Jointly Optimal Power Control And Congestion Control For Ad Hoc Networks

Ad Hoc networks are wireless, mobile, self-organizing networks, without the support of the wired infrastructure, and need not set any central control node. The nodes in Ad Hoc networks can communicate with each other in the form of multi-hop. Therefore, the Ad Hoc networks have strong survivability. Ad Hoc networks have broad application prospects, due to its flexibility and quick instauration.Nodes in Ad Hoc network rely mainly on limited battery energy supply, Thus, effective power control technology, which can save the energy and prolong the network lifetime, is a major task in the researches of Ad Hoc networks. the radio channel, used by the nodes, is impacted by the signal attenuation, noise, interference and other factors, so the channel capacity is very limited and is changed over time. Therefore, Ad Hoc networks are more prone to congestion. Power control technology can adjust the powers of the links and the level of interference between the links to improve the channel capacities of the network, while saving energy consumption, and congestion control technology can rationally allocated the data flows in the network to avoid congestion and improve network resource utilization. Consequently, jointly power control and congestion control technology is a very interesting research job.The main contents and innovations of this paper include the following aspects:First, we can make the maximum throughput of the network by solving the traditional Network Utility Maximization problem, but it can not reduce the energy consumption and prolong the life of the network. In this paper, we introduce a cost function of the energy consumption into the objective function of the traditional Network Utility Maximization problem, and propose a "green" Network Utility Maximization problem. Then we adopt primal Lagrangian decomposition technique to deduce the corresponding jointly power control and congestion control algorithms using the gradient project method and analyze the convergence of the algorithms. Also we validate the proposed algorithm through Matlab simulation platform and analyze the performances of the algorithms. So we can get a conclusion that the proposed algorithms in this paper can increase the throughput of the network, meanwhile, reduce the energy consumption and prolong the life of the network.Second, Service Quality Assurance is one of the main issues in the research of the Ad Hoc networks presently. We introduce delivery contract to the "green" Network Utility Maximization problem, so that the network can provide users with a certain Service Quality Assurance, and put forward the corresponding jointly power control and congestion control algorithms to solve this problem. We analyze the convergence of the algorithms and verify algorithms performances through Matlab simulation platform. So we can get a conclusion that the jointly power control and congestion control algorithms limited by the delivery contract can make the network provide users with a certain Service Quality Assurance.Third, the performances of the networks are different, if the networks use the different channel access schemes. Since the Ad Hoc networks are self-organization and have no central control nodes, the channel access schemes chosen is not fixed, but changed with certain randomness. Therefore, we propose a "green" Network Utility Maximization problem based on random scheduling. After transforming the problem into a Lagrangian dual problem, we decompose the problem according with channel access schemes, and then derive the jointly optimal power control and congestion control algorithms for every channel program. Also we verify and analyze the algorithms performances through Matlab simulation platform.