Cyclostationary blind equalisation in mobile communications

Abstract

Blind channel identification and equalisation are the processes by which a channel impulse response can be identified and proper equaliser filter coefficients can be obtained, without knowledge of the transmitted signal. Techniques that exploit cyclostationarity can reveal information about systems which are nonminimum phase; nonminimum phase channels cannot be identified using only second-order statistics (SOS), because these do not contain the necessary phase information. Cyclostationary blind equalisation methods exploit the fact that, sampling the received signal at a rate higher than the transmitted signal symbol rate, the received signal becomes cyclostationary. In general, cyclostationary blind equalisers can identify a channel with less data than higher-order statistics (HOS) methods, and unlike these, no constraint is imposed on the probability distribution function of the input signal. Nevertheless, cyclostationary methods suffer from some drawbacks, such as the fact that some channels are unidentifiable when they exhibit a number of zeros equally spaced around the unit circle. In this thesis the performance of a cyclostationary blind channel identification algorithm combined with a maximum-likelihood sequence estimation receiver is analysed. The simulations were conducted in the pan-European mobile communication system GSM environment and the performance of the blind technique was compared with conventional channel estimation methods using training. It is shown that although blind equalisation techniques can converge in a few hundred symbols in a time-invariant channel environment, the degradation with respect to methods with training is still considerable. Yet, the fact that a dedicated training sequence is not needed makes blind techniques attractive, because the data used for training purposes can be re-allocated as information data. In the concluding part of this thesis a new blind channel identification algorithm which combines methods that exploit cyclostationarity implicitly and explicitly is presented. It is shown that the properties of cyclostationary statistics are exploited in the new algorithm, and enhance the performance of the technique that solely exploits fractionally-spaced sampling. The algorithm is robust in the presence of correlated noise and interference from adjacent users.