Implementation and performance evaluation of an experimental wireless relay sensor network

We investigate the performance of an experimental wireless relay sensor network comprising a source node, a destination node and dedicated relay nodes. Two relaying algorithms, dynamic relaying and fixed relaying, are tested to find out to what extent the relaying algorithms could bring any benefit to a wireless sensor network. A simple scheme is applied at transmit nodes to eliminate the need for synchronization. The tests are carried out in an indoor environment with rare movement of objects. The results show that the network performance can be significantly improved by the presence of relay nodes if data are not protected by a repetition code. However, if repetition coding is deployed the performance difference between the networks with and without relaying is very small. Moreover, it is shown that a single relay node using the fixed relaying algorithm can offer the best trade-off between performance improvement and energy consumption.