Novel adaptive signal processing algorithms for multiuser detection

The paper is concerned with developing novel multiuser detection algorithms. The main stress is on blind decorrelation of weakly stationary processes which can successfully combat multiuser interference (MUI) and intersymbol interference (ISI). The detector architecture includes an adaptive channel identifier (capable of blind decorrelation), followed by a stochastic Hopfield net which performs the task of multiuser detection. The key contribution to mobile communication lies in two factors: (i) blind decorrelation which can identify the channel even in the case of fading; (ii) the use of the stochastic Hopfield net which can guarantee good performance due to its ability to reach the global minimum of the associated quadratic form. The proposed detector structure is proven to be superior to the RAKE or Viterbi receivers. The results are demonstrated by simulations as well.