Chained estimation for data reduction-driven routing in wireless sensor networks

Wireless sensor networks are a precious tool for numerous use cases: studying the fauna and flora, measuring various metrics and so much more. Here we focus specifically on wireless multihop networks, in which the funneling effect is too often responsible of the unbalance of energy consumption, causing in turn nodes failures. In an attempt to mitigate the effects of funneling, the scientific literature proposes among others traffic reduction based on data estimation. The lead idea being that if two phenomena are correlated, then maybe we could estimate the way one goes based on the way the second one do. This effectively reduces the amount of data nodes may have to transmit while preserving the information. We propose in this article a similar method, but here not only limited to a pair of nodes. With our scheme we can reduce the traffic of several nodes over a whole multihop route of communications in a network. We have implemented that scheme and experimented with real hardware which shows good results.