Dynamic Packet Size Optimization and Channel Selection for Cognitive Radio Sensor Networks

Dynamic spectrum access has gained traction in wireless sensor networks (WSNs) because of the scarcity in spectrum caused by the proliferation of wireless devices and services, and it provides spectrum efficient communication for the WSNs. However, the communication between nodes in a cognitive radio sensor network (CRSN) is affected by the transmission power, fading, and interference with licensed users, and these factors hinder the data transmission between the energy constrained cognitive radio sensor nodes. Therefore, there is a need for an adaptive energy-efficient optimization scheme which takes into account the varying environment conditions. Since packet length plays a pivotal role in determining the performance of the network, packet size adaptation that is aware of the channel characteristics may bring about performance improvement. Furthermore, existing packet size optimization or channel selection schemes devised for WSNs and CR networks are not appropriate for the CRSN framework. In this paper, we devise a dynamic packet size optimization and channel selection scheme (DyPSOCS) for CRSNs. We employ a constrained Markov decision process (CMDP) to solve the optimization problem with quality of service (QoS) constraints. Simulation results show improvement in QoS performance as well as energy efficiency when compared to other schemes.