In Chapter 5, we looked at scheduling in wireless networks, where the focus was to design a scheduling algorithm that can support any traffic strictly within the capacity region. Implicitly, we assumed that data flows were regulated by congestion control so the incoming traffic was always within the capacity region. However, in practice, the behavior of congestion control algorithms themselves may be affected by scheduling algorithms. For example, a dual congestion control algorithm which reacts to delays, so the source rates would be regulated based on queueing delays, which could be quite different under different scheduling algorithms. Therefore, to achieve fair resource allocation in wireless networks, we need to revisit the network utility maximization formulation introduced in Chapter 2. In Chapter 2, we assumed that routes from sources to their destinations are given. In this chapter, we will study a more general model in which routes are not given and are part of the resource allocation decision process. We will consider the network utility maximization problem for wireless networks, and answer the following questions.
• How should the Lagrangian duality to derive joint congestion control, routing, and scheduling algorithms that maximize the sum network utility be used?
• How should the performance of the joint algorithms using discrete-time Markov chains be analyzed?
• What is the difference between the network utility maximization formulation for the Internet and that for wireless networks, and what is the rationale behind the difference?
• What is the relationship between proposed models/algorithms and practical MAC/physical layer protocols?
Review the options below to login to check your access.
Log in with your Cambridge Aspire website account to check access.
If you believe you should have access to this content, please contact your institutional librarian or consult our FAQ page for further information about accessing our content.