On Reconfiguring Embedded Application Placement on Smart Sensing and Actuating Environments

Abstract

In a smart home environment appliances and objects have sensing, actuating and wireless communication capabilities. Recent embedded middleware initiatives aim at providing an easy to use programming framework for developing applications that require the cooperation of sensing and actuating nodes. To achieve this, the mobile agent paradigm is adopted under which an application consists of a set of communication agents residing at nodes of the system with adequate resources. Of particular importance in such system is to decide which agent to place where so that network traffic is optimized. In this chapter we discuss the problem of reconfiguring an initial agent placement so that the resulting scheme is more energy efficient. We formulate the aforementioned agent placement problem (APR) which turns out to be particularly challenging due to the various constraints involved. A heuristic algorithm based on graph coloring is proposed and evaluated against a greedy approach under various scenarios.