRF 6 e oad
Men Syvaded value of
Sr
Now Squume VCE)
v CEE 2 Cie
$0
heiiFinal Ans
| yew of ee a
a wate O Nore on ebay iow 0f wncloy end
digitel Gemeuni cation Eashern s\n enone
T
oF noise.
ees Coram wilco! on
. Ye Onvloy sysrem 1 Signals ure contiwou.
«(owe © pnwented signale 9 Uisecty eftects
Ato ee Biqnedl quality. =
: | Even smell amovn: oF noise can distort
[Ane aignel , mulcing it harden Ao wetover
[ Are oniginal messu ae |
+ | PS _noige incteager , sre Signal becomes ters
clear ond more difficult +e sand evateund -
Exumpie - \n an old vudio ; stetitor cruckl{n
Sounds huppens due © noise -
“ae Anvloy Communication syayem renemit
MUSTO using ComMMous ayous ead very
ios ampiieded Frequency, on. phages 1h thee
cresence of mois|| e Date |
= ots 4
ae \ envoy: 3g
poise Sugcep nds eae various syp2s
gre Wighly Sur cep mon 9 a.
of Hoise C therm |
wie en distort 4
Exon eral Ot ones 4
“ Helee eg an eck Yow “signed lease -
a reduction jm fiddity Cer4 > hissing in
Orwdio or Show in video).
S Cumootove efFeetr : gloise accumulateg
over long Ustence ond through mult ple
| Sk8 93 of amplification or pro cegejng |
| Further -degeading the signal U
{, =
a Error Correction + cored 8
INS tems + om
[Awe vodyanved enron f2teeton pase
worrechon mechunigms ,
| digital) — communfaction —
H
~ Afgitol Systems,
Coinury’) |
a
Signal,
Snes Gre Sent af os Ug
1 1S \ OY 9 exh ath @natniatiaten \evel
e Teceiveyr - cam ST dkaur noise becomes Nem, Strong
Poceur: bul error dedection \ ond
korte cron — tesinniques como} Phen un EEK
ane
ample : Amobiie Phone call muy wemain
\ Gs en \ s r aa aa
dan: even vol some baekyround ™ polee .
ee cone oa
pighal communication system rangeit
(nor motion \ binary From CO and =)
won y Aig crete aged \evels
: 2 Woise mmounihy = oigired oiqnals ave—
ey .
| more robugt Fo Mose. AS Yong oy the
wvise dors not exceed av Centein thresholl,
ane original deadu com be perfec}
| ewevovertd
Signe) _Reqeneretion = pigntnl systems: cen
ceqenerobe Signal ot (nt erveda , eliminating
accumuloared Heise end emalnraning signal
Ler egaity over long dtasemtes.
Erno, petection ond correcton 7
pavenced Feclniq ued IKE parity ce neetxs 4
a L_albectksouma : and enror — conve CHINY cod et
un) pstd hn ese uchymesst s-€* gamma
by noise.
—————x&—<<,
s LY | weed of Sourte Coding
1 t
Sourte wiiny is+ne Poocefs oF converting
iInformution™ from & Source inte a formal thet
fe efficient for storage ov +vUNEMission
_ peedad -
+ | Reduce Redundorey |» IRhwe duuliniopiion -
contuing medun Sounrte Coding
_dleent-nutey nis xed nduncy to heduce_
che sive. od)
Efficient —Teansmission iamatler duty sizes
mewn Foster baransinigalon “i 6vey communicétta
Bie. | channels,
r
Efficient storage -
\35 Storage Sp| pros
@ (=)
“eoding =
ool ereproc# ee codin p
4h 1
= acs _ Punter encodin G oo
”
cana cen or Wes oe On ei eS chanr &
.——_____—— ec Hon |
I Ex- \nakead OF Sending pererse Bic ith
cS OT EMIO TS eae CN
