Collaborative Resource Allocation in Wireless Sensor Networks.

Collaborative Resource Allocation in Wireless Sensor Networks   The new millennium heralds the convergence of communication, computing and intelligent control of the physical environment. The rapid advancement of computing and wireless technologies will enable us to employ cooperative real-time nodes in hostile environments in order to accomplish different tasks ranging from space monitoring and surveillance, to environmental protection without human intervention. Under this challenging vision, there will be an extensive deployment of highly dynamic and physically constrained real-time nodes connected together. Let us consider the following examples: Cooperative mobile robots, equipped with visual sensing, used in hostile/dangerous environments to clean up highly polluted spots or to remove mines or to defuse bombs. Cooperative real-time nodes, equipped with acoustic and visual sensing, used for surveillance in wide open spaces. Network of multifunction phase array radars (this is an example of real-time systems with physical constraints) used for air-traffic control or for military purpose such as detecting/tracking hostile targets.   It’s worth noting that all the applications mentioned above are characterized by a high degree of fluctuation in terms of computational and/or networking resource requirements. However, the causes of such a dynamic behavior are different in fact; for example, visual tracking is the main cause of highly variable workload in cooperative robots equipped with visual sensing, while, on the other hand, variable number of tracked targets and state dependent tasks cause highly dynamic workload in radar systems.   When several real-time nodes are connected together, the need for collaboration in a timely manner creates the following challenging problems:   Handling highly dynamic workloads among collaborative nodes. Providing real-times wireless communication.    Under the three major problems above identified, this thesis will focus primarily on issues like collaborative scheduling and prioritized medium access protocols. Specifically, the guidelines of this work are the followings: Collaborative Scheduling: tasks running on different nodes can be tightly coupled in a system where several real-time nodes cooperate. The goal is to develop distributed rate adaptation and collaborative resource reclaiming techniques aimed at mitigating the effects of highly dynamic workloads in distributed real-time system composed of collaborative nodes. It is worth noting that the degradation of performance of one task might affect the performance of other tasks running on different nodes (bottleneck task problem due to local rate adaptation), or locally reclaimed resources could increase a task rate without improving the overall system performance. Prioritized Medium Access with rate adaptive messages: traditional medium access control (MAC) are not suitable to build wireless sensor networks of collaborative real-time nodes because messages exchanged inside the network are mainly periodic and need guaranteed bounded delay. As a consequence, we will try to address the following medium access issues: Prioritizing the medium access to provide messages with bounded delay, and Providing rate adaptive messages in order to achieve the concept of distributed rate adaptation.