Scribd
Upload
10 views

Lecture 05 Space Time Coding

The document discusses space time coding for mobile radio channels. It begins by describing mobile radio channels and the fading effects caused by multipath propagation. It describes statistical models for fading channels including Rayleigh and Rician fading. It then discusses different types of diversity techniques used to mitigate fading effects, including time, frequency, and space diversity. It focuses on space diversity, describing receiver diversity techniques like selection combining, switch combining and maximal ratio combining. It also discusses transciever diversity which provides diversity gain through multiple transmitters.

Uploaded by

劉力瑋
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
10 views

Lecture 05 Space Time Coding

The document discusses space time coding for mobile radio channels. It begins by describing mobile radio channels and the fading effects caused by multipath propagation. It describes statistical models for fading channels including Rayleigh and Rician fading. It then discusses different types of diversity techniques used to mitigate fading effects, including time, frequency, and space diversity. It focuses on space diversity, describing receiver diversity techniques like selection combining, switch combining and maximal ratio combining. It also discusses transciever diversity which provides diversity gain through multiple transmitters.

Uploaded by

劉力瑋
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

LECTURE 05 SPACE TIME CODING

June 18, 2019

Li-Wei Liu
National Chiao Tung University
Institute of Electronics Engineering
Contents

1 Introduction 2
1.1 Mobile Radio Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Fading Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Fading Effect Description: . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 Statistical model for Fading Channels . . . . . . . . . . . . . . . . . 3
1.2.3 Rayleigh Fading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.4 Rician Fading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Space Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1.1 Receiver Diversity . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1.2 Transciever Diversity . . . . . . . . . . . . . . . . . . . . . . 5

1
Chapter 1

Introduction

1.1 Mobile Radio Channels


1. Scattering

2. Reflection

3. Diffraction

4. Line of Sight

5. Mobile Receiver

(AWGN is only applied for WireCommunication and Line of Sight Channel)

1.2 Fading Effect


• Multi-path Propagation → Flat / Frequency Selective Fading → Coherence Band-
width

• Time-Varying → slow and fast fading → Coherence Time

1.2.1 Fading Effect Description:


Divide the propagation into 2 factors, one in δ(t − τn ) the other in e − j 2π f c τn (t )
The Equivalent Low Pass Channel Impulse Response:
− j 2π f c τn (t )
n αn (t )e δ(t − τn )
P
C (τ; t ) =

2
3

C (τ; t ) = αn (τ; t )e − j 2π f c τ
Given Uncorrelated Scattering:
Auto correlation of Channel

φC (τ1 , τ2 ; t 1 , t 2 ) = E [C (τ1 ; t 1 )C ∗ (τ2 ; t 2 )] (definition)


= φc (τ1 , τ2 ; ∆t ) (Wide-Sense-Stationary)
= φc (τ1 , ∆t )δ(τ1 − τ2 ) (Uncorrelated Scattering Form)
= φc (∆τ; ∆t ) (General Form)

Considering MultiPath Propagation Only ∆t = 0 : (Arrival Time Difference)


F our i er Tr ans f or m over τ
c(τ; t ) −−−−−−−−−−−−−−−−−−−−→ C ( f τ , t )
F
φC (τ; 0) = 12 E ∗ [c ∗ (τ1 ; t 1 )c(τ2 ; t 2 )] −
→ φC (∆ f τ ; 0) = E ∗ [C ( f τ1 , t )C ( f τ1 + ∆ f , t + ∆t )]

1
(∆ f )c ≈ (Coherence Bandwidth)
Tm
Considering Time Variation Only∆τ = 0 : (Observation Time Difference)
F our i er Tr ans f or m over t
φC (∆ f τ ; ∆t ) −−−−−−−−−−−−−−−−−−−−→ SC ( f τ ; λ)
Since ∆τ = 0,SC ( f τ ; λ) = SC (0; λ)

1
(∆t )c ≈ ((∆t )C : Coherence Time, B d : Doppler Bandwidth)
Bd

1.2.2 Statistical model for Fading Channels


x l (t ) = kαke j ∠α

1.2.3 Rayleigh Fading


• No Direct Path, consist of reflected wave only

• Two Quadrature component (in-phase and quadrature phase) are uncorrelated Gaus-
sian random process with zero mean and variance (σ2s )

1.2.4 Rician Fading


• Consist of both line of sight and reflected wave
4

• Rayleigh is the special case of Rician (when K = 0)


2
• Rician Factor K = 2σ
D
2 , where D : direct signal power, σs : received signal variance
2
s

- For an AWGN channel, the error rate decreases exponentially with the increasing
SNR.
- For a Rayleigh fading channel (Worst Case Channel Model), the error rate decrease
inversely with the SNR.

1.3 Diversity
1. Time Diversity
(Multiple transmission of same info over different time slot(time spacing > coherence
time)
e.g. ECC + Interleaving

2. Frequency Diversity
(Multiple transmission of same info over different frequency slot(frequency spacing
> coherence bandwidth)
e.g. (Spread Spectrum :CDMA, Multi-Carrier Modulation:OFDM)

3. Space Diversity (Multiple Tx Rx)


Large seperation between antennas (> λ2 )
Polarization Identity (Clockwise & Anti-Clockwise of EM Wave)
Angle Diversity (Cell Base Station Direction)

1.3.1 Space Diversity


1.3.1.1 Receiver Diversity

1. Selection Combining (Choose the highest Strength(SNR also)of Nr Receiver)

2. Switch Combining (Switch as SNR is below the threshold and activate next RF
front end estitmation)

3. Maximal Ratio Combining (Weighted Sum of Strength(SNR also))


Pn R
r = i =1 αi ·r i , where r i : received signal , αi : weight factor, we could further choose
αi = A i e j φi for Phase Cancellation
5

4. Equal Gain Combining (Sum of Strength(SNR also))


αi · r i , where r i : received signal , αi : weight factor, we choose αi = 1e j φi
Pn R
r = i =1
for Phase Cancellation only

- Diversity gain would finally saturate.

1.3.1.2 Transciever Diversity

- Diversity gain would finally saturate, when doing comparison, multi-transciever means
multiple transmitted power. In order the compare the BER, we would normalized the
power, therefore the best scenario case is AWGN signal without fading.

You might also like