低复杂度室内超密集可见光通信网络频谱资源分配算法

针对室内超密集可见光通信网络多小区资源分配算法中存在的吞吐性能与算法复杂度之间的矛盾,提出一种吞吐性能近似最优且复杂度低的资源分配算法。通过建立各小区资源分配问题的最优化模型,得出该问题是凸优化问题的结论;在合理近似处理后,推导了终端资源分配归一化比例因子的解析式,在此基础上提出一种资源分配算法;算法复杂度分析表明本文算法具有多项式复杂度,低于经典最优化内点法。仿真表明,相比于请求速率占比分配法,本文算法的吞吐量性能提升了57%,服务质量满意度性能提升了67%。     

TCP在光突发交换试验网络中性能的实验研究

在光突发交换试验网络中对TCP的性能进行了实验研究.首先研究了丢包对网络性能的影响,表明突发包丢失将导致TCP吞吐量的快速下降,并且丢包率越高,可用的TCP传输带宽越窄.然后分别详细研究了OBS试验网络中的延时损伤和报文段关联增益,以及对TCP传输性能的影响.最后综合考虑这些因素.实验结果表明,在本OBS试验网络中存在一个不随丢包率变化的最优化突发包组装时间250us使得可用TCP带宽最大化.为了最大限度地提高TCP的传输性能,组包时间应该选择等于或稍大于这个最优值.     

基于纠缠交换的树形量子局域网传输模型

本文提出了一种基于纠缠交换的量子局域网络传输模型,结合传统通信网络的树形局域网模型,使用经典通信协议,并根据以太网IEEE802.3一般帧格式设置一种帧格式用来承载建立量子信道的控制信息。用户间依靠该局域网模型通过纠缠交换协议以及经典信道辅助传递控制信息来控制用户间量子信道的建立与释放,实现端到端量子信息的传输。对该网络的吞吐率进行分析表明,当节点时延为1μs,将传输过程中理想条件下的概率PSW、Pbel、Ptes设为0.8,则吞吐率随着网络中纠缠粒子生成概率Pbor增加而显著增加,最大为550kb/s;将传输过程中理想条件下的概率Pbor、Pbel、Ptes设为0.8,吞吐率也随着网络中纠缠交换成功概率PSW的增加而增加,最大为450kb/s;同时当节点时延每增加1μs,相同条件下的网络吞吐率将下降50%。因此,在将来量子局域网设置中应尽量提升纠缠交换过程的成功率,同时优化信道,降低经典信道的时延以提高量子局域网中的的系统吞吐率。     

噪声环境下基于最大后验非线性变换的隐马尔可夫模型自适应算法

由于训练环境和识别环境的失配,识别系统的性能会严重下降。为此,提出了基于最大后验概率非线性变换的环境自适应算法,可以减小由于环境的失配所引起的系统性能的下降。在本算法中,利用分段线性回归近似非线性变换将训练环境下隐马尔可夫模型(HMM)的均值向量变换到识别环境,减小环境的失配,变换参数的估计采用了最大后验概率估计(MAP)。数字语音识别实验证明:该环境自适应算法的识别性能优于MLST,MAPLR和MLLR等算法。     

WSANs中基于蜂巢结构的移动容错恢复算法

旨在研究无线传感器与执行器网络(WSANs)中节点失效情况下恢复执行器(actor)节点服务的算法. 首先说明了WSANs中的实时覆盖模型, 证明WSANs覆盖恢复问题是NP难问题, 给出了近似求解方案. 在此基础上, 提出了一种基于六边形蜂巢结构的移动容错算法HMFR用于恢复失效actor节点, HMFR 算法在限制网络初始部署的条件下拥有很好的性能. 通过实验与现有的恢复算法进行比较, 发现HMFR算法在actor覆盖sensor节点数和移动距离方面有更好的性能.     

