The number of iterations required to accomplish this task is dependent on the channel
conditions, interpolation polynomial used, and desired timing offset requirement.
6.4.3 DFT Based
The next timing error estimation method is based on calculating the Fourier transform of a signal to
create a useful spectral component. The timing error is related to the phase of this spectral component.
The method used to perform the transformation can either be discrete Fourier transform (DFT)
or fast Fourier transform (FFT) based [26].
We can continue to assume the clock controlling the Analog to Digital Convertor (ADC) is free
running as shown in the block diagrams above. The timing error is generated by first squaring the
received signal and then the DFT is computed using LN samples, where N is the oversampling factor.
The block diagram for this technique is given in Fig. 6.31.
324 CHAPTER SIX
r(k) | |2 DFT (??“1/2??)arg{ }
z (k)
FIGURE 6.31 DFT-based timing error estimator.
The timing error estimator is mathematically represented as
(6.103)
where N is the number of samples per symbol, L is the length of the symbols used in the average, and
z(k) is the sampled squared value of the received signal. Hence for each group of samples, a new timing
error estimate is obtained and used for timing correction/compensation.
Pages:
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536