Service oriented wireless sensor network.

摘要

Quality of service (QoS) applications and network lifetime management are key challenges in Wireless Sensor Networks (WSNs). Service-level guarantees that are achievable in Service Oriented Architecture (SOA) can ensure reliability, availability and overall usability in any application. How to create the necessary structure to implement SOA in a WSN to achieve the required levels of quality of service is still a research topic. The SOA paradigm provides the methodology to improve these qualities in WSNs. We present a strategy that uses sensors in a cluster and dynamic service selection to achieve higher levels in quality of service for WSNs, while preserving network lifetime. The realization of these strategies will be achieved through a SOA middleware layer in the WSN. The concept of SOA dynamic service selection is transformed so that it can be applied to WSNs. The transformation requires that we equate a service in SOA to functions that are performed in a WSN. In our architecture, a service or function will be achieved through a group of sensors called a sensor cluster. Quality metrics will be captured at the sensor cluster level each time it is executed within the WSN. We present sensor cluster selection algorithms that will determine the most effective and efficient sensor clusters to participate in a composite function plan. The sensor cluster selection algorithms guarantee a minimum level of quality by selecting services that meet a client's quality constraints while taking into account the network routing requirements to achieve the operation in the shortest amount of time utilizing the least amount of energy. Additionally, we present methods for implementing service selection using the multiple choice, multiple dimension knapsack (MMKP) algorithm within a composite function that has complex workflow. We make use of the WSN's MMKP selection algorithm using dynamic programming as a proof of concept to demonstrate that there is a significant improvement in the energy utilization, network lifetime and overall quality.