Global Optimisation Based Stochastic Selfish KL-UCB for Spectrum Sensing Applications
Aiming at the problem of spectrum resource waste in radio communication,the paper proposes the Globally optimised stochastic selfish KL-UCB(Globally optimized stochastic self-ish Kullback-Leibler Upper Confidence Bound)algorithm.Its goal is to improve the spectrum utilisation and system performance through optimal spectrum selection decisions when terminal devices compete for spectrum resources.In the paper,according to the random selfish KL-UCB characteristics,the algorithm is utilised in the spectrum sensing problem,by initialising,iterating and updating the spectrum hole generation,the spectrum hole generation and the cumulative re-gret value are used as the judging criteria,and the global optimisation of the random selfish KL-UCB algorithm improves the optimisation rate of the global optimisation algorithm by 45%o-ver the random selfish KL-UCB algorithm through the comparison of the data,and it is conclu-ded that the improved random Selfish KL-UCB has a more efficient and holistic has a wide range of application prospects in spectrum sensing.