CAREER: Next Generation Network Coding: Distributed Design Via Coded Feedback

Purdue University, West Lafayette IN

Investigators

Abstract

The exponentially increasing demands for heterogenous services over Internet and wireless networks have imposed new urgency and great challenges for designing next-generation networks that achieve higher throughput, better resiliency to adversary attacks, and stronger security for sensitive information. Recently, a new network coding paradigm has emerged, which potentially achieves the above goals simultaneously in an optimal fashion. Nonetheless, to realize fully its promised benefits requires a clean-slate methodology of network code design that achieves optimal performance and maintains low complexity while reacting quickly to the ever changing network environment. To that end, the investigators study coded feedback, which generalizes the conventional feedback mechanism that acknowledges non-coded traffic by incorporating the unique algebraic structure of network-coding. This research first explores the joint algebraic properties between coded forward and feedback packets, which are the basis for the subsequent development of new, efficient network protocols for next generation network codes. In particular, the investigators are interested in the roles of coded feedback that transform the non-coding network resource allocation algorithms to their coding-oriented counterparts. The rich information carried by coded feedback is then used to design protocols that offer enhanced global and local error protection and data security while self-adapting to network changes. The theoretic studies of this program follow closely practical considerations including the dynamic channel quality and the interference nature of wireless networks and the complexity of the resulting protocols. System-level implementation is used to verify and further refine the theoretic findings of this program. The coordinated approaches of coding theory, graph theory, and network optimizations provide valuable interdisciplinary training for students with Purdue University.