基于多智能体强化学习的动态频谱分配方法

针对认知无线电网络中多个异质用户具有不同的服务质量(QoS)要求,提出一种基于多智能体强化学习的动态频谱分配方法。该方法从用户满意度角度出发,以用户体验质量(QoE)作为系统的评价指标,构建多个虚拟智能体,模拟多个用户以合作方式与环境进行交互学习,融合各个用户的学习和频谱决策结果,实现频谱资源优化分配。仿真结果表明,在未知主要用户使用规律和信道动态特性条件下,相比基于传统强化学习的动态频谱分配方法,提出的方法能有效提高次用户的QoE,降低用户间的冲突概率。     

恶意干扰下基于斯坦伯格博弈的无人机动态频谱分配

针对恶意干扰场景下无人机群动态频谱分配问题,构建了基于斯坦伯格博弈的动态频谱分配模型,干扰机为斯坦伯格博弈的领导者,无人机群为斯坦伯格博弈的跟随者,设计了不同博弈参与者的效益函数,并证明了该博弈存在稳定的斯坦伯格均衡解。在此基础上设计了一种分层动态频谱分配算法,针对领导者采用Q学习选择干扰信道的场景下,跟随者采用随机学习自动机来确定信道分配策略。仿真结果表明,所提算法能够得到无人机用户的最优信道分配策略,有效提升无人机用户的总吞吐量性能,实现效益最大化。     

认知无线多跳网中保证信干噪比的频谱分配算法

为解决无线多跳网络在固定频谱分配方式下所固有的信道冲突等问题,利用认知无线电的动态频谱分配技术,提出了一种适用于次用户组成的无线多跳网络的、underlay方式下的全分布式频谱分配算法。该算法将频谱分配问题建模成静态非合作博弈,证明了纳什均衡点的存在,并给出了一种求解纳什均衡点的迭代算法。大量仿真实验证明,该算法能实现信道与功率的联合分配,在满足主用户干扰功率限制的同时,保证次用户接收信干噪比要求。     

基于频谱数据库的分布式动态频谱接入算法

频谱数据库是一种直接获得频谱信息的方式,针对在复杂多变的网络环境中用户之间缺少信息交互,研究了数据库协助和没有数据库情况下的动态频谱接入算法。一方面,次用户通过历史感知数据估计信道的可用性而进行信道选择,另一方面,次用户通过数据库获得更加可靠的频谱信息从而做出策略。证明了用户之间的博弈是一个超模博弈,通过提出的分布式学习算法能收敛到一个纯策略纳什均衡点。仿真结果表明,提出的联合数据库感知算法和数据库协助算法比依靠感知的结果收敛更快,获得的系统吞吐量接近最大,减少了用户决策的时延,提高了频谱利用率。     

面向数据传输稳定性的无人机群频谱接入方法

针对频繁变化的无线通信环境中无人机群频谱资源使用问题,在保障无人机数据传输稳定的前提下,以降低无人机间用频冲突率、提升信道利用率和系统吞吐量为目标,设计了基于多用户非耦合排队频谱接入方法。该方法中无人机之间不需要信息交互和中心控制节点,仅需根据自身数据传输需求、互扰功率和历史信道使用状态信息来构建学习效益函数,生成频谱使用策略。仿真结果表明,所提方法在确保无人机数据传输稳定和兼顾无人机用频公平性的前提下,能有效降低无人机间用频冲突,提升无人机群吞吐量。     

基于地理位置的频谱共享系统功率分配

针对次用户地理位置分布的随机性,提出一种基于地理位置的多信道underlay频谱共享模型,并设计了一种基于凸优化的功率分配算法以提高频谱资源利用率。首先根据不同地理位置的次用户对主用户干扰的差异,设计一种基于产出投入比的公平性信道分配准则,将复杂的多信道资源分配问题转化为单信道功率分配问题;继而采用一种基于凸优化的迭代寻优方式进行功率分配,最终获得次用户的系统信道容量。仿真结果表明,相比按接入时间先后分配信道的方法,基于地理位置的多信道underlay频谱共享模型可以获得更高的信道容量。     

认知无线电网络基于F范数的频谱共享

针对多用户多输入多输出认知无线电网络的频谱共享问题,提出一种在保证授权用户服务质量要求的前提下,以认知网络容量最大化为目标的基于F范数的频谱共享方法.该方法利用信道矩阵的F范数选择认知用户以获得认知网络的多用户分集增益,并采用两次选择的方式降低算法的复杂度,通过将认知用户的发射信号投射到干扰信道的零空间来避免认知用户对...     

基于强化学习的合作频谱分配算法

为了解决认知无线电网络中的频谱分配问题,提出了一种基于用户体验质量的合作强化学习频谱分配算法,将认知网络中的次用户模拟为强化学习中的智能体,并在次用户间引入合作机制,新加入用户可以吸收借鉴其他用户的强化学习经验,能够以更快的速度获得最佳的频谱分配方案;并且在频谱分配过程中引入了主用户和次用户之间的价格博弈因素,允许主用...     

