Towards Efficient Spectrum Sensing for Cognitive Radio through Knowledge-Based Reasoning

In this paper, an efficient approach to Medium Access Control layer spectrum sensing algorithm is introduced for the purpose of peer-to-peer communications with cognitive radio. The proposed algorithm is designed specifically for the application scenario of information gathering in a full-loaded legacy voice and data band, where the cognitive radio tries to utilize other available licensed bands through spectrum sensing. Efficiency is achieved in the proposed algorithm through the use of a knowledge-based reasoning approach to fine sensing, where the optimal range of channels to finely sense is determined using proactive fast sensing and channel quality information as priors. Simulation results demonstrate that the proposed spectrum sensing algorithm is capable of achieving improved performance over existing techniques by drawing a good balance between performance and operational complexity.