It is well documented how cooperative links can offer considerable performance gains at the physical layer, but it is unclear what kind of network support would be required to attain the sought gains. Cooperative links violate the simple collision model for broadcast transmission, a model that has been instrumental so far in allowing the parallel evolution of communication theory and network theory. Recognizing the absence of a correct taxonomy to use cooperative links at the network and multiple access layer, the objective of this collaborative project is to investigate theoretically and experimentally the interplay between a cooperative decentralized physical layer and the wireless network architecture as a whole. More specifically, the project will develop viable link abstractions, multiple access protocols, end-to-end network transport models, appropriate algorithms to support the introduction in wireless mobile networks of two technologies that are rapidly advancing: 1) cooperative transmission, that consists of multiple network nodes operating as a decentralized multi-antenna modem and, 2) distributed source coding, that allows the decentralized compression of correlated observations and, thus, is relevant to the design of a decentralized receiver. The project will also use the GNU software radio platform to test cooperative links and assess their feasibility and degradation when facing real limitations of transceiver synchronization, carrier offset, clock jitter and computation delays. Algorithms and theoretical results will be disseminated through the standard tools of research publications. Experimental results will be also documented online where the software will be shared to serve as an educational tool as well as to foster new technological advances in mesh networks. This project will bring future wireless networks closer to achieving the physical limits of communications.
