Scheduling algorithms for tdma wireless multihop networks

Current and future wireless standards use TDMA to provide guaranteed Quality-of-Service (QoS) in the network. While these standards specify how transmissions should occur, they do not discuss scheduling algorithms to find when transmissions should occur (transmission schedules). Despite the technological advances, the question of finding transmission schedules has existed for the past twenty years without a satisfactory answer. This thesis presents a new class of scheduling algorithms for Time Division Multiple Access (TDMA) wireless multihop networks. These algorithms have three major advantages. First, they take into account overhead and delay. With reduced overhead, transmission schedules have much higher throughput than what is possible with previous approaches. The algorithms can also be customized to produce schedules with specific delay properties. Scheduling to achieve a specific delay opens up a new dimension in wireless scheduling that was previously not possible. Second, the algorithms provide a simple and computationally efficient way to specify exact constraints on end-to-end flows in the network. These constraints provide us with a way to solve two important cross-layer design problems in TDMA wireless multihop networks. Third, the algorithms can be distributed in a very natural way. The distributed versions of the algorithms achieve the same performance as the centralized algorithms. The practical value of the algorithms is shown by simulating their usage in two different types TDMA wireless multihop networks.