The Upper Bounds Reasearch On Qos In The Cluster-Tree Wireless Sensor Networks Based On Network Calculus

With the growing number of Wireless Sensor Network (WSN) applications, it’s of crucial importance to supply corresponding control of Quality of Service (QoS) under diverse circumstances. Based on the analysis framework of Network Calculus, this paper has analyzed and researched the upper bounds of QoS in the worst case of data aggregation in cluster-tree WSN applications. The duty circles of router nodes are set according to their depth of the cluster-tree structure. Moreover, this paper introduces a sleep time period during the scheduling cycle of Time Division Cluster Scheduling strategy to adjust the duty circles of the clusters at different depth for better control of energy consumption of the whole WSN. Under the circumstance of IEEE802.15.4/ZigBee specifications, this paper analyzes how the sleep time period will impact the guaranteed service of wireless sensor nodes. The upper bounds of single hop delay, backlog and end-to-end delay are derived and analyzed with numerical results based on deterministic network calculus afterwards.For improving the QoS under the worst circumstance, this paper applies the Strict Priority Scheduling Strategy in the cluster-tree WSNs After analyzing the how the SP scheduling would impact the guaranteed service provided to data flows, this paper derives the relevant arrival curves and service curves of different data flows based on network calculus. Moreover, the upper bounds of delay and backlog are derived and analyzed with numerical results and the result shows that the SP scheduling has positive impact on the QoS of WSNs.