A Push-based Paradigm for Environment Adaptive Application Reconfiguration in Clustered Sensor Networks

Abstract

Application reconfiguration is essential in order to complement the flexibility and adaptability for sensor networks in the environment monitoring domain. In this paper, we first describe an effective environment adaptive application reconfiguration (EAAR) mechanism. Then we present a push-based paradigm to carry out EAAR, where instead of a pull-based paradigm proposed in our prior work. This paradigm suits for clustered sensor networks, which provides a software entity, reconfiguration agent, to perform the process of reconfiguration on cluster head nodes. Reconfiguration agent is a novel concept, and has great potential in providing energy-efficient reconfiguration processing with low latency. A prototype is constructed to show the workflow of reconfiguration agent. Particularly, we design two metrics: execution time and energy consumption, and use the simulation platform to quantitatively measure the performance of different paradigms in the reconfiguration procedure. Experimental results show that push-based paradigm performs better than pull-based one