基于跨层信息的信息中心网络缓存算法研究

为了提高信息中心网络(ICN)的性能,研究了ICN缓存算法的优化设计。针对现有的缓存算法冗余度高、命中率较低的问题,提出了一种基于跨层设计的全新的ICN缓存算法CRCache,该算法根据网络层信息与应用层信息之间的相关性,有选择地将内容缓存在与其相匹配的路由器上。通过这种跨层的缓存设计,CRCache缓存算法能够消除网络中不必要的冗余内容、提高缓存命中率、减少网络流量。设计并实现了基于真实用户访问日志与真实中国网络拓扑的大规模仿真实验,通过实验证明,该算法在缓存命中率、网络流量减少率、服务器负载等指标上均优于现有的缓存算法。     

跨层设计的AODV路由设计与仿真

随着网络对无线、移动和全IP化的需求,Ad Hoc网络成为人们日常通信活动中所普及的网络。本文借助跨层设计理念,在AODV协议中添加跨层参数,对AODV协议路由发现过程的路由更新操作进行改进。根据Ad Hoc网络通信特点和AODV路由协议运行原理,结合跨层设计理念,设计并实现了一种基于跨层设计的AODV路由——CL-AODV。     

自组网中一种基于跨层负载感知的蚁群优化路由协议

将蚁群优化和跨层优化方法结合起来,提出了一种基于跨层负载感知和蚁群优化的路由协议(CLAOR)。协议将整个路径中各节点MAC层的总平均估计时延和节点队列缓存的占用情况结合起来,共同作为路由选择和路由调整的重要度量标准进行按需路由发现和维护,通过拥塞节点丢弃蚂蚁分组以及借助部分兼具蚂蚁功能的数据分组实现正常路由表的维护等方法,减少了控制开销,增加了算法的可扩展性,较好地解决了自组网中现有基于蚁群优化的路由协议中普遍存在的拥塞问题、捷径问题和引入的路由开销问题。仿真结果表明,CLAOR在分组成功递交率、路由开销以及端到端平均时延等方面具有优良性能,能很好地实现网络中的业务流负载均衡。     

移动自组网选播路由协议ARAD

针对移动自组(ad hoc)网的特点,提出了一种适合ad hoc移动网络的选播路由协议——ARAD协议.该协议在充分考虑选播路由特点的基础上,将ad hoc按需距离矢量路由(AODV)协议的路由发现机制与动态源路由(DSR)协议的缓存源路由技术相结合,在动态变化的ad hoc网络中实现了一种高效的选播路由.模拟结果表明,在动态的网络环境下,ARAD协议能够取得良好的性能.     

基于BP神经网络算法的多参数在线测量物联网系统

针对现有物联网在线测量系统存在的参数测量周期长,异常反馈不及时的问题,设计了基于BP神经网络算法的多参数在线测量物联网系统。该系统通过感知层、网络传输层及应用层完成参数的采集、分析处理及显示,其核心在于系统应用层加入BP神经网络算法对实时采集参数进行优化处理,来缩短参数测量时间。通过算法仿真实验可知,本文所提出的基于BP神经网络算法优化的在线测量系统,对于加快数据采集时间,提高制造过程系统反馈调节速度,有着很好的效果。     

基于广播树的移动自组织网络合作路由协议

为了激励自组网中的自私节点使其参与网络合作,提出了一种基于VCG机制计算转发节点支付的BEC协议,此协议对LDTTO协议的支付方案进行了改进,使源节点失去作弊动机.在路由发现阶段,创建以目的节点为根的广播树,将节点拓扑信息沿广播树以单播方式发送到根节点,降低了网络控制负载开销和路由发现延迟.BEC协议引入了容错机制,通过在混杂模式下对父节点的监听来确保数据的正确传输.在BEC协议中,高度为O(lgn)的广播树,路由发现过程的消息负载为O(nlogn),低于LOTID协议的O(n2).仿真实验结果表明,与现有的几种重要协议相比,BEC协议具有更低的消息负载和网络延迟.     