IEEE802．22 WRAN网络中基于拍卖的频谱租借算法

在IEEE802．22WRAN网络中,基站间的频谱共享是一个重要的研究方向。若每个WRAN基站伺机占用信道,将带来信道分配不均的问题。利用拍卖理论激励基站间平衡信道分配,提出一种动态的频谱租借算法。该算法将频谱共享情况建模为一个拍卖基站和多个竞拍基站的拍卖模型,同时考虑竞拍基站间的信道需求和拍卖基站的收益,建立频谱共享的数学优化模型。仿真结果表明,与已有的频谱拍卖机制相比,该算法能获得较高的拍卖收益。     

基于业务请求数据包长度的机会频谱接入算法

针对非时隙主用户网络,研究了单个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和接入模型,提出了一种基于次用户请求业务数据包长度的机会频谱接入算法。该算法根据每个时隙分配给次用户业务数据包长度,自适应调整机会频谱接入策略。仿真结果表明,所提算法能够在干扰水平要求较高情况下,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中;同时当外部环境发生变化时算法具有较强的鲁棒性。     

Resource allocation based on dynamic hybrid overlay/underlay for heterogeneous services of cognitive radio networks

In cognitive radio networks (CRNs), hybrid overlay and underlay sharing transmission mode is an effective technique to improve the efficiency of radio spectrum. Unlike existing works in literatures where only one secondary user (SU) uses both overlay and underlay mode, the different transmission modes should dynamically be allocated to different SUs according to their different quality of services (QoS) to achieve the maximal efficiency of radio spectrum. However, dynamic sharing mode allocation for heterogeneous services is still a great challenge in CNRs. In this paper, we propose a new resource allocation method based on dynamic allocation hybrid sharing transmission mode of overlay and underlay (Dy-HySOU) to obtain extra spectrum resource for SUs without interfering with the primary users. We formulate the Dy-HySOU resource allocation problem as a mixed-integer programming to optimize the total system throughput with simultaneous heterogeneous QoS guarantee. To decrease the algorithm complexity, we divide the problem into two sub-problems: subchannel allocation and power allocation. Cutset is used to achieve the optimal subchannel allocation, and the optimal power allocation is obtained by Lagrangian dual function decomposition and subgradient algorithm. Simulation results show that the proposed algorithm further improves spectrum utilization with simultaneous fairness guarantee, and the achieved Dy-HySOU diversity gain is satisfying.     

A Parallel Repeated Auction for Spectrum Allocation in Distributed Cognitive Radio Networks

Cognitive radio (CR) is applied to solve spectrum scarcity. Although the auction theory and learning algorithm have been discussed in previous works, their combination is not yet researched in the distributed CR networks, where secondary users (SUs) can occupy several channels simultaneously by assuming that one channel can be accessed by at most one SU. A parallel repeated auction scheme is proposed to solve resource allocation in multi-user multi-channel distributed spectrum-overlay CR networks. A novel bid scheme in the light of the first-price sealed auction is designed to balance the system utility and allocation fairness. The proposed auction scheme can be developed based on a learning algorithm and be applied to the scenarios where the cooperation among SUs is unavailable. Under the assumption of limited entry budget, SUs can directly decide whether or not to participate in spectrum auction by comparing the possible bid with access threshold which can be applied into situations that SUs have different transmit power. Theoretical analysis and simulation results show that, compared with original myopic scheme and original genie-aided scheme, the proposed auction scheme can obtain a considerable improvement in efficiency and fairness, especially with adequate available resources.     

Cooperative spectrum sharing of multiple primary users and multiple secondary users

This paper proposes a multiple-input multiple-output (MIMO) based cooperative dynamic spectrum access (DSA) framework that enables multiple primary users (PUs) and multiple secondary users (SUs) to cooperate in spectrum sharing. By exploiting MIMO in cooperative DSA, SUs can relay the primary traffic and send their own data at the same time, which greatly improves the performance of both PUs and SUs when compared to the non-MIMO time-division spectrum sharing schemes. Especially, we focus on the relay selection optimization problem among multiple PUs and multiple SUs. The network-wide cooperation and competition are formulated as a bargaining game, and an algorithm is developed to derive the optimal PU-SU relay assignment and resource allocation. Evaluation results show that both primary and secondary users achieve significant utility gains with the proposed framework, which gives all of them incentive for cooperation.     

A unified spectrum sensing and throughput analysis model in cognitive radio networks

Cognitive radio (CR) is a newly developed technology for increasing spectral efficiency in wireless communication systems. In the CR networks, there exist two traditional spectrum‐sharing technologies called spectrum overlay and spectrum underlay. A new hybrid overlay/underlay paradigm has also been discussed in the literature. In this work, we create a unified spectrum sensing and throughput analysis model, which is suitable for overlay, underlay, and hybrid overlay/underlay paradigms in the CR networks. In the proposed model, the energy detection scheme is employed for the spectrum sensing in the network in which the co‐channel interference is present among primary users and secondary users (SUs). The SUs' throughput in the proposed CR system model is then analyzed. The simulations are also carried out for demonstrating the performance of overlay, underlay, and hybrid overlay/underlay paradigms. Copyright © 2013 John Wiley & Sons, Ltd.     

