Collective Real-Time QoS in Wireless Sensor Networks

Wireless sensor networks (WSN) are composed by sensors which are often used for event detecting and target tracking. In the traditional, wire-based networks, applications often concern about data transport between two specific nodes. While in typical applications of WSN, the data sensed about event or target is reported by a group of sensors, which is quite different from the traditional end-to-end model. Thus, in WSN, the conventional definition of quality of service (QoS) is no longer applicable. Particularly, the metric of real-time service has to be advanced in order to reflect the data-centric, i.e. collective characteristic of WSN. To address the need, we firstly choose the appropriate metric of the real-time services. Moreover, based on the collective metric, a novel collective real-time service protocol (CRTS) is proposed to meet different collective real-time requirements of events simultaneously. To achieve our goal, different QoS classes are dynamically assigned to different events according to their actual and demanded collective real-time performance. In addition, a queuing model based on QoS class is employed to allocate the bandwidth of intermediate nodes among events. By using the strategy we want, every event acquires enough bandwidth to ensure collective real-time service while minimizing the bandwidth waste. The result of simulation verifies the effectiveness of our approach