无人机网络中基于分层博弈的干扰对抗频谱接入优化

针对无人机通信网络中的干扰对抗问题,提出了一种基于分层博弈的自适应频谱接入优化机制.考虑到无人机网络节点的动态特性,将干扰器视为分层博弈领导者,无人机用户视为分层博弈跟随者,不同层级间博弈参与者具有不同的效应函数,采用斯坦伯格均衡分析所构建的博弈并证明了均衡解的存在性和唯一性.在此基础上设计一种分层信道选择学习算法来求...     

基于图论的动态频谱分配

在认知无线电系统中,动态频谱分配是决定能否充分高效利用有限频谱的关键.文中建立起网络拓扑模型,分析了基于图论着色模型的动态频谱分配技术的发展过程、研究现状.并讨论了其未来的研究方向,主要是在算法改进上.     

基于潜在博弈的认知无线电频谱分配研究

为解决认知无线电网络中认知用户、授权用户共存情况下的频谱分配问题,提出了一种基于潜在博弈的认知无线电频谱分配模型,该模型以最小化系统总干扰水平为目标,认知用户采用避免机会浪费的改进型策略动态调整规则进行频谱选择,经过有限改进路径快速收敛到博弈的纳什均衡点。在30个认知用户和2个授权用户共存的场景下进行仿真实验,算法以较快的收敛速度实现了频谱分配的目标,证明了模型及算法的可行有效。     

基于干扰消减的认知无线电频谱分配算法

在认知无线电网络的频谱分配过程基础上,提出了一种基于干扰消减的频谱分配算法.该算法通过将可用频谱分配给能够同时无干扰地接入同一频谱的所有认知用户来提高授权频谱的使用率.同时,该算法参考各个认知用户在初始阶段的可用频谱数量来为未分配到频谱资源的认知用户进行频谱分配,对频谱分配过程的公平性进行了优化.仿真结果表明,该算法能够在认知用户数量较多、可用频谱紧张的情况下获得较高的吞吐量.     

一种基于功率控制和潜博弈的动态频谱分配算法

功率控制是感知无线电重配置的重要内容之一。结合功率控制的动态频谱分配策略可以有效提高频谱分配的性能。文章构建了一种潜博弈模型,通过在授权链路保护算法中加入功率控制机制,进一步降低对系统中授权用户和其它感知用户的干扰,提高通信链路的信干噪比,改善分配算法的性能。仿真结果表明算法达到了预期的效果,保证了感知无线电系统可靠通信。     

优先级保护的分布式动态频谱分配算法

对认知无线网络中动态频谱分配算法进行了研究,在分析原有的ADP（Asynchronous Distributed Pricing）算法的前提下,运用合作博弈的理论提出了具有干扰价格因子的ADP算法,分配过程中所有用户发布自己的干扰价格,用户在选择信道和功率的时候需要考虑到自身的效用和对其它用户的干扰,可以在有效保证高优先级用户性能的前提下最大化网络效用。仿真结果表明,该算法可以达到纳什均衡并且具有极快的收敛速度,可以提高高优先级用户的信干噪比。     

异构无线网络中基于覆盖概率的动态频谱分配算法

该文研究异构无线网络下行链路动态频谱分配(DSA)问题。首先考虑用户分布对下行链路干扰分析的影响及不同业务用户的信干比(SIR)要求,利用用户分布密度函数将小区内用户的SIR性能转化为小区基站的覆盖性能,在阴影衰落信道条件下,提出基于覆盖概率的干扰控制模型;基于该干扰控制模型,以网络效益最大化为目标,将DSA建模为非线性约束组合优化问题,进而结合贪婪思想设计了基于覆盖概率的DSA(CP-DSA)算法。仿真结果表明,CP-DSA算法能够提高网络效益,有效控制基站之间的干扰,满足用户的SIR要求。     

基于动态联盟的无人机集群协同干扰方法

针对无人机集群对多个异构电磁目标进行协同干扰时面临的多智能体协同学习决策问题,提出了基于动态联盟的无人机集群协同干扰方法。运用马尔可夫决策对协同干扰进行建模,构建干扰联盟博弈模型,优化集群干扰中的频率和功率联合决策以及干扰资源的动态调配,实现对多个用频模式不同、压制功率需求不同的电磁目标进行干扰的系统效能优化。首先,利用协同强化学习算法对任务目标的信道决策进行学习,联盟内无人机协同干扰;然后,根据干扰效果进行干扰联盟结构的动态调整。所提算法可从强化学习进程中根据干扰效能的差异动态调整干扰资源,并能适应目标功率的变化,相同场景下,系统干扰效能值较基线算法提升约12.5%。     