[oourrce ling compresses t4_ Udi
tredhed! AKO) oteiraeim teen Y
1 erainnel es aaa OES
of bits. we)
w
ory RW
aot numb
y_ | Coding “EFF cheney
7 es
I coding efficiency oa measure of how Close. ug
|| Ye “average ~ \engiy of a code Cutter -
compression) %6 to Fhe enbopy of 4h eL
Sourte,
Formuda
EFFicieney
2
WG) = Entaop
| eretdenc
poo!')Wavy
fas)
ees
Dols tuba ionic TT ne
MY eG 9 og ao, alae
othe average coding
= Erenu erty, ottusriny ; aynbo\ yer sino
prea -saiiex “iyi oe tg. be “ae
a) demdiny merneds +© paewent
© | Requires _ Speck
—jarmniaig vi ty (ey Preete: coder) ___
je = _CHuetmen “ealing)
_ Given Symbol \ “u i
ple with “di Frerent probubili Fee’
a Lmost Frequent) =O
a =i
: c= $h0
pe 4thend hela _ probe shy
aynbols epedys
= Ue, So IS seeelieed lalena Sy = 16
NANG 1 Se =) | 16 sy 64 = Se Soir ae oes
*— SS esal So
eker mine _ | ound. LeendOa Gece
LEninopy We - Pyloya PZ
pveruge Wade enyth L= > Pili
EFF Geant Va ‘
Sort symbols by proioobi lity
Sort Yc given symbols in descendiny order
Cer pee
[Poe een Sp se
oo i
on sl a
Conssrust — shannon- Fano tode_
Owide he War inte Moo qeoure with sume
of proba pilittes sai sof Sot Flip -Fop
pe vietenctontnep ll Varela)
This 16 the slumber of input bite per needing
M grep. ——-
Code rare Ck\n)
i
weber of inpurs brie Aivided by
™
of outpue: ite py Saree
Krom "She Ate
nit + \ngi y ee Ut bite
ans Codemete = |
amy The Constraint engin 18 sche pamber of ince
YS thet ofFert the outpud , hich ig cayel te %
7
K = GCnumber of memory elements) 1 = a4 = 9
con S4rcun4 \engih =z
Sy) Glen contealies Seyuenced
«|| Binet 6lpy + seo of cumment olf M4) mit Gleam
SV hem rotor polynomined
Gye Debit}
+ ecco: eben: MOP obinemndry
heer tor pero
Cie = ED ot
Creme Tox ng Seyuency Cin ocial)
ai!
3 Gucee (OAs
S01 Generoting sey ences = C1 sin ot
{
wv) | ole Seyuened for Meseeje Seguenve
me}, + ye es i)Concupeniel the o|P 35
olP Sequ ence = tv Likjo to, O90, uhos
6 @
P| fer 2 eae aS" Linewr
i ee ke jigite
pike cher 4
he he
ies hee
\. Pe eecee
ee We i ie
“yous code
oiet Wives
U
1s ewdent thet
Ge. |i ee Oe
ence&)
C) | athe minimum -distence between any tod
|| codeword is 9, Hence, this je Gingle
errey cortentiny Code. Sinte herd cre
|_ _6 single esa cond 7 $4ndaemed ____
| wwe cam “Comreat el! _gingle “@acors emd ___
one + dowhle. e-paper Putter .
| Peewee ane ne obtained from -
| 62 eat g
ee, :
el Jolo}_|o |e] arhe ' codewords in oa iene ltocls code are
| locks of gyeniboolg hut are. encoded wing
more gyro} Atown oe original velue
Ao be Sent. : 2
Ac \neur Code of Vengtn Arong mits sblocks
containing n symbols. for ‘example ,
she (114.3) Hamming code bu Viner binuny
code whith repre okt 4 bit message
| wing 4-14 wdewords
q
| \+is aasimele exror condeo| coding terhnique_
| used Gor erroy detection Od covrertion.
[inform ution data is partitioned into
rioeks of jenyrh Pier Fo~ exumple
L Informution woord 6 Athen coded into eo lou
| of lenyth n bits cotled a codeword.
