Control protocols for wireless mesh networks must carefully deal with the cross-interactions within the network in both the vertical (i.e., across protocol layers) and the horizontal direction (i.e., across network nodes). These cross-interactions offer both an opportunity to use cross-layer/network-wide optimization to achieve high capacity, and the challenge to maintain scalable, robust and modular solution architecture. The goal of this project is to develop a high-performance mesh-network control plane (MCP) that is loosely-coupled, i.e., only a minimal amount of information is exchanged between layers and network components to optimize network efficiency, and further, if changes, errors, or inefficiencies occur at one layer/component, the overall solution shows graceful degradation. Thus, this new control plane achieves both efficiency and modularity. The PIs will develop a solid theoretical framework for such a loosely-coupled MCP across multiple protocol layers, based on which robust, decentralized, and easy-to-implement algorithms and protocols will be developed to address challenges in congestion control, admission control, link scheduling, multi-access, power control and routing. The solution will be implemented and evaluated on the Mesh@Purdue testbed.
Broader Impact: The results will have a broad impact on wireless industry sectors by providing new fundamental insights, novel control structures, distributed algorithms, and practical protocol stacks. Through the Center for Wireless Systems and Applications (CWSA) at Purdue, the PIs will actively share their findings with industrial collaborators. Students supported on this project will have a balanced exposure to a wide variety of theoretical and applied techniques, which will be valuable in their future careers.
