A Novel Blind Adaptive Prediction&Subspace Code-Aided AR Narrowband Interference Suppression for CDMA Systems

In this paper, a projection approximation subspace tracking with deflation (PASTd) algorithm based on subspace technique for narrowband interference (NBI) suppression of code division multiple access (CDMA) is proposed to improve the performance of recursive least square (RLS) algorithm based on direct matrix inverse (DMI) code-aided technique firstly. Specifically, this technique is against the type of NBI, namely, autoregressive (AR) stochastic process. Then a recursive least square (RLS) prediction&PASTd code-aide algorithm is proposed to solve the problem of difficult determination on low order with above PASTd algorithm. In this paper, its signal to noise and interference rate (SINR) performance against NBI and MAI is compared with DMI RLS algorithm. It is seen that this method outperforms current DMI RLS technique.