Rate-Aware Network Codes for Video Distortion Reduction in Point-To-Multipoint Networks

This paper considers a wireless point-To-multipoint network in which a base station needs to broadcast a real-Time video sequence to a set of devices with heterogeneous channel capacities. In such a scenario, a packet transmission is successfully received at a given device if the adopted transmission rate is lower than the channel capacity of that device. To reduce the video distortion of all devices before the deadline, this paper employs instantly decodable network coding (IDNC) and formulates the video distortion minimization problem as a Markov decision process. Given that the optimal policy suffers from a high computational complexity, an online maximal clique selection algorithm over a rate-Aware IDNC graph is proposed to heuristically select a transmission rate and a packet combination at each transmission. This heuristic reduces the individual video distortions of all devices by incorporating the unequal importance of video packets, the hard deadline, and the various channel capacities into the coding decisions. Furthermore, this heuristic is modified to propose a fairer solution that delivers a good quality video to individual devices regardless of their channel conditions. Simulation results over a real video sequence reveal that the proposed IDNC algorithms improve the received video quality as compared to existing rate-Aware IDNC algorithms.