An analytic traffic model with adaptive QoS control in an unreliable wireless sensor network

An analytic traffic model is developed for an unreliable wireless sensor network. We model the dynamics of traffic flow from the source node through a set of intermediate nodes to the sink node by using single-server queues. These single-server queues with finite buffers are linked in tandem. To analyze the performance of the sensor network, we decompose the tandem queuing network into individual nodes with modified arrival and service processes and modified queue capacities. In the individual node modeling, we consider the impact of the unreliable transmissions, i.e., node/link failure events, by involving the immediate upstream node and downstream node of the separated node. The steady-state solutions of the individual nodes are determined iteratively. A performance metric source-to-sink delay is derived and selected for studying the quality of service (QoS) control. Adaptive QoS control schemes are developed and their performance is validated by simulation.