Joint power allocation and beamforming with users selection for cognitive radio networks via discrete stochastic optimization

Cognitive radio is a promising technique to dynamic utilize the spectrum resource and improve spectrum efficiency. In this paper, we study the problem of mutual interference cancellation among secondary users (SUs) and interference control to primary users (PUs) in spectrum sharing underlay cognitive radio networks. Multiple antennas are used at the secondary base station to form multiple beams towards individual SUs, and a set of SUs are selected to adapt to the beams. For the interference control to PUs, we study power allocation among SUs to guarantee the interference to PUs below a tolerable level while maximizing SUs?? QoS. Based on these conditions, the problem of joint power allocation and beamforming with SUs selection is studied. Specifically, we emphasize on the condition of imperfect channel sensing due to hardware limitation, short sensing time and network connectivity issues, which means that only the noisy estimate of channel information for SUs can be obtained. We formulate the optimization problem to maximize the sum rate as a discrete stochastic optimization problem, then an efficient algorithm based on a discrete stochastic optimization method is proposed to solve the joint power allocation and beamforming with SUs selection problem. We verify that the proposed algorithm has fast convergence rate, low computation complexity and good tracking capability in time-varying radio environment. Finally, extensive simulation results are presented to demonstrate the performance of the proposed scheme.     

Dynamic spectrum sharing in cognitive radio networks using truthful mechanisms and virtual currency

In cognitive radio networks, there are scenarios where secondary users (SUs) utilize opportunistically the spectrum originally allocated to primary users (PUs). The spectrum resources available to SUs fluctuates over time due to PUs activity, SUs mobility and competition between SUs. In order to utilize these resources efficiently spectrum sharing techniques need to be implemented. In this paper we present an approach based on game-theoretical mechanism design for dynamic spectrum sharing. Each time a channel is not been used by any PU, it is allocated to SUs by a central spectrum manager based on the valuations of the channel reported by all SUs willing to use it. When an SU detects a free channel, it estimates its capacity according to local information and sends the valuation of it to the spectrum manager. The manager calculates a conflict-free allocation by implementing a truthful mechanism. The SUs have to pay for the allocation an amount which depends on the set of valuations. The objective is not to trade with the spectrum, but to share it according to certain criteria. For this, a virtual currency is defined and therefore monetary payments are not necessary. The spectrum manager records the credit of each SU and redistributes the payments to them after each spectrum allocation. The mechanism restricts the chances of each SU to be granted the channel depending on its credit availability. This credit restriction provides an incentive to SUs to behave as benefit maximizers. If the mechanism is truthful, their best strategy is to communicate the true valuation of the channel to the manager, what makes possible to implement the desired spectrum sharing criteria. We propose and evaluate an implementation of this idea by using two simple mechanisms which are proved to be truthful, and that are tractable and approximately efficient. We show the flexibility of these approach by illustrating how these mechanisms can be modified to achieve different sharing objectives which are trade-offs between efficiency and fairness. We also investigate how the credit restriction and redistribution affects the truthfulness of these mechanisms.     

Efficient error detection in soft data fusion for cooperative spectrum sensing

The primary objective of cooperative spectrum sensing (CSS) is to determine whether a particular spectrum is occupied by a licensed user or not, so that unlicensed users called secondary users (SUs) can utilize that spectrum, if it is not occupied. For CSS, all SUs report their sensing information through reporting channel to the central base station called fusion center (FC). During transmission, some of the SUs are subjected to fading and shadowing, due to which the overall performance of CSS is degraded. We have proposed an algorithm which uses error detection technique on sensing measurement of all SUs. Each SU is required to re-transmit the sensing data to the FC, if error is detected on it. Our proposed algorithm combines the sensing measurement of limited number of SUs. Using Proposed algorithm, we have achieved the improved probability of detection (PD) and throughput. The simulation results compare the proposed algorithm with conventional scheme.     

Prediction‐based MAC‐layer sensing in cognitive radio networks

We consider a cognitive radio network which coexists with multiple primary users (PUs) and secondary users (SUs) transmit over time‐varying channels. In this scenario, one problem of the existing work is the poor performances of throughput and fairness due to variances of SUs' channel conditions and PUs' traffic patterns. To solve this problem, we propose a novel prediction‐based MAC‐layer sensing algorithm. In the proposed algorithm, the SUs' channel quality information and the probability of the licensed channel being idle are predicted. Through the earlier predicted information, we schedule the SUs to sense and transmit on different licensed channels. Specifically, multiple significant factors, including network throughput and fairness, are jointly considered in the proposed algorithm. Then, we formulate the prediction‐based sensing scheduling problem as an optimization problem and solve it with the Hungarian algorithm in polynomial time. Simulation results show that the proposed prediction‐based sensing scheduling algorithm could achieve a good tradeoff between network throughput and fairness among SUs. Copyright © 2014 John Wiley & Sons, Ltd.