A Comparative Performance Investigation of Adaptive Beamformers with Iterative Weight Refinement for GSM Signal Detection

In a multiple-access communication system, many users utilize the same spectral bandwidth by transmitting, for example, during assigned time slots using time-division multiple access (TDMA). The users of a cellular system based on TDMA can encounter cochannel interference because carrier frequencies are reused in nearby cells. In this paper, we investigate adaptive beamforming and equalization algorithms for cochannel signal detection in the Global System for Mobile communications (GSM). In particular, we are interested in iterative weight refinement techniques where the beamformer weights are successively adapted using re-encoded TDMA data to improve signal estimation. The data of a GSM channel are processed in two stages that include the following series of steps: frame synchronization, adaptive beamforming, signal detection, adaptive equalization, and soft-input decoding. The algorithms are evaluated using the number of frames passing the parity check as the performance measure for real GSM data collected with an array of antennas