1

EMLAB

Wireless propagations

2

EMLAB

channelRFDSP데이터

송신

RF DSP

수신

데이터

1. Wireless channel model

• 무선 통신 기기 사이의 전파 송수신 예측 어려움 .

• 단말기 설계를 위해 channel 특성 파악 필요 .

3

EMLAB

근사 모델 #1: Free space model

2

4

RGGPP RTTR

R

Rc

j

RT eR

GGH

4)(

(Friis formula)

• 거리 감쇄 효과만 고려 .

• 초기 설계에 유용 .

4

EMLAB

Antenna modeling

)/(4

cRtR

R

)ˆˆ(),(4 hT,vT, TT

jkR

TTTT

in hhR

eG

P

E

안테나에서 나온 신호는 거리 R 떨어진 지점에서 진폭이 1/R로 줄어들고 시간 지연이 R/c 만큼 생긴다 .

)(t

5

EMLAB

Field test route (outdoor)

Guidelines for drive testing to accurately estimate average cell radius or for tun-ing a path loss model.

6

EMLAB

Field test route (indoor)

Typical in-building measurement route and signal strength.

7

EMLAB

Measured path loss

8

EMLAB

Physical model

Impu

lse

resp

onse

Time

dthstxt

R )()()(

)(th

Measurement Simplified model

9

EMLAB

Propagation channel model

tfjtfjii

iiR

cic eettstx 2)(2)]([)(Re

Impu

lse

resp

onse

Time

Spec

tru

m

Frequency

Amplitude variation

time spread

Mathematical model

αi(t), τi(t) 가 channel 특성 결정 .

10

EMLAB

Hierarchical fading process : statistical model

11

EMLAB

Propagation loss mechanisms : scaled model

1. Large Scale Fading:due to distance

2. Medium Scale Fading: shadowing and obstacles

3. Small Scale Fad-ing: due to multipath

12

EMLAB

The three scales of mobile signal variation

13

EMLAB

(1) Large scale fading

)/()(][)( 00 dddPLdBdPL

( 거리와 path loss 를 fitting 하여 β 값을 찾음 .)

14

EMLAB

근사 모델 #2 : Two-ray Ground Reflection

21

21

4)(

R

eR

R

eGGH

Rc

j

S

Rc

j

RT

15

EMLAB

• Measure PL(d0) or calculate Free Space Path Loss.

• Take measurements and derive empirically.

근사 모델 #3 : Log-Distance Model

Environment Exponent (β)

Free space 2

Urban area 2.7~3.5

Shadowed urban area 3~5

Indoor LOS 1.6~1.8

Indoor non-LOS 4~6

)/()(][)( 00 dddPLdBdPL

16

EMLAB

Classification of large scale propagation models1. Macro cell model

① Empirical model

a. Log-distance path loss

b. Okumura-Hata

② Physical model

a. COST231/Walfisch-Ikegami Model

③ ITU-R model

2. Micro cell model

④ Empirical model

a. Dual slope model

⑤ Physical model

a. Two-ray model

b. Street canyon model

c. Random waveguide model

d. Ray tracing model

⑥ ITU-R model

3. Pico cell model

⑦ Empirical model

⑧ Physical model

17

EMLAB

Okumura-Hata model (COST207)

1. 150MHz~1.5GHz 에서 유용 .

2. 측정을 통해 얻은 데이터임 .

3. Open area, Suburban area, Urban area 로 분류함 .

ERBALdB log

CRBALdB log

DRBALdB log

1.1)56.0log56.1()7.0log1.1(

3001.1))54.1(log(29.8

30097.4))75.11(log(2.3

94.40log33.18)(log78.4

4.5))28/(log(*2

log55.69.44

log82.13log16.2655.69

2

2

2

2

CmC

Cm

Cm

CC

C

b

bC

fhfE

MHzfh

MHzfhE

ffD

fC

hB

hfA

For large cities

For medium to small cities

frequencycarrier :[MHz]f

mobile andstation base

between distance circlegreat :R[km]

1.5m) (oftemheight

antennastation mobile:][

C

mhm

Urban areas :

Suburban areas :

Open areas :

18

EMLAB

COST231-Hata model

GERBFLdB log

areasan metropolit3

areassuburban and cities sized medium0

log82.13log9.333.46

G

hfF bC

1. 1.5GHz~2GHz 로 Hata 모델 확장

2. 측정을 통해 얻은 데이터임 .

