| A novel paradigm of communication, multi-hop wireless networks, have recently emerged both as a promising and cost-effective architecture to meet the ever-growing demands and expectations of the users. In these class of networks, a collection of wireless nodes dynamically establish and maintain connectivity among themselves, thus, enabling users and nodes to seamlessly internetwork in areas with a little or no communication infrastructure. In spite of the multiple aspects of advantages, however, research efforts have shown that when nodes are randomly or arbitrarily placed in a 2-D region, the capacity available to each source-destination pair becomes vanishingly small, as number of nodes grows to a large level. Motivated by such an issue, in this dissertation we are concerned with the problem of optimizing the capacity of multi-hop wireless networks. First, we propose to use a combination of cooperative communications and multiple channels, which together has great potential to evade various issues that limits the capacity of wireless networks. Further, using the insights of the proposed approach, we design a channel allocation protocol at the MAC layer for wireless networks employing cooperative communications. We also construct an analytical model to optimize the parameters used in the MAC protocol design. Second, we study the performance improvement in a multi-hop wireless network by coupling it with the coverage and capacity of infrastructure networks, referred to as hybrid wireless networks. In doing so, we point out severe flaws in the existing research efforts and design a simple and practical power-aware routing policy, that can adapt to the operating environment, for hybrid wireless networks. In comparison to existing works, we clearly show the gain one could obtain on delay as well as on capacity in executing our design. Lastly, we propose to use transmission power of nodes to increase the amount of information sent across each wireless link. While prior solutions rely on minimum transmission power to improve spatial reuse or lifetime of nodes, we look at the power problem from a different perspective and show that one can obtain a significant gain in capacity by judiciously enhancing the power in a multi-channel multi-hop wireless network. To prove this interesting result, we essentially introduce the novel concept of a co-channel enlarging effect and then quantify the maximum power at which nodes can communicate on a given channel, without causing harmful interference to other simultaneously communicating pairs. We conclude this dissertation by identifying open issues that need further investigation. |
