认知无线网络中时域机会频谱接入策略研究

发布时间：2018-02-16 10:25

  本文关键词： 认知无线网络 时域机会频谱接入 Q学习 周期性感知 请求数据包长度 序贯决策 能量有效性　出处：《解放军信息工程大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着无线通信业务的迅猛发展,现有无线频谱资源分配殆尽加之传统静态频谱分配方式频谱利用率较低,造成了目前无线频谱资源的危机。基于认知无线电的时域机会频谱接入技术能够在避免对主用户产生有害干扰的前提下,“伺机”接入空闲频段对其进行二次利用,进而提高频谱资源的利用率。然而,次用户如何动态接入短暂、可变的频谱机会使得时域机会频谱接入技术面临一系列新的挑战,为了能有效地发现并合理利用无线频谱机会,本文对认知无线网络中的时域机会频谱接入策略展开研究,主要工作如下:1.在时隙主网络中,为了提高次用户在未知环境先验知识条件下的机会频谱接入收益,给出了一种基于Q学习的信道选择算法。该算法克服了传统信道选择算法对环境先验知识的依赖性,使得智能体能够通过与未知环境不断的交互、学习,选择长期累积回报最大的信道接入。在学习过程中,算法综合考虑所选信道的业务负载和信噪比对回报函数的影响,同时引入Boltzmann实验策略,运用模拟退火思想实现了资源探索与资源利用之间的折衷。仿真结果表明,所给算法能够有效提高次用户的接入吞吐量和接入信道的平均信道容量。2.在非时隙主用户网络中,为了优化次用户机会频谱接入时间,提高次用户的接入吞吐量,研究了次用户在周期性感知框架下的机会频谱接入问题。通过建立次用户信道感知、接入模型,详细推导了非时隙主网络中次用户空闲时长的概率表达式,进而提出了有效传输吞吐量的概念,以此为优化目标,给出一种基于次用户业务请求数据包长度的机会频谱接入算法。该算法在选择可用频谱时考虑本时隙次用户传输数据包长度的大小,自适应选择满足空闲时长的可用频谱接入。仿真结果表明,所给算法能够在碰撞约束要求较高情况下提高次用户平均有效传输吞吐量的同时,实现有效吞吐量与碰撞概率的折衷;同时当外部环境发生变化时算法具有较强的鲁棒性。3.在非时隙主用户网络中,为了提高次用户的能量有效性,基于序贯决策理论给出了一种新的自适应机会频谱接入算法。该算法以最大化次用户能量有效性为目标,通过建立感知接入联合优化模型,使得次用户能够以较优的信道感知时间和传输功率接入信道。同时,在功率控制过程中引入次用户请求数据包长度,使得传输功率可以基于数据包长度自适应控制。上述目标优化问题求解过程中,借助于非线性分式规划理论将其转化为线形规划问题,并运用二分搜索算法寻求最优解。仿真结果表明,所给算法能够实现感知性能和感知能耗的折衷,有效提高次用户的能量有效性。4.在VS2010编程平台上借鉴分层模块化设计思想,实现了机会频谱接入策略仿真测试平台,搭建了不同主用户业务模型下的网络环境,进而仿真评估了前面章节给出的机会频谱感知策略,接入策略。测试结果表明本平台具有一定的实用性和可靠性。
[Abstract]:With the rapid development of wireless communication services, the existing wireless spectrum resource allocation. In addition to the traditional static spectrum allocation of spectrum utilization rate is low, resulting in the current crisis of wireless spectrum resources. Time domain opportunities for cognitive radio spectrum access technology can avoid the premise of harmful interference to primary users based on the "opportunistic" access to the free band it used for the two time, and improve the utilization of spectrum resources. However, the secondary users how to dynamically access short variable time domain spectrum opportunity makes opportunistic spectrum access technology is facing a series of new challenges, in order to effectively discover and rational use of wireless spectrum opportunity, this paper studies the time opportunity in cognitive radio network spectrum access strategy, the main work is as follows: 1. in the time slot in the primary network, in order to improve the secondary users in an unknown environment prior knowledge of machine Spectrum access revenue, presents a channel selection algorithm based on Q learning. The algorithm overcomes the traditional channel selection algorithm on a priori knowledge of the environment dependence, making the agent through learning and interaction, the unknown environment constantly, choose a long-term cumulative return maximum channel access. In the learning process, the algorithm considers the selected channel traffic load and the effect of SNR on return function, while the introduction of Boltzmann experimental strategy with simulated annealing approach and explore the tradeoff between resource utilization. The simulation results show that the algorithm can effectively improve the throughput of secondary users access and access channel of the average channel capacity of.2. in unslotted primary users in the network, to the user the opportunity spectrum access time optimization, improve the throughput of secondary users, the secondary users know under the framework of the machine cycle in sexy Will the spectrum access problem. Through the establishment of user access channel sensing model, deduced the probability expression of unslotted primary network of secondary users in spare time, and then put forward the concept of effective transmission throughput, as the optimization goal, an algorithm based on user service request opportunity spectrum access algorithm of data packet length. The algorithm considers the time slot users transmit data packet length size in the choice of the available spectrum, the adaptive selection of the available spectrum access to meet the idle duration. The simulation results show that the algorithm can in collision constraint to improve the user under the condition of higher average effective transmission throughput at the same time, realize the effective throughput and collision probability tradeoff at the same time; when the external environment changes when the algorithm has stronger robustness to.3. in unslotted primary users in the network, in order to improve the energy efficiency of the secondary user, A new adaptive opportunistic spectrum access algorithm based on sequential decision theory is proposed to maximize the user energy efficiency as the goal, through the establishment of joint optimization models of perceptual access, makes users to better channel sensing time and power transmission access channel. At the same time, the introduction of a user request packet length the power control process, which makes the transmission power can be based on data packet length adaptive control. The optimization problem solving process, based on the theory of nonlinear fractional programming is transformed into linear programming problem, and use the two point search algorithm to seek the optimal solution. The simulation results show that the algorithm can achieve the sensing performance and perceived energy compromise.4., effectively improve the energy efficiency of the secondary users in the VS2010 programming platform using hierarchical modular design, realizes the opportunistic spectrum access strategy A simulation test platform is built, and the network environment under different main user business models is built. Then the opportunistic spectrum sensing strategy and access strategy presented in the preceding chapters are simulated and evaluated. The test results show that the platform is practical and reliable.

【学位授予单位】：解放军信息工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925

【参考文献】

中国期刊全文数据库 前5条

1 张凯;刘洋;赵彪;李鸥;;认知Ad Hoc网络能量有效频谱接入策略[J];信号处理;2013年07期

2 任丙印;王大鸣;崔维嘉;胡捍英;;基于碰撞概率约束的认知系统子信道分配算法[J];太赫兹科学与电子信息学报;2013年02期

3 张凯;李鸥;杨白薇;;基于Q-learning的机会频谱接入信道选择算法[J];计算机应用研究;2013年05期

4 张晶;朱洪波;;基于跨层优化的机会频谱接入算法[J];通信学报;2010年11期

5 张永靖;冯志勇;张平;;基于Q学习的自主联合无线资源管理算法[J];电子与信息学报;2008年03期

，

本文编号：1515294