6. Conclusion
As more multimedia contents are consumed on mobile devices, streaming traffic over wireless networks have been growing exponentially. However, wireless networks often suffer unpredictable failures such as high packet-loss ratio and transmission path disconnection. Multi-hop cooperative relay networking is a promising solution proposed to transmit video streams more stably over wireless networks. Although many studies have been conducted on how to improve the performance of video streaming over multi-hop cooperative relay networks, they have largely focused on fundamental approaches such as cross-layer design, and relay network scheduling. In this paper, a novel approach for video flow routing that takes video complexity into account was presented. A proposed video distortion model was developed for heterogeneous motion-level video streams, because video complexity is one of the main characteristics of video sequences. The proposed flow routing algorithms, CRVC-D and CRVC-I, are comprised of three phases: path determination, CR assignment, and rate assignment. The proposed algorithm considers the video complexities of heterogeneous motion-level video sequences during path determination and rate assignment. The simulation results show that providing high priorities to fast- or slow-motion videos for selection of an available routing path in the different communication conditions improves the performance in terms of min PSNR and PSNR variance without impairment of the total PSNR. It was also shown that the support of CRs and an appropriate hop-count limit improves performance.
