基于用户体验的VLC/WiFi异构网络切换算法

针对可见光通信(VLC)与WiFi组成的异构网络中的网络选择问题,现有工作大多仅关注实际物理网络指标的优化,如信干噪比(SINR)和吞吐量等,而忽视了终端用户的异构需求和服务体验。基于此,该文提出了一种以用户体验质量(QoE)为优化目标的网络切换算法。该算法首先基于平均主观评分(MOS)机制建立离散化的QoE等级模型,然后根据不同网络对应的实际QoE水平给出了相应网络切换准则,当且仅当新网络使得用户QoE等级提升时才进行网络切换,从而在一定程度减小盲目追求容量优化引起的网络切换开销。仿真结果表明,该算法相比现有算法能够在保证QoE水平的前提下显著减小整体的网络切换次数,从而降低网络开销,提升网络性能。      

基于强化学习的异构网络垂直切换方法

网络切换技术不仅可以保证用户的网络连接,同时能够以较强的信号传输网络数据。网络切换技术的性能对网络服务质量(quality of service, QoS)的影响至关重要。然而,现有的切换算法多数存在严重的乒乓效应,这不仅造成网络资源的严重浪费,还会损害用户的QoS。为此,提出了一种基于强化学习的异构网络垂直切换方案,主要从触发切换、网络选择及判决切换等方面进行优化。在触发切换时将垂直切换考虑成必要切换和择优切换,通过Q-Learning（QL）算法在选择网络时优化垂直切换;然后以QoS为条件,在判决切换时加入驻留定时器,从多角度减少用户切换次数,降低乒乓效应对异构网络垂直切换的影响。仿真结果表明,基于强化学习的异构网络垂直切换方法可以在保证QoS的条件下有效减少用户切换网络的次数,短时间内频繁切换的情况也有所改善,降低了乒乓效应的影响。     

卫星动态链路上的可预知确定性流量调度

随着卫星网络的高速发展，卫星之间的动态链路上也出现了确定性传输需求。结合卫星链路的可预知动态特性，提出了运动平台上的时间敏感网络(Time Sensitive Network, TSN)调度机制，研究适用于链路变化的动态调度算法。在基于信用整形(Credit Based Shaper, CBS)算法模型基础上，研究分析CBS在动态变化网络场景中的时延特性。结合所建立的CBS分析模型，研究了带宽占比对于不同等级链路所起的作用。同时根据分析结果设计带宽调整方案，改进CBS算法，并提出链路切换保护机制，用以避免数据在链路状态切换过程中出现丢失。提出的动态方案不仅能提高卫星网络的数据转发确定性，并且能够根据不同数据帧的时延要求灵活分配带宽。     

基于QoS决策的异构VLC/WiFi网络联合收发端需求的接入点选择

针对可见光通信(VLC)下行链路存在多接入点(AP)重叠覆盖、VLC AP光路易被遮挡和干扰的问题,提出了一种基于服务质量(QoS)决策的异构VLC/无线保真(WiFi)网络联合收发端需求的接入点选择算法。该算法在收端为用户建立面向QoS的加权信息矩阵,为用户偏好提供高信干噪比、高接收功率、低负载、低历史中断概率的VLC AP;在发端为VLC AP偏好选择抗干扰性强的用户服务。基于引力理论改进收发端选择权重值,优化用户的AP接入决策。仿真结果表明:该算法在方形均匀布局和混合圆方形布局中比现有算法的吞吐量分别增大了77.3%和11.4%,业务公平性分别减小了53.1%和41.1%。     

LTE-Advanced异构中继网络研究

文中提出一种LTE-Advanced异构中继网络架构,并设计合理的中继接入/退出机制,确保用户通过中继传输获得通信质量的改善;提出了中继传输用户的切换机制,有效避免切换过程中用户数据在回程链路上重复转发的问题,大大减少了中继切换过程中的网络延时;最后,结合Full Buffer业务模型对该中继系统用Opnet进行仿真分析,证明在加入中继并合理设置中继接入控制阈值后有助于充分发挥中继系统的作用,使系统性能得到极大改善。     

车联网环境下的4G和DSRC异构网络切换机制研究

面向车联网的实际应用需求,构建了一种高速公路环境下的基于4G和DSRC的异构网络通信场景。针对车辆在该场景下可能会在不同网络信道间频繁发生垂直切换导致较高的传输时延和丢包率的问题,在分析研究基于4G与DSRC的高速公路异构网络切换过程的基础上,引入网络跳数、连接次数和行驶轨迹作为切换判决条件,基于TREBOL路由协议设计了一种异构网络垂直切换算法。最后利用Veins仿真平台对提出的异构网络垂直切换算法的切换性能进行对比测试分析。测试结果表明,与基于RSSI的异构网络垂直切换方法相比,所提切换算法能增大车辆与DSRC网络连接的持续时间,并有效减少"乒乓效应"。     

