Minimum-BER linear multiuser detection for DS-CDMA signals under impulsive noise environments

In this paper, the minimum-bit error rate (MBER) linear multiuser detection (MUD) for DS-CDMA signals is considered in impulsive noise environments, which are modeled as two Gaussian noises with different variances mixed by a Bernoulli distribution. A conjugate gradient algorithm for MBER detector is developed, which provides an efficient means of finding a MBER solution. Computer simulations for the MBER linear multiuser detector show that the performance under Gaussian mixture noise environments is inferior to that under Gaussian noise environments.