Lossy data aggregation with network coding in stand-alone wireless sensor networks
NEW2AN'11/ruSMART'11 Proceedings of the 11th international conference and 4th international conference on Smart spaces and next generation wired/wireless networking
| Hi-index | 0.00 |
In wireless sensor networks, in-network data aggregation is an efficient way to reduce energy consumption in network. However, most of the existing data aggregation scheduling methods try to aggregate the data from all the nodes at all time-instances. It is neither energy efficient nor practical because of the link unreliability and spatial and temporal data correlations. In this paper, we propose anew data aggregation paradigm which allows estimated aggregation at the sink node. In our scheme, we will selectively let some nodes sample data and aggregate them to the sink node. Two different cases will be studied. Firstly, we assume that the links are reliable and the error between the value obtained from the data of all nodes and that from the data of sampled nodes is bounded. We give detailed analysis on the error bound when the confidence is given a priori. Secondly, we assume that the links are unreliable with a given probability and obtain that the error is still bounded under a given confidence when the probability of link unreliability is not too high or the success probability of retransmission is high enough. We also study how to assign the confidence level among the root nodes such that each root node can calculate the minimum number of sampling leaf nodes based on corresponding confidence level. Through analyzing, we show that it can surely save energy to adopt our method when the link is reliable. When the link is not reliable, the energy still can be saved if the success probability of retransmission is high enough.