Distributed Media Rate Allocation in Multipath Networks

The paper addresses the media-specific rate allocation problem in multipath networks. The streaming rate on each path is determined such that the end- to-end media distortion is minimized, when the receiving client aggregates packets received via multiple network channels. As it is difficult for the media server to have the full knowledge about the network status, we propose a distributed path selection and rate allocation algorithm. The network nodes participate to the optimization strategy, based on their local view of the network status. This eliminates the need for end-to-end network monitoring, and allows for the deployment of large scale rate allocation solutions. We design an optimal rate allocation algorithm, where the media client iteratively updates the best set of streaming paths. According to this rate allocation, each intermediate nodes then forwards incoming media flows on the outgoing paths, in a distributed manner. The proposed algorithm is shown to quickly converge to the optimal rate allocation solution, and hence to lead to stable rate allocation solutions. We also propose a greedy distributed algorithm that achieves close-to-optimal end-to-end distortion performance in a single pass. Both algorithms are shown to outperform simple heuristic- based rate allocation approaches for numerous random network topologies, and therefore offer an interesting solution for media-specific rate allocation over large scale multi-path networks.

Pascal Frossard, Dan Jurca

The paper addresses the media-specific rate allocation problem in multipath networks. The streaming rate on each path is determined such that the end-to-end media distortion is minimized, when the receiving client aggregates packets received via multiple network channels. As it is difficult for the media server to have the full knowledge about the network status, we propose a distributed path selection and rate allocation algorithm. The network nodes participate to the optimization strategy, based on their local view of the network status. This eliminates the need for end-to-end network monitoring, and allows for the deployment of large scale rate allocation solutions. We design an optimal rate allocation algorithm, where the media client iteratively updates the best set of streaming paths. According to this rate allocation, each intermediate nodes then forwards incoming media flows on the outgoing paths, in a distributed manner. The proposed algorithm is shown to quickly converge to the optimal rate allocation solution, and hence to lead to stable rate allocation solutions. We also propose a greedy distributed algorithm that achieves close-to-optimal end-to-end distortion performance in a single pass. Both algorithms are shown to outperform simple heuristic-based rate allocation approaches for numerous random network topologies, and therefore offer an interesting solution for media-specific rate allocation over large scale multi-path networks.

2006

Media-Specific Rate Allocation in Multipath Networks

Pascal Frossard, Dan Jurca

We address the problem of joint path selection and rate allocation in multipath streaming in order to optimize a media specific quality of service. An optimization problem is proposed, which aims at minimizing a video distortion metric based on sequence-dependent parameters, and transmission channel characteristics, for a given network infrastructure. Even if in general, optimal path selection and rate allocation is an NP complete problem, an in-depth analysis of the media distortion evolution allows to define a low complexity algorithm for an optimal streaming strategy. In particular, we show that a greedy allocation of rate along paths with increasing error probability leads to an optimal solution. We argue that a network path shall not be chosen for transmission, unless all other available paths with lower error probability have been chosen. Moreover, the chosen paths should be used at their maximum available end-to-end bandwidth. Simulation results show that the optimal rate allocation carefully trades off total encoding/transmission rate, with the end-to-end transmission error probability and the number of chosen paths. In many cases, the optimal rate allocation provides more than 20% improvement in received video quality, compared to heuristic-based algorithms. This motivates its use in multipath networks, where it optimizes media specific quality of service, and simultaneously saves network resources, with very low computational complexity.

2005

Packet Selection and Scheduling for Multipath Video Streaming

Pascal Frossard, Dan Jurca

This paper addresses the problem of choosing the best streaming policy for distortion optimal multipath video delivery, under delay constraints. The streaming policy consists in a joint selection of the video packets to be transmitted, as well as their sending time, and the transmission path. A simple streaming model is introduced, which takes into account the video packet importance, and the dependencies among packets, and allows to compute the quality perceived by the receiver, as a function of the streaming policy. We derive an optimization problem based on the video abstraction model, under the assumption that the server knows, or can predict the state of the network. A detailed analysis of the timing constraints in multipath video streaming provides helpful insights that lead to an efficient algorithm to solve the NP-hard streaming policy optimization problem. We eventually propose a fast heuristic-based algorithm, that still provides close to optimal performance. Thanks to its limited complexity, this novel algorithm is finally demonstrated in live streaming scenarios, where it only induces a negligible distortion penalty compared to an optimal strategy. Simulation results finally show that the proposed scheduling solutions perform better than common scheduling algorithms, and represent very efficient multipath streaming strategies for both stored and live video services.

2004