In this paper, we propose a resource allocation algorithm to transmit multiple video bitstreams over multiuser orthogonal frequency division multiplex (OFDM) networks. The proposed algorithm explores a low-complexity rate-distortion function and diversity of multiuser OFDM systems subject to constraints on delay and limited system resources. A set of relative importance is imposed in resource allocation to assure that each user can achieve the required data rate and the quality of service. We first present a simple and effective rate-distortion function including a temporal error propagation effect and relative importance. Then we propose the cross-layer multiuser resource allocation algorithm that consists of subcarrier assignment and power allocation for minimizing the overall video distortion. The proposed subcarrier assignment algorithm utilizes the subgradient method adopting a shadow price mechanism by assuming equal power distribution for each user. Based on the result of subcarrier assignment, the power allocation algorithm maximizes the sum of distortion reduction while the relative importance for each user is maintained. In simulation results, the proposed resource allocation algorithm outperforms the time division multiple access (TDMA) and the previous resource allocation algorithm maximizing the sum capacity in terms of the worst received video quality among users. The proposed algorithm achieves PSNR gain by 1.0 s(-) 3.0 dB, compared to the previous resource allocation algorithm considering rate-distortion function. It is shown that the proposed resource allocation algorithm distributes the video distortion among users more fairly with relative importance compared to the previous algorithms.
Bibliographical noteFunding Information:
This research was supported by the MKE (Ministry of Knowledge Economy, Korea, under the ITRC (Information Technology Research Center) support program supervised by the IITA (Institute of Information Technology Assessment) (IITA-2008-C1090-0801-0038). The research of C. Yim was supported by the Seoul R&BD program (10581).
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications