Computationally Efficient Joint Demodulation For Sensor Arrays

Statistical Signal Processing Inc, Yountville CA

Investigators

Abstract

Computationally Efficient Joint Demodulation for Sensor Arrays Abstract The objective of the proposed research is to realize order-of-magnitude improvements in the interference mitigation capability of an antenna array with its associated multi-channel digital signal processing algorithms, compared with performance presently attainable using conventional linear processing. The motivating commercial applications include high-capacity cellular telephony, personal communications systems, and wireless local area networks. Such huge increases in interference mitigation capability translate to similar magnitude increases in system capacity. With no end in sight to the accelerating demands on communications capacity, the commercial value of a breakthrough of this type would be tremendous. The signal processing method being investigated is joint demodulation of multiple signals using a multichannel Viterbi algorithm modified to trade signal quality for computational cost. The goals are to 1) characterize the number of interfering signals that can be separated and 2) characterize the degree of reduction in complexity of the Virterbi algorithm that can be realized while maintaining acceptable bit-error-rate, both as functions of the number of antenna elements in the array, and parameterized by signal-to-noise ratio. The interference environments simulated model cellular telephone channels with GSM and/or IS-136 signaling.