| etbytg dota D byt 4 codeword
ed et a
EncoderJ ;
Vinewr__Brouls
tl sys ems:
vey “pei
L\inewe : =the wde Follows gimple. math rudy |
6
codes ate O-
code vsed in agiel
abun maid
corme ering
Sond comet
dota ig sens
Bice code + The messoye
Bork _t Fixed = sire qerous
tke eddiny bis “ina KOR. i
codeword - Cosh message plore (5 ebunged
inte a longer block Ccotled w codeword) i
‘es
by addin exiva bite called purity bits
atoege help Bind and ie errors ~
Apolicubion -
oigvel communication 1 St et
G@HON 8
used In mobile ghoneg ) Susel\ite omm: af
iz Tine . Tec
ond wrrelas net
orks to Aetert
ond + FR
Awansfey ,ee
(3)
y | compurey sledwortes +
Yelps ensure thet duta ge
o
Cee ae ee
Tanternet Gite emaiia on § jeg igri aan
_enecveety
Gpate | Communication
vor useful | 9 sending detu to emd .feom
cudel\\*tS or Spacecraft 4 where enrors
ean €asily oceum,
S)_ | Broadeust Sygrem ,
3 used in digit) ory ond. radio beoudcastin
+o gaovide cleay slynals even In bud weethe
ot poor Signal acends,Wow a vocoder coors
Analysis
Modulation signe) (voice)
ny
re eS FARIaNS
LN Co Commer sinnel (ayntnesizer)
aN
Step -2 winduleition
Modulation signoh Coren)
Coin tnestoey)Votoders
Avowser analyre the spectral churueledoe |
—~ oof a meodulster aignel vogtenp speculaas |
— get thet 19 7 Fommetion 40, S9ePe -acanmier
a | etsood ' sa picolly Synthesired Soumd ,
oe ereoulhoy \n a synthetztd sound segue
with the “ATA bre! OF the moduluter ,
[=-k5 Perocess essemMully ena lyre jal
_gpeech , Compreses the Information into
ope cites) eney envdooes , end then epolics
hose evelooes to a different sound
Soudce »
s anakygis
the Votodey omulycey the WMeodulutor gi anc
Cen a voice) by dividing 4 Into mutiple_
medueat loands using a ei\ter ben.
Q. Emevelope Gdrothon
_ fer coon Freyuem4 bond athe vocoder nal
ache txiruched amplitude envelopes arte The
used jv condro\ ashe amplitude of cormsponding
frequency lownds in he cuarien signal Cera
, :y
a
I
My 2
nett
wep — wd 5
nati See sh gunrenized eee 35
ia aia antaee silane © 7 of tre 7
ee ae the sound”
she Sphase! aigne)
we
are imposed sn Hie cus
Le ©4_poncepis -
x] Medal stor jet
Gi aaa
thot paovides the spectre
The di gine
chuvactemsHeu -Cusualty geeeh)
; 7
¥ | cuwrler - =
| vine Gitined tet ge, Shaped py Whe
| Cu i
ound S0uere D. a
= Filter Bun :
A__Collectinon of band “poss Fiitera hed
Viuide the cudis (gna Jato APE bere
foreq umes Oe,
= Arp\\tude En ope -
AEP then LH OD
CS ee
of Anos Ahe. amplig d
ovis me
Fae uanuy ix™M Ghend for oaobal Syatremn fer mobsi\e_
commaniconion . Hi a Syfael moe Newark |
common wati\ized by mobile Prone users
amoung the world. at sscini he head
yw fs Whe most Populus of the thece dfgtiah
wire legs a elpinony SAsrems CTT DMA 1_GiSs9 ,
ond epmA) ond vag rhe combination
of FOMA) ond “Toma
\e we 4 different Frequensy bends |
85D Pte, Goomts , \g00 MIN, ond Woo Mie.
GSM converts ond tompresseds duta before
endiny it alony a chunnd with two other
a 7 f
arenmnG of user: dota, each with it, +ime
Shot
asm 4 Uffernt Sizey of cells
Moen -
7
\n_+nig give of dhe cell a ‘Bee
Staton antenna 16 ingtulled,
\n tnge size of Cali Gmtenna height
5 Neg Yum tne WwveTuge Poe \evehaf palin’ oureet ee 5 '
‘ _| spect il eta Oone es el eee 7
7. rece Win Sr =
_ hy Lo® power? aad deAces-
: Invernational 604 compen st SY: 4
\ La? bem Le Lost’ ss +
& New feuture omd setNicedi a
the Aroniecrure of HS) =
[trate phetworte Compras of ry
|. Fav clteel cwentks swonich combe boodly os
divided into . ©
—-
‘ | The Mobile Station (M43)
=a Sucne Stesrion Subogystem (855)
e Network gwoiihing,
|| —the operuh 07) ae Sub aistem me)
H Bip port Subs ateen ,ohers Calso cabled Morr speeeth
Grey
are the algorthms used (A Lge
as yem for mobile Communication) Networks Fo
See on, decor (ee a
“exe coders ploy, a cumcied nol ios
Ligited inenemission oF
eso oak S .
a enablin cP. ct ent
é woice: oven mobile
x | a tar coder ts a voice rode Lioder ~dewodsr )
shut ae
convets onaloy volte signals into digi te
formes -
© | compresses he dain for efficient Jeunimission
‘ pe compress {4 od the TeeeiVer'6 and
for pluyburk .
oak
«Tips - analog Voice feom the User,
7
S| digitizexon Ff converts vomuley voice to digit
a aoe oh eT Converter),
: Compmessions + Apes speed ‘compression Wing
_ospefie -valgortthm «ike Q@PE -LTP.