一种基于异构网络的自适应垂直切换算法

提出一种基于WSN与WLAN异构网络环境下无缝移动通信的垂直切换算法。该算法针对日益普及的传感器网络自身特点与室内应急通信应用的特殊性,提出网络选取优先级参数。针对传感器网络在应急通信等特殊场合的应用．制定了专门的切换判决策略。实验分析表明,该算法可以一定程度地提高通信网络对现场场景变化的适应程度,满足室内环境下异构网络通信的需求。     

一种基于IEEE802.16无线MESH网络集中调度机制的时隙分配方法

在基于IEEE 802.16的无线Mesht网络中,时隙分配算法对网络性能有重要影响.针对现有时隙分配算法只研究上行链路且时隙分配的结果导致中继节点在转发数据时频繁在相邻时隙间进行收发切换的问题,提出了一种上下行链路通用的时隙分配方法,对于上行链路,跳数较小节点的数据优先传输,而对于下行链路,跳数较大节点的数据优先传输,在传输路径上采用逐跳传输的策略.仿真结果证明了该算法的有效性.     

室内可见光通信异构组网关键技术研究进展

移动通信网络发展迅速，移动宽带用户以及新兴信息业务与日俱增，可见光通信(VLC)成为室内异构组网的重要补充手段之一，其相关技术亟待突破。首先，围绕研究背景与意义进行简要概述，其次，深入讨论并分析了室内可见光通信异构组网架构设计、垂直切换方案设计和多用户接入机制优化三个关键技术的主要进展。最后，从实际应用需求角度出发，对三者进行总结与展望。     

基于灰预测模型的蓝牙网络切换算法

蓝牙网络接入点切换问题是构建蓝牙网络必须考虑的关键问题,而蓝牙特点给接入点切换的实现带来了极大挑战。提出了一种基于灰预测模型的蓝牙网络切换算法,该算法对链路质量信息的数量和分布特征没有苛刻要求,仅用少量样本数据就可以预测出下一时间链路质量,当预测值低于阈值时,接入点搜索网络拓扑结构找出最佳接入点集合,并实时监测切换终端与集合中接入点之间链路质量,预测出下一时刻链路质量,当切换终端链路质量预测值低于另一阈值时,接入点选择最佳接入集合中预测值最佳的接入点完成切换。该算法预测过程可以动态地调整参数,实现自适应预测,提高了预测的精度。仿真结果表明,使用该算法减少了终端切换时延以及误切换概率。     

Transmit beamforming and power control for cellular wirelesssystems

Joint power control and beamforming schemes are proposed for cellular systems where adaptive arrays are used only at base stations. In the uplink, mobile power and receiver diversity combining vectors at the base stations are calculated jointly. The mobile transmitted power is minimized, while the signal-to-interference-and-noise ratio (SINR) at each link is maintained above a threshold. A transmit diversity scheme for the downlink is also proposed where the transmit weight vectors and downlink power allocations are jointly calculated such that the SINR at each mobile is above a target value. The proposed algorithm achieves a feasible solution for the downlink if there is one and minimizes the total transmitted power in the network. In a reciprocal network it can be implemented in a decentralized system, and it does not require global channel response measurements. In a nonreciprocal network, where the uplink and downlink channel responses are different, the proposed transmit beamforming algorithm needs to be implemented in a centralized system, and it requires a knowledge of the downlink channel responses. The performances of these algorithms are compared with previously proposed algorithms through numerical studies     

基于可见光通信的室内光电混合网络组网关键技术

针对可见光通信在室内的应用,分析了采用可见光与无线射频构成光电混合网络的可行性,提出了一种可行的网络架构;指出了该网络架构组网中的关键技术问题,并针对网络中的光源布局、多址接入、小区切换、异构网络融合等给出了技术路线。     

An Effective Channel Control Algorithm for Integrated IPTV Services Over DOCSIS CATV Networks

In this paper, we present an effective IPTV channel control algorithm that guarantees a seamless channel change and provides a fast data communication service simultaneously via controlling the usage levels of uplink and downlink resources over DOCSIS CATV networks. The proposed algorithm treats the number of adjacent channels and the assigned grant interval as control variables to reduce the network delay part of channel zapping time and improve the quality-of-service of the uplink and downlink. Finally, the experimental results are provided to show the performance of the proposed algorithm.     

