Intelligent Spectrum Sensing Using Optimized Machine Learning Algorithms for Cognitive Radio in 5G Communication

Abstract

Spectrum Sensing plays an important role in cognitive radio which is used to resolve the co-existence issue and to optimize the available spectrum efficiency. However, the upcoming 5G communication involves the diversified scenarios with distinguished characteristics which makes spectrum sensing more difficult to serve different application in terms of high performance and flexible implementation. Also motivated by this challenge, the paper proposes the new algorithm which implemented the novel bat optimized multi-layer extreme learning machine and works on different input vectors such as received signal strength, distance, Energy and channel ID and classifies the users for the better classification and sensing. Moreover, to prove the suitability of the proposed algorithm in terms of performance under 5G scenario for health care applications, we have compared the other spectrum sensing techniques and parameters such as sensitivity, selectivity and channel detection probability. Ultimately the results demonstrate that the proposed spectrum sensing has outperformed the other algorithms and shows its capability to adopt for various 5G scenario.