Collaborative routing and channel access in heterogeneous wireless networks

We consider the fundamental questions on how channel access interacts with the routing protocols in the wireless networks and propose corresponding cross-layer design solutions.Traditionally most wireless routing protocols work independently of the channel access method, even though it is not true that routing in wireless networks occurs over a pre-existing network topology and the transmission over one link does not impact the transmissions over other links, as it can be done in a wired network. A comprehensive understanding on the essentials of protocol interaction is indispensable to the design and optimization of protocol stacks, tailored to the unique dynamic environments in wireless networks.In this thesis, we first present an approach that takes advantage of multi-packet reception (MPR) to reduce the negative effects of multiple access interference and therefore increase the capacity of an wireless network. We analyze the performance upper bound of joint routing and scheduling for wireless networks that embrace interference by using MPR. We formulate the optimization problem under a deterministic model and seek to maximize the aggregate network throughput subject to minimum rate requirements. We then propose a polynomial-time heuristic algorithm aimed at approximating the optimal solution to the joint routing and channel access problem under MPR. We show the effectiveness of our heuristic algorithm by comparing its performance with the upper bound.Second, we propose a novel analytical model that captures the functionality of the routing protocols together with the characterization of the performance of the medium access control protocol (MAC). It reveals the interplay between the protocol functionality and network parameters, and provides new insight on the routing and channel access protocol design for wireless networks.Third, we present the CROWN (Collaborative ROuting, scheduling and frequency assignment for Wireless Networks) scheme. CROWN is a cross-layer optimization approach for spectrum-agile nodes to adjust their spectrum allocation and transmission scheduling according to the underlying traffic demands. Instead of choosing the optimal route based on predetermined transmission scheduling and frequency assignment results, CROWN incorporates the efficiency of the underlying frequency assignment and scheduling information into the routing metric calculation, so that the route with the maximal joint spatial and frequency reuse is selected. Simulation results show that CROWN efficiently exploits the frequency diversity and spatial reuse features of spectrum-agile radios.