3. Open area, Suburban area, Urban area 로 분류함 .

19

EMLAB

Walfisch-Ikegami Propagation Model

sdmsdFdB LLLL Lf : free space loss

Lmsd:multiple knife edge diffraction

Lsd : single diffraction

20

EMLAB

(2) Medium scale model : shadowing

SdB LLL 50

21

EMLAB

Typical variation of shadowing with mobile position at fixed BS distance

Probability density function of shad-owing. Measured values are pro-duced by subtracting the empirical model.

Log-normal distribution : shadowing

2

2

2exp

2

1)(

S

S

L

S

LLp

22

EMLAB

LzxSz SS

zQdx

xdLLpzL

L /

2

2exp

2

1)(Pr

Shadowing effect on coverage

23

EMLAB

(3) Small scale fading

tfjc

cetuttftats 2)(Re)(2cos)()(

)()()()( )( tjytxetatu tj

Probability density function of real part of NLOS fading signal→ Gaussian

Multi-path fading : NLOS

24

EMLAB

Equal delay ellipse : multi-path fading

25

EMLAB

Complex samples of NLOS fading signal Theoretical and experimental Rayleigh distributions

Rayleigh fading

26

EMLAB

Rician fading

2

2 2/

s

P

Pk

NLOS

LOS

kr

Ieer

rp kr 2)( 0

)2/(2

22

Rice factor :

27

EMLAB

c

,1

12

)cos1()( υk

][ || tυrrr

c

kυ

kkk ||

][|| HυEEE

][|| EυHHH

HE,

HE ,k

υ

동일한 전자기 현상이지만 관찰자의 상대적 운동에 의해 E, H, k, 주파수 등이 변한다

2c

ttrυ

Frequency shift by Doppler effect

28

EMLAB

1f

2f

Beat frequency due to Doppler effect

)cos1( ff

υ

다중 경로 신호의 주파수가 달라져 beat frequency 관찰됨 .

29

EMLAB

The received signal of a time varying channel where the transmitted signal is a time-harmonic signal with frequency f =10Hz.

Time varying channel – Doppler shift

Dc f

T16

9

30

EMLAB

Time varying channel : auto-correlation

Power spectral density

)2()()( 0* Dlll fJPtataE

otherwise0

if)/(1

2

)2()( 20

D

DDD

ffffffJFfS

( fD : maximum Doppler frequency.)

( )S F

Dff

Jakes model

For time delay profile

Dc f

T16

9Coherence time :

31

EMLAB

Wideband fast fading

c

rr 21

]1[

)()(])([

1

])([2

])([1

11

2211

jtj

tjtj

eea

eaeaH

)]cos(1[2)( 1 aH

32

EMLAB

Multi-path Fading

Impu

lse

resp

onse

Time

• Flat fading :

• low data rate.

• Ex) AMPS, ASK, ...

Time

Impu

lse

resp

onse

Sd T

• Frequency selective fading :

• high data rate 인 경우에 생김 .

• Ex) WCDMA, DMB 등 ...

Sd TInter-symbol interference

ST

STSymbol Time

d

d

33

EMLAB

Multi-path Fading

Multi-path Fading 의 영향

•협대역 신호인 경우 영향이 작다 .

광대역 신호인 경우 신호 spec-trum 에 미치는 영향이 크다 .

34

EMLAB

l

tffjll

llcetxtaty ))((2)()(Re)(

…shift in time …

…shift in frequency …

… attenuation…

(this causes small scale time variations)

paths

Wide-band fast fading channel model

35

EMLAB

Channel sounding

n

iii

T

PP

10

n

iii

Trms P

P20

21

rmscB 2

3

Coherent B. W. :

36

EMLAB

Overcoming narrow band fading

Diversity

37

EMLAB

Overcoming wide band fading

Linear equalizer

38

EMLAB

Adaptive equalizer

39

EMLAB

OFDM receivers

OFDM Subcarriers in the Frequency Domain

40

EMLAB

. . .

0w 1w 2Nw 1Nw

12

각도 추정

.

.

.0w 1w 2Nw 1Nw

RF, modem

다중 빔 안테나

각도 추정 알고리즘 다중 빔 생성

Adaptive antenna