Hybrid Opportunistic Scheduling in Cognitive Radio Networks

In a cognitive (secondary) multiple-access network which is subject to interference power constraints imposed by a primary system, it is desirable to mitigate the interference on the primary and to harvest multiuser diversity gains in the secondary. To simultaneously achieve these goals, a two-step (hybrid) scheduling method is proposed that pre-selects a set of secondary users based on their interference on the primary, and from among them selects the user(s) that yield the highest secondary throughput. The optimal number of active secondary transmitters is characterized as a function of the primary interference constraint, the secondary transmit power, and the number of secondary transmitters n. The secondary sum-rate (throughput) of the proposed algorithm grows optimally (proportional to log n). We investigate the tradeoff between scaling the secondary throughput and reducing interference on the primary, and characterize the optimum tradeoff in the regime of large n. Finally, we study user scheduling under fairness constraints, which is necessary when the channel statistics of secondary nodes are not identical. A modified hybrid scheduling rule is proposed to ensure user fairness, while still achieving the optimal growth rate for the secondary throughput.