基于改进量子粒子群和主动PI模型的自适应无线传感器网络拥塞控制算法设计

针对无线传感器网络中路由节点需要转发大量数据导致网络拥塞,从而引起节点丢包率高和网络吞吐率过低的问题,提出了一种基于主动PI模型和改进量子粒子群优化算法的拥塞控制方法;首先定义了丢包率和队列长度计算方式;设计了改进的PI主动队列管理模型,然后,为了改进PI模型的控制效果,采用改进的量子粒子群算法对PI主动队列管理模型中的比例系数和积分系数kP和kI进行参数优化,从而得到能实现WSN自适应控制的主动PI控制模型;最后,对基于量子粒子群算法和PI主动队列模型的网络拥塞算法进行了描述和说明;仿真实验表明:文中提出的拥戴控制方法能有效实现WSN拥塞控制,是一种适用于WSN的有效拥塞控制方法,与其它方法相比,具有较短的平均队列长度和较大的网络吞吐率,具有很强的可行性。     

执行依赖启发式动态规划PMSM控制策略研究

永磁同步电机算法中的PID参数固定不变,自适应能力较差。近似动态规划具有自学习能力,能动态调整参数,增强系统的自适应能力。本文将近似动态规划中的代表性算法——执行依赖启发式动态规划与类PI网络相结合,根据环境反馈对各个策略进行评价,选取评价最优的策略在线调整动作网络参数,对永磁同步电机进行自适应最优控制。仿真结果表明,执行依赖启发式动态规划加PI控制算法在系统速度变化和负载变化过程中,动态效果和稳态精度优于传统PID控制。     

基于马尔可夫决策模型的异构无线网络切换选择算法

针对异构环境下不同业务类型用户对于接入网络的不同服务质量(quality of service,Qo S)需求,该文提出了一种基于马尔可夫决策模型的切换选择算法.建立基于软件定义网络(software defined network,SDN)的异构无线网络架构,以实现对异构网络的通透控制.利用马尔可夫过程预测下一时刻的网络状态以得到采取动作后的一次回报,依据网络的不同状态属性针对实时用户和非实时用户分别构建立即回报函数,并采用层次分析法确定属性权重;基于状态动作对构建期望回报函数,采用逐次逼近的迭代方式得到使长期期望回报最大的切换策略.仿真结果表明,该方法针对不同业务类型用户均能选取最优切换策略,同时降低阻塞率,提高了用户的Qo S和无线网络的资源利用率.     

基于DIJKSTRA的无线传感器网络分簇路由算法

为加快无线传感器网络最优路径搜索速度、减少路径寻优能量消耗和延长网络寿命,提出了基于改进的DIJKSTRA算法的无线传感器网络分簇路由算法;运用DIJKSTRA算法在无线传感器网络内以多跳接力的方式来搜寻从源节点到目的节点的最短路径;结合能耗优化策略,避免网络能耗热点问题,实现网络能耗均衡;通过与基于蚁群算法的路由算法对比分析,基于Dijkstra的网络分簇路由算法能优化网络分簇并建立较优传输路径,其快速收敛性能减缓了网络中簇头节点的能耗,延长了网络寿命,提高了网络鲁棒性。     

基于阶梯网络与交叉融合的端到端图像去雾

针对卷积神经网络类图像去雾方法存在的细节丢失、颜色失真、去雾不彻底等问题,提出一种基于阶梯网络与注意力交叉融合的端到端图像去雾算法.整体网络模型包含特征提取、特征融合、图像重建三个模块,其中特征提取包括有雾图像细节和轮廓特征的提取,由阶梯网络的不同阶梯层提取实现;特征融合模块以注意力机制的交叉融合实现,并结合自适应残差...     

