Adaptive arrays are becoming an essential part of future mobile radio systems where they can be used to provide cancellation of co-channel interference. Current full-band adaptive arrays based on the LMS algorithm offer simplicity and robustness but, in some applications, the LMS algorithm suffers from slow convergence rates. In such applications, faster algorithms would normally have to be used at the expense of increased complexity and cost. Recently, however, a novel quadrature mirror filter- based- multirate sub-band adaptive beamforming technique (QMF-MSAB) has been presented by Khalab and Ibrahim (see Proc. IEE 3rd European Conf. on Satellite Communications, Manchester, UK, p.55, 1993, and Electronics Letters, vol.30, p.1194-1195, 1994). This technique offers considerable improvements in terms of convergence rates and cancellation performance over full-band beamforming using the normalised LMS (NLMS) algorithm. In sub-band multirate adaptive filtering, the adaptive process is applied in adjacent sub-bands in parallel at lower sampling rates than if a single band were to be used. The improvements in convergence rates are a direct result of applying the adaptive process in sub-bands, where the adaptive step size of the LMS algorithm could be matched to the smaller amounts of energy in each sub-band. A parallel and a serial version of a two-band, QMF-MSAB are investigated. Computer simulation results, that confirm the potential of these techniques for use in future mobile communication systems, are presented.
展开▼
