Optimised Joint Channel Network Coding for Sensor Network with Correlated Sources

In this paper, we consider a Generalized Joint Channel Network Coding (GJCNC) scheme applied to Physical Layer Network Coding (PLNC) for a sensor network. For this purpose, M correlated sources desire to send information to one receiver by the help of a relay. The main goal is to estimate the information being sent from the last source at the destination by using the informations from sources and relay. For this purpose, we propose firstly a classic joint channel network coding at the relay which consider each source signal separately. Then, an optimization of this scheme is done by grouping by pair the nodes in the sensor network. The GJCNC decoding algorithm for each pair of sources is performed at the relay to improve performance. So, an iterative decoding scheme is proposed at the destination to extract the information sent from the last source. Simulation results show that by exploiting the correlation between sources, a significant gain can be achieved.