Constructing quasi-optimal subfamilies of GMW sequences suitable for CDMA applications

Gordon-Mills-Welch (GMW) sequences possess a desirable two-valued autocorrelation function which helps a RAKE receiver track multipath returns. Unfortunately, their cross-correlation functions are less well controlled. Therefore, it is necessary to construct their subfamilies (taking advantage of its reasonably large family size) which offer satisfactory cross-correlation characteristics. The authors study several algorithms for constructing those quasi-optimal subfamilies in terms of minimised bit error rate under cochannel interference. The results show that performance of resultant subfamilies is very sensitive to the algorithms applied and subfamily sizes. A new criterion based on combined (even and odd) maximum cross-correlation is introduced for subfamily construction and two proposed highest-peak-deleting and most-peak-deleting algorithms are effective to construct the GMW quasi-optimal subfamilies