一种动态频谱分配的多层博弈模型构建分析

采用随机几何和博弈论相结合的方法,研究了蜂窝网络中用户对于移动网络运营商(Wireless Service Provider, WSP)的选择与WSP频谱分配方案的制定之间的联系;构建了一个层次化的博弈框架来模拟用户和WSP之间的复杂交互,采用演化博弈模型来描述用户之间的竞争,用非合作博弈模型来描述WSP之间的竞争,用多领导者多跟随者的Stackelberg博弈模型来描述用户和运营商之间的循环依赖问题;分析了演化博弈模型中演化均衡的存在及其渐近稳定性,并在此基础上证明了非合作博弈模型中纳什均衡的存在。     

基于单频段多赢家拍卖的动态频谱分配

针对传统拍卖方法在认知无线网络动态频谱分配中的低效问题,提出了一种基于单频段多赢家拍卖的动态频谱分配算法。该算法在原始贪婪算法的基础上增加了多重贪婪策略,以较低的计算复杂度获得了较优的解;改进了VCG(vickery-clarke-groves)机制,在保留占优策略激励兼容特性的基础上,提高了卖家的收益,有效抑制共谋的发生。仿真结果表明,该算法的频谱分配效率接近最优分配效率,同时提高了拍卖的经济收益。     

Dual-target WOA spectrum sharing algorithm based on Stackelberg game

In order to solve the complex spectrum allocation problem,a dual-target whale optimization algorithm (WOA) with strong parallel computing capabilities was introduced,and a Stackelberg game model was proposed that could effectively reflect the actual spectrum requirements,and a dual-target WOA optimized distributed antenna system (DAS) spectrum sharing scheduling algorithm was designed.Simulation results show that performance comparison is performed from multiple indicators such as optimal pricing and user benefits.The proposed algorithm has a good spectrum sharing allocation effect,can achieve fair and effective spectrum allocation,and provides an important reference for the future communication network spectrum sharing mode.     

基于干扰图的家庭基站分簇频谱分配算法

为了减小宏用户和家庭基站之间的干扰,有效分配频谱,本文提出了基于图论着色的分簇信道分配算法。该算法在保证满足宏用户信干噪比(signal-to-interference plus noise ratio,SINR)要求的前提下确定每个家庭基站可用的子信道集,再根据构造的家庭基站系统干扰图,动态地给每个用户分配所有可用的频谱资源。本文分别构建无向干扰图和有向权重干扰图,使用了一种基于簇的改进算法,在保证宏用户信干噪比的条件下,尽量提高家庭基站的吞吐量。仿真结果表明,本文算法可以降低宏用户和家庭基站的中断率,同时频谱效率得到提高。     

Stackelberg game theoretic pricing algorithm for bandwidth allocation in cooperative access

Cooperative access among user devices by sharing wireless access bandwidth opens a new paradigm in heterogeneous networks.However,how to stimulate cooperative relay nodes forwarding service data for others and allocating corresponding bandwidth to support it are two key issues in the cooperative access.This paper proposes a Stackelberg game based framework which is benefit participants including relay nodes and client nodes.This framework generalizes the pricing based bandwidth allocation algorithm by the Stackelberg game model,which optimizes the profit of the cooperative relay nodes while guaranteeing the bandwidth requirements of client nodes.We transform the profit maximization problem into a convex problem and solve it using the convex optimization method.The simulation results demonstrate that the proposed framework and corresponding algorithms outperform the bidding weight proportional fairness and fixed value bandwidth allocation ones significantly.     

