| This dissertation studies wireless scheduling by jointly optimizing the physical (PHY) layer and the medium-access-control (MAC) layer. It consists of the following three main thrusts.;In the first thrust, rate optimization for multicast communications at the medium access control (MAC) layer is studied. For Threshold-T based multicast policies, the transmission rates are optimized to maximize the throughput in stable networks and in saturated networks, respectively. To ensure multicast reliability when no retransmission is required at the MAC layer, transport layer erasure coding is used for reliability enhancement.;The second thrust focuses on distributed opportunistic scheduling (DOS) for throughput maximization in wireless ad hoc networks based on random access. In such networks, DOS involves a process of joint channel probing and distributed scheduling. The desired tradeoff, between the throughput gain from better channel conditions and the cost for further channel probing, boils down to judiciously choosing the optimal stopping policy for channel probing. First, the collision model is assumed for random access, and the optimal scheduling policy is characterized from two perspectives: (1) For the network-centric case, it is shown that the optimal DOS is a pure threshold policy. (2) For the user-centric case, the problem of threshold selection is treated as a non-cooperative game, and the existence and the uniqueness of Nash equilibrium is explored. Since there is an efficiency loss at the Nash equilibrium, a pricing-based scheme is proposed to mitigate the loss. The study on distributed opportunistic scheduling is also generalized to the network models with multiple-input-multiple-output (MIMO) links and imperfect channel estimation.;Next, DOS under the PHY-interference model is characterized. Different from the collision model, where at most one link can transmit successfully in a one-hop neighborhood, under the PHY-interference model, multiple links can transmit successfully simultaneously and the number of simultaneously transmitting links is also random. Optimal stopping theory is used to explore this problem, and it is shown that the optimal policy for distributed scheduling still has a threshold structure. Observing that the network throughput depends heavily on the contention probabilities of all links, a two-stage algorithm is proposed to jointly optimize the rate threshold and the contention probability.;The third thrust is on developing a cross-layer optimization framework for effective interference management towards understanding the fundamental tradeoffs among possible MIMO gains in multi-hop networks. Under such a common thread, solid PHY-layer interference models are developed, and a set of feasible rates and signal-to-interference-plus-noise-ratio (SINR) requirements are extracted. Based on the PRY-interference models, the structural property of the optimal scheduling policy is characterized, and the problem of joint stream multiplexing and link scheduling is formulated as a combinatorial optimization problem. Efficient centralized algorithms are devised by using a multidimensional knapsack approach, for both the extended network model and the dense network model, respectively. A (contention-based) distributed algorithm is also developed, in which links update their contention probability based on local information only. |
