Joint CFO and Channel Estimation for ZP-OFDM ... - Semantic Scholar

Report 2 Downloads 80 Views
Joint CFO and Channel Estimation for ZP-OFDM Modulated Two-Way Relay Networks

Gongpu Wang† , Feifei Gao‡ , Yik-Chung Wu∗ , and Chintha Tellambura† †

University of Alberta, Edmonton, Canada, ‡ Jacobs University, Bremen, Germany ∗ The University of Hong Kong, Hong Kong Email: [email protected] WCNC’10

Outline ■

Introduction



Previous Results



Problem Formulation



Proposed Solution



Performance Analysis



Simulation Results



Conclusion

Outline Introduction Previous Results Problem Formulation Proposed Solution Performance Analysis Simulation Results Conclusion

2

Introduction ■

Outline Introduction

Two-way relay networks (TWRN) can enhance the overall communication rate [Boris Rankov, 2006], [J.Ponniah, 2008].

Previous Results Problem Formulation Proposed Solution Performance Analysis Simulation Results Conclusion

 T1   f1

h1

-  R  fr

h2

 T2 -  f2

Figure 1: System configuration for two-way relay network.

3

Previous Results ■

Most existing works in TWRN assumed perfect synchronization and channel state information (CSI).



Channel estimation problems in amplify-and-forward (AF) TWRN are different from those in traditional communication systems.



Flat-fading and frequency-selective channel estimation and training design for AF TWRN has been done in [Feifei Gao, 2009].



Our paper will focus on joint frequency offset (CFO) and channel estimation for AF-based OFDM-Modulated TWRN.

Outline Introduction Previous Results Problem Formulation Proposed Solution Performance Analysis Simulation Results Conclusion

4

Joint CFO and Channel Estimation Problems in TWRN ■

With CFOs, the orthogonality between subcarriers will be destroyed in TWRN.



Even with completed estimation, data detection is not simple as circular convolution no longer exists.



How to estimate the mixed CFOs and channels and how to faciliate data detection?



We introduce some redundancy and modify the OFDM TWRN system to facilitate both the joint estimation and detection.

Outline Introduction Previous Results Problem Formulation Proposed Solution Performance Analysis Simulation Results Conclusion

5

Signals at Relay ■ Outline

The relay R will down-convert the passband signal by e−2πfr t and obtain

Introduction Previous Results

rzp =

Problem Formulation Proposed Solution

2 X

(N ) Γ(N +L) [fi − fr ]Hzp [hi ]si + nr ,

(1)

i=1

Performance Analysis Simulation Results Conclusion

where Γ(K) [f ] = diag{1, ej2πf Ts , . . . , ej2πf (K−1)Ts } and   x0 . . . 0    .. ..  .. si,0  . . .    si,1       .. [x] , si = FH ˜si =  .  H(K)  . x0  zp  xP   ..   . ..  .. .  . si,N −1 . .  0 . . . xP {z } | K columns



Next, R adds L zeros to the end of r and scales it by the factor of αzp to keep the average power constraint. 6

Signals at Terminal T1 ■

T1 will down-convert the passband signal by e−2πf1 t and get

Outline

(N +L) yzp =αzp Γ(N +2L) [fr − f1 ]Hzp [h1 ]rzp + n1

Introduction Previous Results

(N +L) =αzp Γ(N +2L) [fr − f1 ]Hzp [h1 ]   2 X (N ) × Γ(N +L) [fi − fr ]Hzp [hi ]si 

Problem Formulation Proposed Solution Performance Analysis Simulation Results Conclusion

j=1

(N +L) + αzp Γ(N +2L) [fr − f1 ]Hzp [h1 ]nr + n1 {z } |

(2)

ne



Next, using the following equalities h

i

(K) H(K) [f ] = Γ(K+P ) [f ]H(K) Γ(K) [−f ]x , zp [x] Γ zp

(3)

h i (P +1) Γ [f ]x Γ(K) [f ].

(4)

and Γ

(K+P )

[f ]H(K) zp [x]

=

H(K) zp

7

Signals at Terminal T1 ■ Outline

yzp can be rewritten as (N +L) (L+1) (N ) yzp =αzp Hzp [Γ [fr − f1 ]h1 ]Hzp [h1 ] s1 + ne

Introduction Previous Results

(N +L) (L+1) (N ) + αzp Γ(N +2L) [f2 − f1 ]Hzp [Γ [fr − f2 ]h1 ] × Hzp [h2 ] (5)

Problem Formulation Proposed Solution Performance Analysis Simulation Results Conclusion

(N +L)

(N )

(N )



We further note that Hzp [x1 ]Hzp [x2 ] = Hzp [x1 ⊗ x2 ] where ⊗ denotes the linear convolution.



