An intelligent closed-loop learning automaton for real-time congestion control in wireless body area networks

Recently, a considerable literature has grown up around the theme of wireless sensor networks (WSNs) that are investigated in many applications. So far, wireless body area networks (WBANs) have emerged as a flexible solution for remote monitoring of mobile patients, nurses and elderly people. Nevertheless, WBANs are challenged by the real-time constraints of medical data. In another hand, WBANs have limited communication and computation capabilities. So, the collected physiological data should be aggregated before being sent to the base station. Moreover, the aggregation process may lead to the congestion problem. In this paper, we propose a closed-loop learning automation that is based on conditional probabilities to assign each packet to the appropriate queue given the criticality of the sensed data. Our OMNeT++ simulation results have been analysed using the Z-test. They confirm the performance of our scheme in terms of the drop ratio and the throughput.