. caonsmission - sendy thes comprased -dade
sve the Gisrt vhedoont ,
Receiver decodiener | 9
I = Auyoam «
\ve
<< a ye eis
Deca ete ne oaks oat =
ne — oo ptt
I “Tree disyaum ———— og) represen
ee it ee Se
| Bhows ol\ po Gs
ed aeadture, euab lise,
pe “decision ene hyers
F _| 1416 + commonly 2A in 1
babii gomnd toding Aneo™ » Pure
L_convelwHonal enioding tak
I key Geatuorey *
L
. | Stents fom wstey sinyle soot Ande (int!
Syeute ) s!
a Loon level conresponds to a nevus Hme
rep or input “pi+
Bruched represent Possible
bosed on tne inpur
Eorh pan Prom he vooe as
AN ey Re a
one WG
Maren stay i MsExample We cose’ Bi
in_convolutionw Codiny , oo ree Ae. ure cle
can, Awe ow Ai fe an input “lpite. ¢ Meee
ourpur Se ff A aint on over tim
_aTralig ia yarn 19 wo -Compuct goa e'n a
meprtgentution of ne shade Lewooietbne of
La finite = stute machine over time + Yt
commonly used in convolutional cd
Virewt decoding sand Siynal processing
‘Ss a OG
Le Features |
Bibione pce Vigua erie ache hace
cred wndcaney hy reusing poles “(sha4e3) ever
swe.
Modes Cueviriter) represent -Gsores ot ore
Aime Gre-p
ce dyer [ioe eagle ghee opel Arungitions
ered on Sip bfx and show olp bite,
| key Features
| Sivat low hoo Aveo. Ava yerarn + PE seu aittbc ih
aed nduniy loa _vtusing nodes ( chates
onent cli
Modes ( vertices) Fepretent Shares at
| eon -diene shep
Bags Cbeonctnes ) Le paee§ eon arn: HoneeI
yoru
pret did
. p Nicwd!
ow aes
oe ercicivo®
oe spo le dots ok Aue
|| gen ances
4p Syed
| Gey t S
Ip Sonn jos Se
amie inblagtest *
| eonvoluHonnl coded »—_—
pr enialirgs miele eS
maciwat sm anaes pa 2
ch Coxyuunre of ghades.) —— $=
—
‘ cfu plles (RUA yan. cuore Ce na Um?
communi Cotton jac
=
x, Use Cte! -
Lm dewdiny o convolutional cade, the_ a
ArOis_- Ainqram helps ind the mos
—
Vikedy sAwune mitted bite See} hee Nin ;
dy amie (Pov yorumenin
3
| Gucnmar ee
| se
Lead re, Tee (Agen ee
H
Srvguce
Stanoniny growin
Sy uce
Qepertient puss |2
luge Sede. amderrd Sending —______che codin 1
encodinu Ponisi wi bewbtralge
iS
ear For aul = Scale exumple4 bay & Stod
geste tes es lice
decoding
Sree piuyrum for convoluHonw code.So) Grras
|_mepine P0e ond Shee The pow petit) oe
| at of deal ad 3 re
{east ee ded ae ee el
ocean opvinnal) stile a |
ona. past lon Vue.
comers o0e jetn)e ¥ ak
denoted up f (x) 1 gubrati ep :
\
ptasx Palo “subsystem Cage }+
Network end Gwitching Gub-
: 39S A
a Spen-otioN sulbay tern
8: BeneFi te
4) tnhwned decrerity
9) Vice cublind
3) intemetton Roming Serwe
b\ Tweet mesoye Si
5) guste -mobile- Internet Seance
4 Corto Whoert Hor
4 Yves = ppionit Por signed) Pe ccedion
i) torre + of e394 wad in broduced
in 13k | wboarderied LANG Geos i
Wid uisra ee comp,