Designing GA based effective transmission strategies for intra-WBAN communication

Abstract

Wireless Body Area Networks enable a new trend of proactive health care exploiting battery-powered wearable and/or implantable sensor nodes. The transmission of biomedical signals consumes most of the scarce energy resources in a WBAN. State-of-the-art transmission mechanisms either focus on energy efficiency or timely data delivery. However, an optimal balance between the two subject to various ambient conditions is a necessity. Accordingly, in this paper, an optimal data transmission policy is formulated at the node level, considering node lifetime, node heating, and reliable data delivery for various ambient conditions. The proposed model is divided into two phases. In the pre-deployment phase, each node is stochastically modeled following the Markov Decision Process. The formulation involves a reward function that attempts a trade-off between energy consumption and reliable data delivery. Then, a genetic algorithm is applied to obtain a set of optimal transmission decisions for each node based on the reward function for a range of ambient conditions. Finally, in the deployment phase, the results of the previous phase are mapped to an algorithm to be executed at WBAN nodes. The work is experimentally evaluated via simulation and compared with the state-of-the-art approaches. The performance is also observed based on the real-life dataset. Results demonstrate that around 85–90% of data packets could be delivered to the sink with minimal energy so that the nodes hardly get heated.