基于节能的无线传感器网络跨层优化设计研究

无线传感器网络的特点决定了解决能量问题是WSN优化的关键。传统分层协议优化模式下层与层间信息难于共享,增加了通信开销以及信息冗余,没有充分利用网络资源。考虑到链路质量、剩余能量会在很大程度上影响路由方式,拓扑结构会影响信道的使用方式,本文在LEACH-L协议的基础上进行设计优化,提出一种基于跨层优化思想的分簇路由协议Leach-CLO。     

一种基于跨层设计的UWB无线Ad hoc网络MAC层协议

运用跨层设计思想和方法研究了超宽带（UWB）无线自组织（Ad hoc）网络媒体访问控制（MAC）层协议的设计,提出了一种可利用UWB技术定位性好的优势和实现跨层协作的MAC层协议。该协议通过物理层、MAC层和网络层之间的跨层协作来解决自组织网络单信道无线传输过程中的隐藏终端和暴露终端问题以及网络能量节约问题,以提高网络的性能。仿真实验表明,该协议在平均吞吐量、平均端到端时延以及能量开销等性能指标上,均优于已有的IEEE802．11和MACA—BIMAC协议。该协议的设计思想和方法为下一步的UWB无线Ad hoc网络研究奠定了理论和实验基础。     

多节点下Ad Hoc网络路由协议模拟

在windows环境下,主要介绍基于NS-2网络模拟器的Ad Hoc网络路由协议的模拟实现。首先对协议模拟软件NS-2进行介绍,接着对现有的Ad Hoc网络的路由协议进行说明,最后对NS-2的方法及工作流程进行解释,并在此基础上针对不同网络模式、不同移动场景和不同的传输负载对表驱动路由协议,按需驱动路由协议进行模拟。结果表明对实时性要求较高的场合适宜使用表驱动协议,而对数据质量和网络带宽有严格要求的场合应使用按需驱动路由协议。     

一种融合MAC层拥塞通告的混合网络TCP协议

在研究无线网络媒体接入控制(MAC)层拥塞测度的基础上,提出了一种跨层的显式拥塞通告(ECN)机制,即:当数据包中记录的请求发送(RTS)次数超过给定阈值时,通过ECN向传输控制协议(TCP)源端发送拥塞通告,从而启动TCP拥塞控制.这种跨层设计是对有线网络中基于主动队列管理(AQM)的拥塞控制的有效补充,由此可以得到一种与已有的协议无缝连接的混合网络TCP模型.通过在网络模拟器NS2中构造多流无线局域网和多跳无线/有线混合网络,对所提出的方法进行了仿真,实验结果说明该方法能够提高混合网络的性能,并且具备良好的扩展性.     

Cross-layer design for tree-type routing and level-based centralised scheduling in IEEE 802.16 based wireless mesh networks

Infrastructure wireless mesh network, also named as mesh router, is one key topology for the next generation wireless networking. In this work, the performance optimisation for the infrastructure wireless mesh network is presented and the sub-optimum solution mechanism is investigated. A cross-layer design for tree-type routing, level-based centralised scheduling and distributed power control theme is proposed as the sub-optimum solution strategy. The cross-layer design relies on the channel information and the distributed transmission power control in the physical layer, and the wireless scheduling in the medium access control (MAC) layer, as well as the routing selection mechanism in the MAC upper layer. In this work, a modified distributed power control algorithm is proposed first. In addition, a tree-type routing construction algorithm for centralised scheduling is presented to improve the network throughput by jointly considering interference and hop-count to construct the routing tree. Simulation results show that the proposed cross-layer design strategy can effectively improve the network throughput performance, decrease the power consumption and achieve better performances.     

