Performance of cognitive radio reinforcement spectrum sharing using different weighting factors

This paper introduces a distributed spectrum sharing scheme in the context of cognitive radio which enables efficient usage of spectrum. This is achieved by using the past experience based on reinforcement learning. It shows that reinforcement spectrum sharing provides a good solution and has the potential to significantly improve the system performance. Several learning strategies based on different sets of weighting factors are investigated. Comparisons of system performance using different learning strategies are given to illustrate the importance of weighting factors in the spectrum sharing process.