In two-way relaying networks, two sources exchange messages with each other supported by a relay. Using a two-phase protocol, both sources transmit simultaneously, resulting in a superimposed signal at the relay. Based on the fact that the relay is not interested in the individual messages from the sources, a network coded signal is directly estimated from the received signal, which is broadcast back to the sources in another time slot. Such a processing strategy to map the received signal to the network coded signal is termed Denoise-and-Forward (DNF).

The goal of this work is to design the DNF mapping rule adapatively based on the instantaneous channel knowledges such that the overall system throughput is increased. To successfully accomplish this work, solid knowldges in CT I+II and Advanced Topics in Digital Communications are required. The analytical researches should be verified by MATLAB simulations.