Ad hoc网络中一种跨层流量分配算法——CLATA

针对ad hoc网络各协议层的功能都相互关联的特点,提出了一种在ad hoc网络中基于网络平均时延最小的跨层自适应流量分配算法（CLATA）。该算法将网络层自适应流量分配信息传递给媒体接入控制（MAC）层,以改进MAC层中的冲突退避算法,实现网络平均时延最小化,提高网络的利用率。仿真实验结果表明,该算法可以动态调整链路之间的流量,并具有快速的自适应性,优化网络资源的利用。     

Quality of service for multimedia traffic using cross-layer design

Quality-of-service (QoS) guarantees are critical for the transmission of multimedia traffic over mobile wireless networks. Currently, wireless networks provide QoS guarantees using the legacy layered protocol architecture where each layer provides a separate, independent solution, with its own optimised adaptation and protection mechanisms. Cross-layer design has been proposed as a methodology to extend that paradigm in wireless links where there is interdependence between the layers and hence opportunity for information sharing. Recently, cross-layer adaptation mechanisms have been proposed which attempt to solve the QoS provisioning problem. However, most of these mechanisms only use the lower (physical and data link) layers and the possibility of using higher protocol layers remains unexplored. As a result, restrictions are placed on the system which introduces functional and efficiency limitations. Here, one such limitation is highlighted, namely the inability to insert more than one class of traffic in a physical layer frame. A physical and application layer cross-layer adaptation mechanism is then proposed, which overcomes this limitation. The performance results of the scheme show that the cross-layer mechanism can be efficiently applied for the purpose of providing QoS guarantees for multimedia traffic.     

Cross-layer binding update for TCP performance enhancement over Mobile IPv6 networks

Mobility support in Internet Protocol version 6 (IPv6) is a de facto standard for allowing nodes to remain connected while moving around in the IPv6 networks. To provide better handoff latency and performance, numerous optimisation techniques were developed. Some of them are single-layer optimisation such as the hierarchical mobile IPv6 architecture, which aims at reducing the registration time in network layer. Others are cross-layer optimisation such as the fast handover mechanism, which tries to reduce the address resolution time in network layer based on the data-link layer detection. Most of these techniques are focused on data-link and network layer optimisation. The authors propose a cross-layer mechanism, which uses the standard binding update packet in mobile IPv6 handoff to enhance the overall transmission control protocol performance in transmission. The theoretical analysis and simulations show promising performance enhancements.     

基于改进粒子群算法的分布式天线系统跨层功率分配方案

结合数据链路层的队列状态信息和物理层的信道状态信息定义了系统的吞吐量系数和公平性系数,建立了分布式天线系统跨层功率分配的离散速率集优化模型。对粒子群优化算法的初始群体产生、粒子更新等步骤进行了改进,形成了改进粒子群算法(IMPSO)。利用IMPSO进行了动态功率分配和跨层优化。数值仿真结果表明,IMPSO能够取得很好的队列时延性能,选取不同的权重可对系统吞吐量性能和公平性性能产生重要影响,IMPSO获得的系统吞吐量性能和公平性性能较之功率固定分配算法(AP)均有较大幅度的提高。     

三端背靠背柔性直流输电的虚拟同步发电机控制策略及其在配电网中的应用

探讨了一种在10 kV配电网中引入三端背靠背（back-to-back,BTB）柔性直流输电（voltage source converter based HVDC,VSC-HVDC）系统的接线方案,通过VSC-HVDC的控制实现提高供电可靠性和抑制环流的目的。VSC-HVDC系统中换流站在传统控制模式下几乎没有转动惯量,难以有效地参与电网调节。为了提高电网受端系统频率的稳定性,改善系统发生故障后的运行特性,在中压（10 kV）交流配电网的背景下,提出了虚拟同步发电机（virtual synchronous generator,VSG）控制策略在三端BTB VSC-HVDC系统中的应用。首先在10 kV系统中加入三端BTB VSC-HVDC互联装置建立交直流混合配电网,建立换流站在传统控制和VSG控制下的数学模型,然后通过PSCAD/EMTDC平台进行仿真,在2种控制方式下得到系统在受到扰动和发生故障时的运行特性。结果表明,三端BTB VSC-HVDC系统受端换流器使用VSG控制能有效地参与电网调节,增加系统惯性,改善系统的暂态特性,提高电网运行的可靠性。     

