Adaptive Channel Estimation in Wireless Communications - Adaptive Signal Processing in Wireless Communications - page 61

Viterbi Detector: K = N = 1, L = 2, T = 420, T s = 25µs, TIR = 0.3, P = 7, f d = 200 Hz, 500 runs.

10 0

10 −1

10 −2

SI&DPS: step1
SI&DPS: 1st iter.
SI&DPS: 2nd iter.
SI&DPS: 3rd iter.
SI&CE: step1
SI&CE: 1st iter.
SI&CE: 2nd iter.
SI&CE: 3rd iter.
TM&DPS
TM&CE

10 −3

10 −4

10 −5

0 5 10 15 20 25 30

SNR (dB)

FIgure 2.8 BER versus SNR for f d = 200 Hz.

Viterbi Detector: K = N = 1, L = 2, T = 420, T s = 25µs, TIR = 0.3, P = 7, f d = 200 Hz, 500 runs.

0

−5

−10

−15

−20

SI&DPS: step1
SI&DPS: 1st iter.
SI&DPS: 2nd iter.
SI&DPS: 3rd iter.
SI&CE: step1
SI&CE: 1st iter.
SI&CE: 2nd iter.
SI&CE: 3rd iter.
TM&DPS
TM&CE

−25

−30

−35

−40

0 5 10 15 20 25 30

SNR (dB)

FIgure 2.9 MSE versus SNR for f d = 200 Hz.

statistics-based estimator (denoted as step 1 in the figures), and the DML approach after
one, two, and three iterations (denoted as 1st iter., 2nd iter., and 3rd iter. in the figures),
and TM training approaches (denoted as TM in the figures). From the four figures, we
can see that after iterations, superimposed training-based estimation and detection per-
formances improve a lot, because the information data that are viewed as interference
by the first-order statistics-based estimator are now exploited to enhance the channel
estimation for the next iteration. Therefore, the self-interference is effectively removed