Efficient Data Gathering for Decentralized Diversity Combining in Heterogeneous Sensor Networks

We study reliable communication in heterogeneous sensor networks. In many application scenarios, more energy-constrained (mobile) nodes are distinguished from more powerful base stations (or ground nodes). Wildlife monitoring is just one of many examples within the Internet of Things research community. In order to improve the communication reliability (and, thus, also the energy footprint), these ground nodes often apply macro-diversity to reduce transmission failures and to avoid costly retransmissions. In recent years, the concept of using distributed sensor networks as antenna arrays for receive diversity has been proposed. We address one of the key challenges in such networks, which is the huge cost of forwarding signal samples to a (central) sink through the ground network, where diversity algorithms are eventually applied. In particular, we present two algorithms, a cluster and a tree based one, that help reducing the data transfers in the ground network. Ground nodes try applying diversity techniques early whenever they receive signal samples from multiple receivers. In extensive simulations, we show that the algorithms substantially outperform naïve centralized solutions.