Analysis of Distributed Noise Canceling

This paper tackles the issues which arise when the process of porting conventional algorithms, especially signal processing ones, from their normal centralized form to a parallel distributed form is undertaken. It deals specifically with those all too familiar issues particular to wireless sensor networks of limited power, memory, computational capabilities, and bandwidth. What is the proper approach one should take to ensure that goals of efficiency and performance are met, and to what degree of granularity should the parallelism available within an algorithm be exploited are some of the questions which will be addressed here. The issue of modeling the communication and computation scheme for the distributed version of the algorithm is introduced here along with the preliminary prototype for doing so