Optimization of distributed free-viewpoint video synthesis

New type of video streaming may appear in the future networks, called Free-viewpoint Television (FTV) allowing customers to control their viewpoint and generate new views of a dynamic scene from any 3D position. The desired viewpoint is synthetized from two or more camera views requiring significant network and computational resources. We propose to distribute the viewpoint synthesis functionality in the network in so called proxy servers. In this paper we investigated how to localize the proxy servers in order to generate the lowest traffic load, but avoid the overload of the disposable resources. Finding the optimal layout is a NP-complete problem, so our goal was to analyze the achievable gain in a simple hierarchical network with several nodes, where the runtime of the brute force optimization algorithm is still acceptable. The obtained results paper showed that even 42% traffic decrease can be realized using distributed viewpoint synthesis for FTV services, so it is worth to continue our work and investigate new alternative optimization algorithms that are capable to find near optimal proxy server layouts and fast enough to adapt the proxy server arrangement continuously according to actual user requirements.