Optimal flow splitting for multi-path multi-interface wireless data streaming networks

We propose optimal flow splitting for wireless data streaming networks consisting of nodes that are equipped with more than one wireless communication interface. The objective is to efficiently distribute data traffic among multiple wireless interfaces, leveraging an additional axis of freedom capable of providing improved performance. For that purpose a distributed framework based on convex optimization is developed that achieves optimal resource utilization for multi-interface multi-path network configuration. Although flow splitting can be performed using single wireless interface, our performance evaluation results show an improved performance when employing multiple wireless interfaces. Performance evaluation results reveal that 50% energy saving can be achieved when using optimal flow splitting for dual interface compared to single interface node architecture. In addition, the proposed solution is inherently robust to link outages due to the availability of multiple wireless interfaces. The proposed framework is quite flexible and can be easily extended to integrate any other performance parameters of interest.