Statistical Delay QoS Driven Power and Rate Allocation for Cognitive Multi-Relay DF Networks

In this paper, we propose a statistical delay Qualityof-Service (QoS) driven power and rate allocation scheme for cognitive multi-relay Decode-and-Forward (DF) networks, which aims at maximizing the relay network throughput under the given delay QoS constraint. Specifically, by integrating the information theory and the concept of effective capacity, our derived optimal power and rate allocation scheme can maximize the effective capacity of the cognitive relay network subject to the given delay QoS constraint, which can be characterized by the QoS exponent θ, and a serize of power constraints, such as the average total transmit power constraint and the average and peak interference power constraints imposed by the Primary User (PU). Simulation results show that the relay network throughput decreases when the delay QoS constraint becomes stringent and the relay network can achieve better throughput performance when the number of relays increases. Moreover, we also find that while utilizing more relays, the average interference power perceived by the PU will reduce and the primary network can achieve better performance.