Cognitive beamforming with cooperative feedback

Abstract

In cognitive radio networks based on spectrum sharing, cognitive beamforming (CB) enables the multi-antenna secondary user (SU) transmitter to suppress the interference to the primary user (PU) receiver and yet maximize the SU link throughput. In our prior work, a design framework was proposed for realizing CB based on finite-rate cooperative feedback from the PU receiver to the SU transmitter. The above cooperative feedback design is based on the orthogonal CB (OCB) algorithm. Specifically, the beamformer at the SU transmitter attempts to null the interference to the PU receiver by restricting the beamformer to be orthogonal to the SU-to-PU channel direction fed back by the PU receiver. In this paper, a new CB algorithm, called non-orthogonal CB (NOCB), is proposed by relaxing the orthogonality constraint for OCB. The NOCB algorithm computes the SU beamformer based on the feedback from the PU receiver including the quantized channel direction information (CDI) and interference power control (IPC) signal that is designed to regulate the SU transmission power. Furthermore, feedforward of the secondary-link CDI from the SU transmitter to the PU receiver is proposed, which enables more efficient IPC feedback design for the SU receiver and thereby allows larger SU transmission power. The proposed NOCB algorithms with cooperative PU feedback and SU feedforward are shown to substantially improve the SU link performance over the existing OCB designs.