Adaptation Techniques and Enabling Parameter Estimation Algorithms for Wireless Communications Systems - Adaptive Signal Processing in Wireless Communications - page 17

1.4.1.1 Channel Estimation with A Priori Information
Channel estimation is an integral part of standard adaptive receiver designs used in dig-
ital wireless communications systems. For conventional, coherent receivers, the effect
of the channel on the transmitted signal must be estimated to recover the transmitted
information. As long as the receiver estimates what the channel did to the transmitted
signal, it can accurately recover the information sent.
The estimation of time-varying channel parameters is often based on an approximate
underlying model of the radio channel. In fading environments, the coefficients of a
channel model exhibit typical trends or quasi-periodic behavior in time, frequency, and
space. The ability to track channel variation depends on how fast the channel changes in
time, frequency, and space. As mentioned before, this is related to Doppler spread (time
variation), delay spread (frequency variation), and angle spread (space variation). By uti-
lizing a priori information about the channel variation, adaptive algorithms with larger
memories can be designed without sacrificing tracking capability [15]. In contrast to the
algorithms that do not exploit this information, adaptive algorithms provide a means of
extrapolation of the channel coefficients in time, frequency, and space [13, 52]. For exam-
ple, in [53], the step size of a simple least mean square (LMS) channel tracker is changed
using the Doppler spread information. Similarly, the window size of a sliding window
(moving average filtering)-based channel tracking algorithm can be adapted depending
on Doppler spread and SNR information [54]. Wiener filtering, which is one of the most
popular techniques for channel estimation using interpolation, is an excellent example
in exploiting a priori information, as the optimal Wiener filter design requires knowl-
edge of Doppler spread and noise power. In most conventional Wiener filtering designs,
the worst-case expected Doppler spread values are used, degrading the performance of
the algorithm for other Doppler spread values [55]. Recently, two-dimensional interpo-
lation using Wiener filtering for OFDM-based wireless communications systems gained
significant interest [28]. In this case, both Doppler spread and delay spread information,
as well as noise variance estimates, can be used to optimize the channel tracker perfor-
mance. Although we have mentioned a few examples, the usage of a priori information
in channel estimation has been considered by many other authors. Further information
can be found in [47, 48].
Figure 1.7 shows a simple coherent receiver structure with an adaptive channel
tracker. The receiver includes a parameter measurement block that estimates the neces-
sary parameters for the adaptation of the channel tracker. The necessary parameters
can be estimated using the received signal and the output of the detector as described
before. The detector requires the channel estimates that can be obtained from the chan-
nel tracker.
1.4.1.2 Adaptive Channel Length Truncation for Equalization
Time dispersion in wireless systems can cause ISI, which degrades the performance, often
severely. Equalization is a technique used to counter the effects of ISI. In the Telecom-
munications Industry Association/Electronics Industry Association/Interim Standard
136 (TIA/EIA/IS-136, or simply IS-136) system, the channel can be assumed to be flat
(nondispersive) with respect to the symbol duration most of the time. Equalization does