【作者】 张悦；

【导师】 杨博；

【作者基本信息】 上海交通大学 ， 控制科学与工程， 2015， 硕士

【摘要】 智能终端的迅速普及给人类生活带来了极大便利,然而由于无线通信系统的广播特性,这也给企业以及个人用户数据的隐私安全带来了新的威胁。以往学者们主要在网络高层研究安全性问题,但近年来,物理层安全性研究由于其优势正逐渐成为研究热点。然而,根据香农信息论,较高的保密需求往往限制着更大的网络流量,因此我们需要寻找一种最优载波分配方案。在下一代无线网络中,多址频分接入(OFDMA)技术由于其资源分配的灵活性和频谱利用的高效性特征,正成为一种较理想的接入方案。与此同时,智能终端高速率传输增大了基站的传输功耗,对有限的资源供给带来了很大的挑战。因此,本文聚焦于如何利用绿色新能源与基站的无线充电技术来应对日益加剧的环境污染问题和资源不足的瓶颈。综上,本文主要针对以下三个无线传输场景设计了OFDMA下行链路网络的跨层资源分配算法:1.基于混合能源供电的基站流量动态定价及资源分配。为了减少系统的电费开支,本文首先为基站配置了新能源和电网的混合供电系统,该系统根据实时的电力价格以及基站功耗需求进行充放电决策。接着考虑到网络盈利系统,我们对每位有随机数据传输请求的用户收取动态的流量费,一方面可以最大化网络净收益,另一方面通过流量控制来达到网络的稳定性和资源的最优分配。基于上述网络模型,我们设计了跨层的PCHA算法,并通过严谨的理论证明和基于matlab平台的仿真验证了该算法的网络稳定性及最优性。2.基于新能源供电的中继选择及基站流量控制和资源分配。为了改善边缘地区用户的数据传输速率,本文构建了OFDMA中继协作网络模型,通过为中继配置新能源充电设备降低了安装电力电缆线和电网电费的成本。此外,我们利用了动态队列技术和李雅普诺夫优化方法来应对复杂的网络结构以及较难获取的统计信息带来的挑战;设计了跨层的RCNA在线算法,在分配功率和子载波的同时优化中继选择,并证明了实际数据队列有确定性的上界,保证了网络稳定性,并且算法性能可以通过调节控制参数V无限接近最优值。3.基于无线充电技术的基站模式选择、时延控制及资源分配。为了解决室内用户无法获取室外新能源充电的难题,同时提高低数据请求下的频谱利用率,我们在第一部分安全传输的OFDMA网络模型基础上,拓展了基站的工作方式——无线充电模式。然而,基于安全传输的无线充电模式会给用户带来较高的传输延迟。因此,我们构建了虚拟时延队列和新颖的李雅普诺夫函数,设计了跨层DFRA资源调度算法:当基站的随机数据请求较少,用户的时延需求较低或者移动终端电量较低时,基站会通过无线电波给用户终端充电。值得一提的是,该算法将数据缓存压力从实际队列转移到了传输层的虚拟队列,即网络吞吐量与时延上界存在折衷的关系;但是实际数据队列有即时性的明确上界,并且不受控制参数V的影响,即当算法吞吐量无限逼近最优时不会影响基站处数据缓存队列的长度。更多还原

【Abstract】 With the rapidly increasing popularity of intelligent terminals, great convenience has been brought to human life. Meanwhile, due to the broadcast nature of wireless communication systems, problems concerning with security requirements remain a major challenge to enterprise and individual users. Traditionally, most of security work is undertaken on upper layers based on computational complexity. Only in recent years,physical layer approach to achieve security has attracted considerable attention for it quanti?es the measure of secrecy and greatly facilitate cross-layer design. According to information theory, however, secure transmission undesirably antagonizes larger ?ow rate, which is a bottleneck in the course of research. To address this problem, it’s an urgent task for us to design an optimal subcarrier allocation algorithm. Besides, OFDMA is becoming the best candidate for the user interface in the wireless network of next generation because of its high e?ciency on the utilization of frequency spectrum and the high ?exibility on the allocation of resource. At the same time, high-speed transmission of smart terminals increases the transmission power of the base station,which brings about a big challenge for limited supply of resources. In this paper, we are interested in how to utilize renewable green energy and wireless power transfer technical to tackle the increasingly intensifying environmental pollution problems and inadequate resource bottlenecks. Based on the above facts, the objective of this paper is to design cross-layer resource allocation algorithms of OFDMA downlink network to the following three wireless transmission situations:1. Dynamic pricing and hybrid energy management with security requirements. To reduce electricity cost of base station, we equip BS with a hybrid energy supply system consisting of renewable energy and power grid, which operates based onthe observation of time-varying electricity price under real electricity markets and the power demand at the base station. To establish a system-wide revenue,for the network administrator, the algorithm charges each user with dynamic admission fee to fully exploit network resources and carry out congestion control at the same time. Based on the above network model, we design the cross-layer PCHA algorithm and verify its network stability and optimality through rigorous theoretical proof and Matlab simulations.2. Joint optimization of base station and relay power allocation for cooperative OFDMA systems with renewable energy harvesting relay. To improve the provision of high throughput for mobiles at the cell edge that always suffer from bad channel conditions, cooperative OFDMA network model is constructed to solve the problem of equipping cables and greatly reduce power cost. Besides, by the means of dynamic queueing and Lyapunov optimization, we convert the time average maximization problem formulation to a series of instantaneous subproblems. The proposed algorithm, RCNA, investigates the joint problem of power allocation and subcarrier assignment with relay selection at the same time. Finally, we prove our algorithm can guarantee network stability and obtain performance extremely close to the optimal by adjusting control parameter V, which is also proved by simulation results.3. Wireless power transmission based BS mode selection, delay control and resource allocation. To handle the problem of outdoor renewable resource unavailable to indoor mobiles and improve frequency e?ciency under low data requests,we extend BS operation mode to wireless power transmission mode. However,secure transmission based wireless power transmission mode may bring about relatively high transmission delay. Thus, we construct virtual delay queue and propose a novel Lyapunov function to design cross-layer DFRA algorithm: in the case that the random data request to BS is relatively low and users’ delay demand is loose or mobile terminals’ battery power is inadequate, BS will transmit power to users through radio frequency.Speci?cally, DFRA transfers data buffer pressure from actual data queue of network layer to the virtual queue of transmission layer, which means there existstradeoff between network throughput and delay bound. What’s more, the actual data queue has speci?c and instantaneous bound, which is independent of control parameter V, that is to say, actual data queue length won’t be increased when performance extremely closes to the optimal.更多还原