Stackelberg game based interference management for two-tier femtocell networks

Femtocell is viewed as a promising option for mobile operators to improve coverage and provide high-data-rate services in a cost-effective manner. This paper considers the uplink interference management problem in a spectrum-sharing femtocell network. Assuming that the macrocell base station (MBS) is rewarded for sharing the spectrum with femtocells by setting a reasonable interference cap (IC) for femtocell users’ (FUEs’) transmissions. Within IC, the FUEs allocate their transmission powers competitively while not introducing much interference to both the macrocell users (MUEs) and other FUEs. A Stackelberg game is formulated to jointly maximize the utility of MBS and the individual utility of FUEs. Specifically, the maximum tolerable interference at the MBS is used as the resource that the leader (MBS) and the followers (FUEs) compete for. Then, the backward induction method is applied to achieve the Stackelberg equilibrium and a distributed power update rule is developed for FUEs. In addition, the implementation protocol is presented, some issues related to the implementations and some future extensions regarding the MUEs’ uplink protection are discussed. Lastly, numerical results demonstrate the performance of our proposed power allocation in detail, and show the effects of varying the number of FBSs and changing other system parameters on the system’s performance.     

认知异构网络中基于克隆选择算法的动态频谱分配

建立了认知异构无线网络的系统模型。基于该模型,以网络效益最大化为目标,考虑接入网的频谱需求,接入网之间的干扰和对应异构网络不同接入技术的多粒度重叠信道之间的干扰约束条件,将频谱资源分配建模为非线性约束0-1整数规划问题,进而提出基于克隆选择的认知异构网络中动态频谱分配算法,并在该算法中设计了一种新的能够同时考虑接入网频谱需求和多粒度信道频谱资源的抗体整数编码方式。仿真结果表明,所提算法相比于贪婪分配算法,增加了网络效益,提高了频谱使用效率。     

基于改进遗传算法的认知无线传感网动态频谱分配方案

将认知无线电中的动态频谱分配技术应用在无线传感网中,针对工作在ISM(industrial,scientific and medical)频段的无线传感网面临的频谱资源紧缺问题,提出一种基于改进自适应遗传算法的动态频谱分配方案.该算法以图论着色模型为基础,以最大带宽收益和最小切换频率为目标函数,在交叉和变异过程中采用自适应交叉概率和变异概率代替固定的交叉概率和变异概率.仿真结果表明,与传统遗传算法和颜色敏感图论着色算法相比,该算法可以实现提高频谱利用率、降低能量消耗的预期目标.     

基于并行分配算法的认知无线电频谱分配算法

针对传统频谱分配方式不合理造成频谱匮乏以及CSGC算法在频谱分配上消耗时间过长影响实时通信的问题,文中提出了一种基于并行分配算法的认知无线电频谱分配算法,其通过同时给各顶点进行上色,从而节省了时间开销。实验证明,并行算法不仅具有与CSGC算法同样的高频谱利用率,且其频谱分配时间开销更低,是一种理想的认知无线电频谱分配算法。     

基于强化学习的动态频谱分配研究

首先介绍了认知无线电技术产生的背景,以及强化学习的发展和应用于认知领域的优势;接着对强化学习的基本原理及其2个常见的模型Q-Learning和POMDP作了介绍,并对其模型定义、思想、所要描述的问题和使用的场景都做了较详细的阐述;然后针对这个方向最近几年的顶级会议和期刊论文,分析了其主要内容;通过最近几年的学术、会议论文中所述的研究现状及成果,说明强化学习的主要特点是能够准确、快速学习到最优策略,能够模拟真实环境,自适应性强,提高频谱感知、分配效率,从而最大化系统吞吐量,这些优势充分证明了强化学习将是认知领域里一种很有前景的技术。     

Stackelberg game for backhaul resource allocation in the two-tier LTE femtocell networks

As one promising technology for indoor coverage and service offloading from the conventional cellular networks, femtocells have attracted considerable attention in recent years. However, most of previous work are focused on resource allocation during the access period, and the backhaul involved resource allocation is seriously ignored. The authors studied the backhaul resource allocation in the wireless backhaul based two-tier heterogeneous networks （HetNets）, in which cross-tier interference control during access period is jointly considered. Assuming that the macrocell base station （MBS） protects itself from interference by pricing the backhaul spectrum allocated to femtocells, a Stackelberg game is formulated to work on the joint utility maximization of the macrocell and femtocells subject to a maximum interference tolerance at the MBS. The closed-form expressions of the optimal strategies are obtained to characterize the Stackelberg equilibriums for the proposed games, and a backhaul spectrum payment selection algorithm with guaranteed convergence is proposed to implement the backhaul resource allocation for femtocell base stations （FBSs）. Simulations are presented to demonstrate the Stackelberg equilibrium （SE） is obtained by the proposed algorithm and the proposed scheme is effective in backhaul resource allocation and macrocell protection in the spectrum-sharing HetNets.