Multiuser scheduling with limited feedback for cognitive radio

In this paper, we propose an underlay spectrum sharing scheme where a secondary broadcast system shares the spectrum with a pair of primary users. Both primary and secondary users are discrete rate-power adaptive systems equipped with limited feedback of the channel quality information. By overhearing the primary feedback and receiving the feedback from each secondary receivers, the best secondary receiver is scheduled at the secondary transmitter which chooses an appropriate power-rate pair from the pre-designed quantization codebook. The secondary quantization codebook is designed with the aim of finding the optimal power and rate allocation which maximize the average secondary rate under the average secondary power constraint and the primary rate loss constraint. Numerical results illustrate that, the secondary performance improves with the increase of secondary feedback load. In addition, for a fixed total feedback load, increasing the number of the secondary receivers brings more benefits than increasing the number of feedback bits at each secondary receiver.