Multi-Class Multi-Load Handling MAC Protocol for WBAN Based on IEEE 802.15.6 Standard Using Reinforcement Learning

The capacity to exchange data and provide real-time monitoring for patient care makes wireless body area networks (WBANs) vital for modern medical treatment systems. WBANs are used for various traffic loads, including high-rate and extremely low-rate continuous monitoring data. The communication standard of IEEE-802.15.6 describes physical and MAC layer functions for WBANs. The IEEE 802.15.6 superframe is used to handle multiple types of traffic load in WBAN based on predefined user priorities. Since various kinds of data traffic have fixed periods, this results in an ineffective use of the superframe. In this study, we develop a multi-class multi-load handling MAC (MMH-MAC) protocol for WBANs. This protocol maximizes the use of the IEEE 802.15.6 superframe through reinforcement learning (i.e., Q-learning). The proposed MAC protocol classifies the data, calculates the node’s Q-value, and allocates the time slot for different loads based on the largest Q-value using the TDMA technique. The operation of the MMH-MAC protocol is examined via comprehensive NS-2 simulations and thoroughly contrasted with the present protocols.