Traffic-Aware Green Cognitive Radio

Cognitive Radio Network (CRN) has emerged as an effective solution to the spectrum under-utilization problem, by providing secondary users (SUs) an opportunistic access to the unoccupied frequency bands of primary users (PUs). Most of the current research on CRN are based on the assumption that the SU always has a large amount of data to transmit. This leads to the objective of SU throughput maximization with a constraint on the allowable interference to the PU. However, in many of the practical scenarios, the data arrival process of the SU closely follows an ON–OFF traffic model, and thus the usual throughput optimization framework may no longer be suitable. In this paper, we propose an intelligent data scheduling strategy which minimizes the average transmission power of the SU while maintaining the transmission delay to be sufficiently small. The data scheduling problem has been formulated as a finite horizon Markov Decision Process (MDP) with an appropriate cost function. Dynamic programming approach has been adopted to arrive at an optimal solution. Our findings show that the average transmitted power for our proposed approach can be as small as 36.5% of the power required for usual throughput maximization technique with insignificant increase in average delay.     

Performance evaluation of improved double-threshold energy detector over Rayleigh-faded sensing and imperfect reporting channels

Cognitive radio (CR) has been viewed as a promising solution to spectrum scarcity. In order to design a reliable CR system, many improvements have been proposed to enhance spectrum sensing performance of secondary users (SUs) in a CR network (CRN). Sensing reliability and transmission throughput of SUs are two important performance criteria, which should be optimized to enhance signal protection of primary user (PU) as well as spectrum utilization rate. In this paper, we consider Rayleigh-faded sensing channels and SUs use improved energy detector (IED) to make their local decisions. The final decision is made in a fusion center (FC) through the cooperative spectrum sensing (CSS) scheme with erroneous reporting channels. We show that the improved double-threshold energy detector (IDED) outperforms the conventional energy detector (CED) in terms of the total error rate. Furthermore, we evaluate the transmission throughput of the CRN through various ED schemes with detection constraints over both perfect and imperfect reporting channels. We show that the IDED has the highest achievable throughput among different ED schemes over imperfect reporting channels.     

Optimal sensing energy and transmission interval in hybrid overlay–underlay cognitive radio networks with energy harvesting

In this work, we present a new concept called “transmission interval” in a hybrid overlay/underlay cognitive radio network. A transmission interval consists of a sequence of time slots during which the secondary user (SU) transmits its data using the optimal mode based on its current state. After the transmission interval ends, the SU has to choose between staying idle for a single time slot to save energy for future possible transmission, transmitting using the underlay mode without sensing to optimize the usage of the limited amount of available energy, or sensing the channel and transmitting using either overlay or underlay mode depending on the primary user (PU) state. The energy harvesting technology is also considered in the presence of multiple PUs and multiple SUs. For the SU network, a sequential decision problem is formulated using the mixed observable Markov decision process to determine the optimal sensing energy and the optimal transmission interval length that maximize the SU network throughput and minimize both the consumed energy and the interference to the PUs. Numerical results show that applying the transmission interval concept increases the SU network throughput and decreases the interference to the PUs compared to conventional models. Moreover, adding the action of underlay transmission without sensing increases the SU network throughput.     

Scheduling Strategy Design Framework for Cyber–Physical System with Non-Negligible Propagation Delay

Cyber–physical systems (CPS) have been widely employed as wireless control networks. There is a special type of CPS which is developed from the wireless networked control systems (WNCS). They usually include two communication links: Uplink transmission and downlink transmission. Those two links form a closed-loop. When such CPS are deployed for time-sensitive applications such as remote control, the uplink and downlink propagation delay are non-negligible. However, existing studies on CPS/WNCS usually ignore the propagation delay of the uplink and downlink channels. In order to achieve the best balance between uplink and downlink transmissions under such circumstances, we propose a heuristic framework to obtain the optimal scheduling strategy that can minimize the long-term average control cost. We model the optimization problem as a Markov decision process (MDP), and then give the sufficient conditions for the existence of the optimal scheduling strategy. We propose the semi-predictive framework to eliminate the impact of the coupling characteristic between the uplink and downlink data packets. Then we obtain the lookup table-based optimal offline strategy and the neural network-based suboptimal online strategy. Numerical simulation shows that the scheduling strategies obtained by this framework can bring significant performance improvements over the existing strategies.     

