Cooperative Networks: A Framework to Derive Scalable Cross-Layer Designs

Cornell University, Ithaca NY

Investigators

Abstract

Technical abstraction boundaries, such as the layers of the OSI model, have been instrumental in the evolution of high performance/power hungry information networks in the past. However, they are currently representing a bottleneck to the development of the large distributed networks such as networks of embedded systems, unmanned vehicles, robots and sensor networks. These networks could have tremendous impact on our society, making it easier to manage and safer. Deploying existing wireless and processor technology in these applications can be dangerous because of the cost of keeping them operative, because of the difficulties in managing them on large scales and because of their vulnerability of the systems to power outages. This research is motivated by the need of a solid scientific framework and a practical set of algorithms that surpass the limiting architectural models used in the present technology and provide working solutions for these new emerging applications. To address this problem this research uses the concept of cooperation at the physical layer as the guiding principle for removing some traditional design boundaries that limit networks scalability and also as a systematic methodology to develop cross-layer solutions. The main partition removed is the network abstraction of "point to point communications", where the information from every source is encoded and transmitted independently and it is received and processed independently. The cooperative networking reference model studied in this research changes all this by allowing the nodes to collaborate in compressing/elaborating the data and in transmitting them. Cooperation seems appropriate in networking sensors, unmanned robots, embedded systems since all the nodes of these networks naturally complement each other activities and information. The study addresses the cooperative modulation, channel acquisition and data compression problems. Level of effort To compress the projects tasks in one year of activity and encompass the main research objectives of our project we will dedicate our focused effort to solve the key issues of channel acquisition, link control and decentralized multiple access in the physical layer cooperative channel model. The aim is to build a new reference model for cooperative networks that clarifies the structure of the communication architecture. A more detailed description of the issues we are going to address is expressed in our original "Statement of Work" [c.f. Section 2.2 in the proposal] and specifically in the first three research tasks described therein. By addressing them we will set the solid framework for the development of fully cooperative and decentralized network architectures. Other research items will be left for future projects.