Novel Multiple Relay Selection Schemes in Two-Hop Cognitive Relay Networks

This paper studies multiple relay selection schemes in two-hop cognitive relay networks to maximize the signal-to-noise ratio (SNR) of the secondary receiver. We first formulate the relay selection as a nonlinear integral programming, and then propose two suboptimal methods, greedy scheme and hybrid scheme. Based on the sorted efficiency function, the greedy scheme selects the secondary relay nodes provided that the sum of the interference power does not exceed the interference power constraint (IPC) of the primary users. The hybrid scheme is proposed to overcome the performance degradation of the greedy scheme whose efficiency function is not the optimal ordering function. Numerical results indicate that the greedy scheme can sometimes, but not always, achieve the near-optimal SNR performance. The hybrid scheme always outperforms the greedy scheme and any best single scheme. Moreover, both the greedy scheme and hybrid scheme are of low complexity compared to the exhaustive scheme.