Research On Network Coding Oriented Scheduling Mechanisms In Wireless Mesh Networks

Wireless Mesh Networks (WMNs) include a group of mobile terminals, which have the ability to organize networks dynamically, and can be considered as one kind of non-infrastructure networks. Due to the characteristics of large capacity, high transmission rate, low cost and good scalability compared with Ad hoc networks, WMNs have gained extensive attention in both industry and academia areas, and become one of popular applications for wireless broadband access to solve "the last mile" problem. As the development of Internet of Things (IoTs) and the ever-increasing requirement of wireless bandwidth, WMNs are widely utilized as the wireless access networks for metropolitan area network, vehicular communications and intelligent terminals, etc.With the advent of big data communication era, wireless multimedia services have replaced the traditional voice and data communications, and become the mainstream of network traffic. Therefore, how to satisfy the ever-increasing wireless spectrum requirement is an important issue. Although various technical solutions have been presented to improve the throughput of WMNs, few existing schemes considered the interactions when multiple transmission methods are utilized simultaneously (for example, the interplay between NC and link scheduling, and the interaction between power control and NC). Furthermore, most of the related literatures assume the Channel State Information (CSI) can be obtained in advance, however, the CSI varies from time to time, which is difficult for estimation in real wireless environment. As the researches of IoTs become hotter and hotter, the characteristics of wireless terminals will be more and more affected by the social relationships of human beings. This will bring new challenges to the emerging social network communication.In order to solve the problems stated above, this thesis includes four parts:a wireless channel estimation-based scheduling scheme; an opportunistic scheduling algorithm using Markov chain; an optimal NC-aware multi-channel multi-radio multi-rate scheduling method, and a NC-oriented scheduling method for social network. The main works and contributions are illustrated as follows:(1) Since wireless environment varies from time to time, the CSI is difficult for estimation, wireless channel gain based scheduling schemes and optimal power distribution methods are presented in Chapter 2. At first, a channel estimation scheme is proposed which bases on Probability of Outage (PoO), and the close-form expressions of PoO for both Analog Network Coding (ANC) and Time Division Broadcasting (TDBC) are derived. Then, a dynamic sampling window is adopted to estimate the values of PoO, based on which the channel gain can be estimated. The presented scheme is adaptive to the change of network environment, and compared with the methods based on physical layer, this scheme is simple for implementation, and is not restricted by network topology. Furthermore, an opportunistic scheduling scheme based on channel gain is presented in order to maximize transmission rate. At last, with the objective of minimizing total power consumption given that the required transmission rates can be satisfied, energy efficient power allocation methods for both ANC and TDBC are proposed.(2) Chapter 3 presents a scheduling method for Physical-layer Network Coding (PNC), in order to utilize wireless spectrum resource effectively for network throughput improvement. A more complicated network topology than the traditional two-way relay channel model is considered, where multiple nodes communicate with each other via one relay node. An opportunistic scheduling method using Markov chain is proposed, where the state of Markov chain is the number of packet that can be received by the relay node at the current moment. Then we derive the transition probability of each state in the Markov chain, and analyze network throughput for both symmetric and asymmetric channel statuses. Our method is more power-efficient in an asymmetric channel situation than a symmetric channel scenario, and therefore is promising since most of the channel states are asymmetric in real network.(3) Although the integration of multiple transmission technologies can improve network performance largely, the research on how to handle the interactions among different transmission methods just starts. In order to maximize network throughput, Chapter 4 presents a NC-aware multi-rate multi-channel multi-radio scheduling method, which supports different kinds of transmission schemes. A NC-aware scheduling method, which considers the interplay between NC and spatial reuse, is proposed at first. Due to the high computational complexity of the formulated problem, the Column Generation (CG)-based method is utilized. In order to reduce the complexity of CG further, a heuristic algorithm is presented to handle the pricing problem in CG. Furthermore, a virtual link aggregation method is presented, which can extend the unicast transmission to multicast transmission. At last, the single-channel single-radio transmission scenario is expanded to multi-channel multi-radio transmission situation, and a heuristic algorithm is presented to solve the channel assignment problem.(4) Due to the increasing popularity of IoTs applications, the wireless terminals will be more and more affected by the social attributes of human beings. According to the social relationship and node selfishness, a NC-aware adaptive scheduling scheme is studied based on social network for relaying method selection. The social relationship of each node is calculated at first. Then a double-auction based next-hop selection method is presented, which encourages long transmission links to be separated into multi-hop transmissions, so that network interference can be reduced, and more opportunities for link aggregation can be created to utilize the promising transmission methods. Furthermore, in order to maximize social welfare and network throughput, a NC-oriented scheduling method for social network is proposed. Due to the high complexity of the formulated problem, an iteration algorithm based on firefly algorithm with low computational complexity is presented at last.In order to verify the effectiveness and advantage of the proposed methods, C++ and Qualnet are used to set up simulation platform. Simulation results demonstrate that, compared with the existing schemes, the presented methods are effective and can obtain better network performance.