Scheduling for end-to-end performance in multi-rate CDMA/HDR systems

An interesting feature of the third generation (3G) cellular networks is their ability to support multiple data rates for data services. Thus, it is important to understand the end-to-end delay and throughput performance over these multi-rate cellular systems. In this work, we consider a multi-rate system such as High Data Rate (HDR) and represent the possible data rates using a Markov chain. Using a M/G/1 queuing model, we calculate the expected data rate and the corresponding link layer delay for each of the initial states. We study the effects of various link layer scheduling techniques for the radio link protocol (RLP) frames destined for multiple user. Though the commonly used link layer scheduling techniques work well, they do not provide the best performance for when end-to-end semantics need to be maintained. In this regard, we propose a scheduling algorithm that aids the transport protocol, for example TCP. Our research shows that the proposed remaining frames first scheduler enhances the transport layer performance compared to the commonly used “best channel scheduler”. Comparative results, based on simulations, are shown with respect to radius of cell, distance of user, number of users, and varying bit error rate.     

Reinforcement Learning-Based Multihop Relaying: A Decentralized Q-Learning Approach

Conventional optimization-based relay selection for multihop networks cannot resolve the conflict between performance and cost. The optimal selection policy is centralized and requires local channel state information (CSI) of all hops, leading to high computational complexity and signaling overhead. Other optimization-based decentralized policies cause non-negligible performance loss. In this paper, we exploit the benefits of reinforcement learning in relay selection for multihop clustered networks and aim to achieve high performance with limited costs. Multihop relay selection problem is modeled as Markov decision process (MDP) and solved by a decentralized Q-learning scheme with rectified update function. Simulation results show that this scheme achieves near-optimal average end-to-end (E2E) rate. Cost analysis reveals that it also reduces computation complexity and signaling overhead compared with the optimal scheme.     

A spectrum handoff scheme for optimal network selection in Cognitive Radio vehicular networks: A game theoretic auction theory approach

The recent strides in vehicular networks have emerged as a convergence of multi radio access networks having different user preferences, multiple application requirements and multiple device types. In future Cognitive Radio (CR) vehicular networks deployment, multiple radio access networks may coexist in the overlapping areas having different characteristics in terms of multiple attributes. Hence, it becomes a challenge for CR vehicular node to select the optimal network for the spectrum handoff decision. A game theoretic auction theory approach is interdisciplinary effective approach to select the optimal network for spectrum handoff. The competition between different CR vehicular node and access networks can be formulated as multi-bidder bidding to provide its services to CR vehicular node. The game theory is the branch of applied mathematics which make intelligent decision to select the optimal alternative from predetermined alternatives. Hence, this paper investigates a spectrum handoff scheme for optimal network selection using game theoretic auction theory approach in CR vehicular networks. The paper has also proposed a new cost function based multiple attribute decision making method which outperforms other existing methods. Numerical results revel that the proposed scheme is effective for spectrum handoff for optimal network selection among multiple available networks.     

Ring-based PON supporting multiple optical private networks using OCDMA technique

This work is a novel attempt to provide local communications for multiple optical private networks (PNs) within ring-based passive optical network (PON). In order to improve network throughput performance, PNs traffic is decentralized from PON traffic let no extra traffic management into optical line terminal (OLT). To achieve multiple secure optical private networking over ring-based PON layout, optical code division multiple access (OCDMA) technique is applied. This technique leads to interconnect optical network units (ONUs) in the same PN sharing the same codeword while other PNs benefiting from different codewords. This scheme can be used in access networks to establish discrete communications between different sites in an enterprise or a university campus or even a residential accommodation. The proposed network architecture is then set up and its bit error rate performance is experimentally demonstrated. Finally, the network scalability and throughput performance of the proposed scheme are analyzed.