Distributed Beamforming and Rate Allocation in Multi-Antenna Cognitive Radio Networks

We consider decentralized multi-antenna cognitive radio networks where secondary (cognitive) users are granted simultaneous spectrum access along with license-holding (primary) users. We investigate the problem of designing beam- formers for the secondary users by maximizing the minimum rate, subject to a limited sum-power budget and constraints on the interference level imposed on each primary receiver. We consider two scenarios: the first one allows only single-user decoding at each secondary receiver whereas in the second case each secondary receiver is allowed to employ advanced multiuser decoding and is free to decode any subset of secondary users. We provide an optimal distributed algorithm for the first scenario and an explicit formulation of the optimization problem corresponding to the second scenario. This problem however is non-convex and hence cannot be efficiently solved even in a centralized setup. As a remedy, we suggest a two-step approach. In particular, the beamformers are first designed assuming single user decoding at each secondary receiver. An optimal distributed low-complexity algorithm is then proposed to allocate excess rates to the secondary users, which are made possible due to the use of advanced decoders at the secondary receivers. Simulation results demonstrate the gains yielded by the optimal beamformers as well as the rate allocation algorithms.