Inter-application Redundancy Elimination in Sensor Networks with Compiler-Assisted Scheduling
Ref: VG-CAS-11 Publication Date: 16, Jun, 2011
Inter-application Redundancy Elimination in Sensor Networks with Compiler-Assisted SchedulingRef: VG-CAS-11 Publication Date: 16, Jun, 2011
Wireless sensor network nodes supporting multi- tasking and multiple concurrent applications are becoming increasingly common. These nodes are typically equipped with multiple sensors of various types. This trend has been fostering the design of wireless sensor networks allowing several concur- rent users to deploy applications with dissimilar requirements. At the same time, the practical burden of programming individual sensor nodes has led researchers to design macro- programming schemes able to program the network as a whole. In this paper, we extend the advantages of a holistic program- ming scheme by designing a novel compiler-assisted scheduling approach (dubbed REIS) able to identify and eliminate redun- dancies across applications. To achieve this useful high-level optimization, we propose to model each user application as a linear sequence of executable instructions; we show how it is then possible to exploit well-known string-matching algorithms such as the Longest Common Subsequence (LCS) and the Shortest Common Super-sequence (SCS) to produce an optimal merged sequence of the multiple applications that takes into account embedded scheduling information. We demonstrate how this novel approach leads to significant network-wide resource savings, including energy.