| Broadband wireless networks are likely to support a large volume of Internet traffic that are currently only viable over wireline networks. Unlike wireline networks, the wireless spectrum is a distributed shared resource whose availability is limited. The rate of a wireless link depends largely on its signal-to-interference-plus-noise ratio and therefore, is highly coupled with the rate of other active links. Provisioning high rates along wireless links over long time horizons necessitates a careful scheduling of links over time and suitable adjustment of their signal powers. Minimizing energy consumption is of paramount importance for the successful development of broadband wireless networks, since it is crucial to supporting high rates. Using multiple routes between origin-destination pairs to transport traffic can be exploited to greatly increase rates.; In this dissertation, we study the problem of joint routing, link scheduling and power control to support high rates for broadband wireless multi-hop networks. We first address the problem of finding an optimal link scheduling and power control policy that minimizes the total average transmission power in the wireless multi-hop network subject to a minimum average rate constraint per link. Multiaccess interference is explicitly modeled. As a byproduct of finding the optimal policy, we find a subgradient of the minimal total power with respect to the rate for each link. We use this subgradient to guide us towards the optimal flow allocation along routes. We discuss some possible generalizations of our approach and outline a low complexity solution to determine an efficient transmission schedule as the rate requirements along links vary.; To scale with network size, we propose a hierarchical approach to link scheduling and power control. In this approach, links are partitioned into groups called clusters. Within a cluster, optimal algorithms are employed to minimize energy consumption. At the top level, different subsets of clusters are scheduled concurrently in each slot. Inter-cluster interference is modeled as static ambient noise at each link. Using an iterative algorithm, we determine the optimal equilibrium point for each cluster. |
