Tradeoff analysis of partial zero-forcing beamforming in large random spectrum sharing networks

In wireless networks, the availability of multiple transmitting antennas can be exploited in several ways. One way is to reduce the interference to unintended receivers by placing beamforming nulls in those directions, as is done in zero-forcing beamforming. Another is to boost the received signal strength at the desired receiver using “matched” transmit beamforming. This work studies a partial zero-forcing beamforming scheme that combines these two traditional transmission schemes, and examines its performance gain in large random spectrum sharing networks. Specifically, beamforming vectors are designed to null the aggregate interference towards a number of selected primary receivers and to boost the received signal strength at the intended receiver. Considering Poisson distributed users and Rayleigh fading channels, the statistical characterizations of signal and interference are analyzed and the optimal number of nulled receivers is discussed. The simulation results show that by employing partial zero-forcing beamforming with a judiciously chosen number of nulls, the permissible density of secondary transmitters can be maximized while protecting the successful transmission probabilities of individual networks.