Design and Analysis of Distributed Mobility Management Schemes for Wireless Sensor Networks

In the present work, we investigate how to continuously inter-operate a large number of mobile nodes in an efficient way in large scale Wireless Sensor Networks (WSN). Mobility traditionally leads to a deterioration of the link quality, to frequent route changes and connectivity loss. In order to address the different challenges encountered in a scenario with strong mobility, we need a context-aware decision module adapting automatically and dynamically to its environment thanks to the combination of a multitude of metrics. Accordingly, we propose a group-weighted fuzzy-based decision module integrated within IEEE 802.15.4 standard scope. This mobility and multicontext management module adapts in real-time the intra-cluster communications depending on the detected local and temporal context patterns (i.e. in terms of nodes activity, mobility, energy autonomy, traffic, etc.) and triggers a handover mechanism ensuring the stability of the whole network. We used WSNet, an event-driven packet-oriented network simulator, to validate and compare the performance of the proposed solution with that of conventional approaches. Results show the importance of managing mobility on WSN and indicate that our approach achieves a higher packet delivery ratio while maintaining a reasonable endto- end delay. It can also reduce the number of disconnections and thus improves the overall network performance.