Opportunistic network coding based cooperative retransmissions in D2D communications

Device-to-Device (D2D) communications can improve the performance of the next generation cellular networks by facilitating the transmission of cellular traffic without additional infrastructure. In D2D communications, cooperative retransmission can be employed to assist the recovery of lost packets at proximity devices by sending coded packets. However, most existing work did not consider the issue of how the transmissions among neighbor devices can be well scheduled to support efficient network coding based retransmissions. Decoupling transmission scheduling from coding decision making can lead to inefficient D2D packet retransmissions and thus reduced overall network performance. In this paper, we study how to use opportunistic network coding based transmission scheduling to assist the cooperative packet recovery in D2D environment. For this purpose, we first formulate the problem of jointly optimization of transmitting device selection and also their coding patterns as minimum latency scheduling problem, which is known to be NP-hard. We further present a low complexity centralized algorithm and a distributed mechanism that does not require synchronization (either global or local) among devices. We derive the complexity of the proposed algorithms. Simulation results show that the performance of centralized algorithm is very close to the theoretical lower bounds and the distributed mechanism can achieve high performance in accelerating packet recovery.