Diversity analysis for two-way multi-relay networks with stochastic energy harvesting

The pairwise error probability (PEP) performance is investigated for the energy harvesting (EH) two-way (TW) multi-relay network with stochastic EH models. In this network, two relay nodes exploit the amply-and-forward (AF) protocol and space-time network coding (STNC) to help two source nodes exchange information simultaneously. We assume only one relay node is solar-powered and equipped with a finite-capacity battery. The optimal transmission policy for the EH relay is proposed to minimize the long-term PEP by considering the causal EH information, battery energy and stochastic channel status. Thus, the design framework is formulated as a Markov decision process (MDP). We reveal that the full diversity order is achievable only if the EH relay´s battery empty probability is equal to zero. Further, the PEP performance can be improved when the battery empty probability becomes smaller. Simulations validate the theoretical analysis and show that the proposed optimal policy outperforms other myopic policies.