Cognitive MAC protocol with minimum sampling time and cross-layer cooperation for low signal-to-noise ratio environments

摘要

The previous Decentralised Cognitive Medium Access Control (DC-MAC) protocol allows Secondary Users (SUs) to independently search for spectrum access opportunities without the need for a central coordinator. DC-MAC assumes that the detection scheme is ideal at the Physical (PHY) layer. In fact, a more complex detection algorithm is impractical in distributed spectrum sharing scenarios. Energy Detection (ED) at the PHY layer has become the most common method because of its low computational and implementation complexities. Thus, it is essential to integrate the DC-MAC with ED at the PHY layer. However, ED requires the Minimum Sampling Time (MST) duration to achieve the target detection probability in low Signal-to-Noise Ratio (SNR) environments. Otherwise, it cannot achieve the expected detection performance. In this paper, we derive an accurate expression of MST for ED in low SNR environments. Then, we propose an Optimised DC-MAC (ODC-MAC) protocol which is based on MST, and which amends the aforementioned problems of DC-MAC with ED. Moreover, the closed-form expressions for the unreliable data transmission probability are derived for both DC-MAC and ODC-MAC. We show that the simulation results agree well with the theoretical analyses. The proposed ODC-MAC can improve the data transmission reliability and enhance the throughput compared to the performance of the traditional DC-MAC.