A bandwidth allocation algorithm with channel quality and QoS aware for IEEE 802.16 base stations

Subscriber stations located in different places encounter various interferences in an IEEE 802.16 network, resulting in that their communication channels experience varying channel conditions. Thus, an excellent bandwidth allocation algorithm should not only satisfy various QoS required by heterogeneous traffic, but also consider the channel quality to maximize bandwidth utilization. In this paper, a bandwidth allocation algorithm with channel quality awareness and QoS guarantee, called CQQ, is proposed. CQQ not only satisfies each connection's QoS requirement, but also dynamically adjusts the downlink/uplink bandwidth to match current downlink/uplink traffic ratio. CQQ allocates more bandwidth to the connections having better channel quality by applying weighted fair queuing strategy to raise the bandwidth usage. CQQ provides lower delay violation ratio and higher goodput than the previous algorithms, as observed from the simulation results. Copyright © 2012 John Wiley & Sons, Ltd.     

Performance of a wavelength hopping MC-VPPM scheme for vehicle-to-infrastructure(V2I) VLC

Multi-coded variable PPM (MC-VPPM) is a modulation technique for visible light communication (VLC), providing high data rate transmission and processing gain improvement at the receiver side. This paper proposes wavelength hopping MC-VPPM supporting multi-user communications in the vehicle-to-infrastructure (V2I) downlink and uplink scenarios of a metropolitan intelligent transportation system and analyzes its performance. By defining outdoor daylight noise and analyzing the sub-band selection effect for wavelength hopping, the communication capability of multi-user V2I downlink and uplink VLC at the daytime was investigated. The simulation results provide the required VLC transmission power for reliable data communication under the predescribed V2I scenario.     

Bandwidth utilization efficiency enhancement for OFDM‐based WSN

Normally IEEE 802.16 (WiMAX) is used for mainly downlink traffic applications. However in the upper tier of 2‐tier (WiMAX‐WiFi) wireless sensor network, the uplink bandwidth faces bottlenecks for high throughput. In this paper, a solution has been proposed for this limitation of uplink bandwidth allocation through the use of queuing theoretic performance modeling. A Markov‐modulated Poisson process traffic model has been formed for orthogonal frequency division multiple access‐based transmission along with discrete time Markov chain system model for queuing. A downlink traffic pattern has been defined for wireless sensor network nodes. Analytical methods are used to estimate the performance parameters like throughput, delay, and probability of packet drop for resource allocation. An algorithm is formulated to find out minimum resource requirement for downlink and to transfer rest of the resources to uplink bandwidth allocation, for throughput enhancement. Uplink frame utilization is determined through another discrete time Markov chain model for adaptive triggering between the proposed maximum and the normal downlink to uplink ratio operations, for efficient distribution of bandwidth resources. Algorithm and simulation results prove outstanding improvement in the uplink throughput around 50%, without degrading the downlink throughput.     

3G系统动态资源分配下的QoS性能分析

动态资源分配（DCA）是解决无线信道紧张的最有限途径之一,文中研究CDMA系统动态资源分配下的QoS性能。从CDMA自干扰受限系统出发,分析信道优先级的排队算法以及上下行信号的接入控制算法,从而进一步提出接入用户的信道资源的优化调整方案,来提高系统服务质量（QoS）等级。文中进行了系统的仿真研究,从系统模型分析和仿真研究结果表明,文中给出的算法使系统掉话率下降,系统的性能得到改善。     

分布式目标下行波束形成

研究了智能天线在下行链路的应用。利用上行链路接收数据,建立分布式目标信道模型。估计移动用户的发射角、扩展角度和发射功率以及下行信道相关矩阵,确定下行链路的波束形成方法。即对频分复用（FDD）系统进行上行相关矩阵变换,构建下行信道协方差矩阵,利用最大合并比准则获得下行链路权值。计算机仿真证实了这种方法的有效性。     

TD网络数据业务AUE和TPE算法融合研究与验证

基于华为TD网络数据业务AUE和TPE算法,在邯郸现网规模实验开启自适应上行增强(AUE)功能及TCP性能增强(TPE)功能,按照室内定点、室外定点、室外路测三种测试方式,分为不同的覆盖场景进行上下行速率性能对比测试;同时针对这两种算法进行融合性验证,找出算法单独开启与同时开启时在不同场景下的性能提升幅度规律,并总结出后续建议,为其他RNC开启AUE和TPE技术提供参考,以全面有效地提升现网用户的数据业务感知。