An Adaptive Frequency Hopping Scheme for Cognitive Radio Networks using Reinforcement Learning
Y R Meena, Kalyan Acharjya, Ramakant Upadhyay · 2024
Adaptive Frequency Hopping (AFH) is an essential generation for Cognitive Radio Networks (CRNs). It no longer bests improves spectral efficiency and quality of service of CRNs but also prevents interference amongst customers. However, traditional AFH schemes need help with essential problems. First, they require a large amount of previous expertise approximately the surroundings. 2d, they remain highly static and inflexible, situation to adjustments in the external environment. This paper presents an Adaptive AFH scheme using Reinforcement to gain knowledge of (RL) for CRNs. First off, an environment-conscious AFH approach based on the user-described frequency occupancy possibility is proposed. The method is composed of a probabilistic hopping sequence generator and an intra-frequency hopping detector. Secondly, an adaptive AFH scheme with temporal-spatial getting to know is proposed, which incorporates an actor-critic-based RL controller and recreation-theoretic modeled surroundings. The RL controller generates a dynamic hopping sequence with low hopping frequency collisions at the same time as the sport-theoretic surroundings fashion the temporal-spatial modifications of the wireless medium. Sooner or later, the proposed scheme is proven in a simulated CRN environment.