Project Details
Project description 

The Internet consists of an increasing number of small devices, connected in most cases via a wireless network to the Internet backbone. Important use cases in this respect are smart cities, inter-vehicle communication and smart transport systems, smart surveillance and remote participation applications. The networks for these use cases are distributed real-time systems consisting of (i) nodes that run on a very limited power budget, perform local processing and need to auto-configure, self-diagnose and self-adapt and (ii) wired networks interconnecting the clusters of wireless nodes. An important enabling paradigm for the mentioned futuristic scenarios is Multiterminal (MT) Source-Channel Coding, which addresses the compression of distributed sources of information. The goal of our research is to design novel MT coding schemes, which offer lightweight encoding, better compression performance, and improved error resilience over state-of-the-art systems. Since the networks are self-organizing, an intelligent cooperation is required between the distributed encoders, and the network management mechanisms both in the wireless networks and the wired networks. Algorithms will be designed for managing software-defined networks and fast allocation of virtualized resources for realizing the above described scenarios. Finally, we will investigate the application of the proposed approaches in cooperative distributed systems for networks employing fixed and mobile visual sensors.

Runtime: 2015 - 2018