Hence yzp is finally written as h i (N ) (L+1) yzp =αzp Hzp (Γ [fr − f1 ]h1 ) ⊗ h1 s1 + ne | {z } azp

+ αzp Γ

(N +2L)

(N ) f1 ]Hzp

[f2 − | {z } v

h

(Γ |

(L+1)

i

[fr − f2 ]h1 ) ⊗ h2 s2 , (6) {z } bzp

where azp , bzp are the (2L + 1) × 1 equivalent channel vectors and v is the equivalent CFO.

8

Joint CFO and Channel Estimation ■

We then obtain

Outline

y = S1 a + ΓS2 b + ne .

Introduction

(7)

Previous Results Problem Formulation Proposed Solution Performance Analysis



Since S1 is a tall matrix, it is possible to find a matrix J such that JH S1 = 0.



Left-multiplying y by JH gives

Simulation Results Conclusion

JH y = 0 + JH ΓS2 b + JH ne . | {z } | {z } G



(8)

n

Joint CFO estimation and channel estimation vˆ = arg max yH JG(GH G)−1 GH JH y, v

ˆ =(GH G)−1 GH JH y, b ˆ ˆ 2 b). ˆ =(SH S1 )−1 SH (y − ΓS a 1

1

(9) (10) (11) 9

Performance Analysis ■ Outline

At high SNR, the perturbation of the estimated CFO can be approximated by

Introduction Previous Results

∆v , vˆ0 − v0 ≈ −

Problem Formulation Proposed Solution

g(v ˙ 0) , E{¨ g (v0 )}

(12)

Performance Analysis Simulation Results

where g(v) = yH JG(GH G)−1 GH JH y.

Conclusion



The NLS estimation of CFO is unbiased and its MSE is 2 σne . E{∆v } = H H −1 H H ˙ ˙ 2b G [I − G(G G) G ]Gb 2



(13)

ˆ is unbiased and its MSE is The channel estimation b H ˙ H ˙ MSE{b} = (GH G)−1 GH Gbb G G(GH G)−1 E{∆v 2 } 2 + σne (GH G)−1 .

(14)

10

Simulation Results

−3

10

Outline

v numerical MSE N=16 v theoretical MSE N=16 v numerical MSE N=32 v theoretical MSE N=32

Introduction Previous Results Problem Formulation

−4

10

Proposed Solution Performance Analysis Simulation Results

−5

Conclusion

CFO MSE

10

−6

10

−7

10

−8

10

−9

10

0

5

10

15 SNR (dB)

20

25

30

Figure 2: Numerical and Theoretical MSEs of CFO versus SNR 11

Simulation Results

0

10

Outline

a numerical MSE N=16 a numerical MSE N=32 b numerical MSE N=16 b theoretical MSE N=16 b numerical MSE N=32 b theoretical MSE N=32

Introduction Previous Results Problem Formulation Proposed Solution Performance Analysis

−1

10

Conclusion

Channel Estimation MSE

Simulation Results

−2

10

−3

10

−4

10

0

5

10

15 SNR (dB)

20

25

30

Figure 3: Numerical and Theoretical MSEs of Channel Estimation versus SNR 12

Conclusion

Outline Introduction Previous Results Problem Formulation

1. Adapt ZP-based OFDM transmission scheme.

Proposed Solution Performance Analysis Simulation Results Conclusion

2. Suggest joint estimation method of CFO and channels. 3. Performance analysis: prove unbiasedness and give closed-form MSE expression.

13

Conclusion

Outline Introduction Previous Results Problem Formulation

1. Adapt ZP-based OFDM transmission scheme.

Proposed Solution Performance Analysis Simulation Results Conclusion

2. Suggest joint estimation method of CFO and channels. 3. Performance analysis: prove unbiasedness and give closed-form MSE expression. Problem: How to obtain individual frequency and channel parameters? (Globecom 2010)

13

Outline Introduction Previous Results Problem Formulation Proposed Solution Performance Analysis Simulation Results Conclusion

Questions and discussion? Email: [email protected]

14

Recommend Documents