Availability Based Multiservice Delivery for Wireless Sensor Networks in Adversarial Environments

Multiservice delivery is an important research issue in wireless sensor networks (WSNs) integrated with cloud computing, especially for heterogeneous applications which require reliable, timely and fair data delivery in adversarial environments. Therefore, both reliability requirements and delay constraints should be taken into account. Moreover, the delivery protocol should be designed to cater for multiple services regarding different traffic types. In this paper, we consider the problem of reliable multiservice delivery in which the source node performs rate control and routing selection based on availability statistics at individual sensors. We model the problem using network utility maximization and incorporate availability estimates into the NUM formulation (ANUM) with delay constraints. In order to handle multiple types of service, the performance of these services is modeled as a utility function of received rate in ANUM. Then we propose a Rate, Routing, and Delay Control (RRDC) algorithm to carry out multiservice delivery in an efficient manner, which is suitable to WSNs in respect of effective utility, latency, and utility fairness. To quantitatively evaluate the impact of unreliable sensors, we define two performance indices, utility degradation index (UDI) and utility fairness index (UFI). Simulation results show that the proposed RRDC algorithm achieves the desired performance among multiple types of service over WSNs in adversarial environments.