Cross-Layer Measurement for the Analysis of DVB-T System Performance

This paper deals with cross-layer measurements for the efficient analysis of Digital Video Broadcasting-Terrestrial (DVB-T) system performance. In particular, valuable relationships between some cross-layer DVB-T system parameters are investigated through measurements conducted at the physical and application layers. Specifically, channel power, error vector magnitude, and modulation error ratio are assessed at the physical layer, whereas objective video quality indices are estimated at the application layer. To this aim, a number of real-life experiments are executed through a purposely developed testbed that enlists a suitable DVB-T transmission platform and real-life receivers. The ultimate goal is to highlight the importance of cross-layer measurements in the analysis and troubleshooting of DVB-T system operation and to show how to deduce from such measurements practical information and hints to be profitably used in the design and setup of any DVB-T broadcasting system.     

Provision of proportional delay differentiation in wireless LAN using a cross-layer fine-tuning scheduling scheme

The proportional delay differentiation (PDD) is one of the most well-known quality of service (QoS) models and has drawn much attention because of its 'controllable' and 'predictable' characteristics. How to provide PDD in a wireless LAN (WLAN) is addressed and a cross-layer fine-tuning scheduling (CFS) scheme with the goal to maintain PDD among all wireless stations is proposed, while improving performance in a WLAN. CFS additionally considers the time-varying channel capacity to schedule packets, finely tunes the contention window and properly arbitrates the backoff time. Also, it operates in a fully distributed manner among all stations and in a cross-layer approach in each station. The simulation results demonstrate that the CFS scheme can provide more satisfactory PDD and higher performance in a WLAN, compared with 802.1 le.     

Sea trials of an underwater, ad hoc, acoustic network with stationary assets

The diverse roles of deployable underwater assets have increased the need for ad hoc networking capability. This capability is defined as the ability to form acoustic communication links by deploying assets in a non-deterministic manner, without depending on a priori route and positioning information. Each asset represents a node in the network, and can act as a data source, sink or router, and initiates, stores or relays data, respectively. Data packets may be hopped over a few nodes before arriving at the designated destination. This ad hoc network consists of two main layers: the medium access control (MAC) layer to resolve node contention, and the ad hoc routing layer to manage routing information. In the MAC layer, both random access and timedivision protocols are applied. The protocol in the routing layer is responsible for route setup and route maintenance. This concept of multiple-hop, ad hoc networking was implemented during sea trials conducted in Portland Harbour, UK, involving up to six nodes. The trial was designed with the aim of gaining application experience in ad hoc network deployment and in assessing the performance of the protocols in a real operating environment. The emphasis was on modem integration, and the stability and reliability of packet transport. The trial results and their subsequent analysis demonstrated the robustness of the network in performing ad hoc routing and resolving node contention in a realistic and acoustically challenging environment.     

CLGA Net: Cross Layer Gated Attention Network for Image Dehazing

In this paper, we propose an end-to-end cross-layer gated attention network (CLGA-Net) to directly restore fog-free images. Compared with the previous dehazing network, the dehazing model presented in this paper uses the smooth cavity convolution and local residual module as the feature extractor, combined with the channel attention mechanism, to better extract the restored features. A large amount of experimental data proves that the defogging model proposed in this paper is superior to previous defogging technologies in terms of structure similarity index (SSIM), peak signal to noise ratio (PSNR) and subjective visual quality. In order to improve the efficiency of decoding and encoding, we also describe a fusion residual module and conduct ablation experiments, which prove that the fusion residual is suitable for the dehazing problem. Therefore, we use fusion residual as a fixed module for encoding and decoding. In addition, we found that the traditional defogging model based on the U-net network may cause some information losses in space. We have achieved effective maintenance of low-level feature information through the cross-layer gating structure that better takes into account global and subtle features. We also present the application of our CLGA-Net in challenging scenarios where the best results in both quantity and quality can be obtained. Experimental results indicate that the present cross-layer gating module can be widely used in the same type of network.