Performance optimization through multi-agent sensing framework based on interference for cognitive radio

Cognitive Radio (CR) can effectively address the spectrum scarcity problem by dynamically utilizing unoccupied bands of licensed users. However, secondary user (SU) needs strong spectrum sensing capability as it may be a heavy burden in the dense and heterogeneous networks. To alleviate SUs burden and improve the performance of the network, a multi-agent spectrum-sensing framework based on interference is proposed. We deduce its average achievable throughput and demonstrate its superior performance through a series of simulations. Considering that Sensing Agents (SAs) are responsible for spectrum sensing, we first propose a new spectrum-sensing frame structure to improve the data transmission duration of SU. We assume that even when the missed detection error of primary users (PU) signal occurs, PU and SU can share the same frequency band with a lower interference. We define the interference probability of each SU and derive the expected level based on the distance between the PU and the SU. We study the average achievable throughput with various signal-to-noise ratio (SINR) thresholds of PU, to investigate the relationship between k and the throughput through k-out-of-n rule. Our results show the performance of the proposed framework is significantly superior compared to the conventional CR and further reveal that reducing the SINR thresholds of PUs will achieve higher throughput performance.