一种无缓存光互连网络中的结点饿死问题及其解决

本文针对一种无缓存的高性能计算机光互连网络BOIN中存在的结点饿死问题,提出了两种不同的解决方法——尽量回避的X优先路由算法和允许丢弃的X优先路由算法。这两种路由算法利用了报文在向X方向发送时其Y方向链路空闲的特点,使得发生冲突的报文可以通过空闲的链路顺利转发。模拟实验结果表明,采用这两种路由算法,能够很好地解决报文在发送时的饿死现象。     

胖树中的分布式动态容错路由

面向云计算的超大规模互连网络增加了对网络容错的要求,容错已成为互连网络的重要问题.为了保证网络的高可用性和高性能,文中基于胖树网络拓扑提出了一种分布式的动态容错路由方法.该方法通过引入一套链路失效消息传播机制和一套基于链路失效信息的动态容错路由算法来实现胖树网络的分布式动态容错.相比已有方法,该方法不增加网络硬件和路由路径长度,并且具有高执行效率和高性能.实验结果表明,在m端口交换机构成的胖树中,该方法可以容忍任意m/2-1条失效链路并以高概率容忍更多条失效链路的组合,同时保持网络的高性能.     

一种基于虫洞交换的竞争预测路由算法

分析了基于虫洞交换技术的片上互连网络路由算法存在的一些问题:固定维序路由当出现竞争时,只能等待,直到链路空闲方可继续前进.热土豆路由在路由器从不等待,任意选择一个空闲的端口进行发送,但却有活锁问题.利用相邻路由节点之间提供的竞争感知信号,提出了一种竞争预测的自适应路由算法,并利用SystemC片上互连网络仿真实现该算法.实验表明,该路由算法的网络传输延时在竞争多发的情况下明显优于热土豆算法与传统的固定维序算法.     

多校区校园网络互连与资源共享研究

多校区网络互连是资源共享的根本,资源共享需要足够的硬盘空间。该文在实践过程中采用专线连接和VPN的方式完成了多区域的网络互连,保障了网络通道的顺畅,采用服务器虚拟化技术扩充了服务器数量,采用虚拟存储技术扩充了硬盘空间,有效利用了空闲分区资源。     

面向片上网络路由器漏流功耗优化的自适应缓冲管理策略

随着单芯片上集成处理器数量的增加,片上网络逐渐成为多核处理器中非常有前景的互连结构.互连网络成为片上多处理器功耗的重要消耗部件之一.而输入缓冲器是路由器漏流功耗的最大消耗单元,采用门控电源是降低其漏流功耗的有效手段.自适应缓冲管理策略能够根据网络中通信量,自适应地关闭/打开缓冲的一部分,从而降低路由器漏流功耗.而为了减小对网络延迟的影响,该策略中采用的提前唤醒技术能够隐藏缓冲的唤醒延迟.在网络注入率较低情况下,两项缓冲不关闭策略下的网络延迟几乎不受唤醒延迟影响.模拟结果显示,在4×4的二维Mesh中,即使网络注入率为0.7,漏流功耗的节约率依然可以高达46％;网络注入率小于0.4时,两项缓冲不关闭策略下的网络延迟最大仅仅增加了3.8％.     

一种二维结构的互连路由设计与实现

针对阵列处理单元之间互连传输拥塞的问题,设计了一种在二维结构中互连的虚通道路由器模型。采用改进的自适应XY路由算法,智能地分配虚通道空闲资源,从逻辑上减少拥塞和等待时间,多路选择器交叉开关完成数据传输。通过ASIC设计,完成虚通道路由器硬件电路,应用Modesim工具进行仿真,达到网络互连传输的目的。     

片上互连网络的功耗特征分析与优化

随着处理器核数的增加,片上互连网络NoC结构日趋复杂,导致片上互连网络功耗所占的比重和功耗分析的难度也在增加。片上互连网络的任务映射,既要保证多处理器核心之间通信的高性能,又要保证耗费尽可能少的功耗和面积,即在有限的功耗和面积开销下获得较高的性能。在进行任务映射时,核心之间的通信距离是减少任务通信功耗的关键。连续且近凸的区域有助于缩短任务的通信距离。分析了一种功耗最优的片上互连网络启发式映射算法（INC）,该算法由区域选择算法和节点映射算法组成。对区域选择算法的2个因子进行了改进,使应用总的通信开销最小化且保证后续应用以很小的通信代价进行区域选择。提出了新的基于选择区域的映射算法。它们在动态到达程序映射问题中的实验结果表明,新的区域选择算法和节点映射算法相比于INC,可以减少12.10%的通信功耗,并且带来11.23%的通信延迟优化。     

一种新的基于预约的拥塞避免机制

由于高速互连网络上的负载不均衡,一些网络结点成为了热点,可能导致部分结点或是链路拥塞,这会极大地降低互连网络的性能。现有的基于预约的拥塞避免技术SRP可以进行主动的拥塞避免,极大地缓解了由于热点问题所带来的负面效应。但是,在热点模式下,其它非热点结点的路由器资源绝大多数处于空闲状态,为了进一步充分利用互连网络的资源,提升互连网络性能,提出了一种基于SRP改进的中间结点缓存技术IRP。IRP可以根据不同的拓扑,例如胖树,有效地利用热点的邻居结点的路由器资源,先利用胖树拓扑的多路径将报文发送给空闲路由器,一旦目的结点路由器可利用,则将缓存报文发送给目的结点,降低互连网络的延迟。     

分级环片上网络互连

在大规模、超大规模片上互连网络中,因为二维互连方式的性能较差而使多维互连方式成为可选方案之一.文中首先基于区域划分设计了一种分级环互连结构,分析了其静态互连特性,然后基于卡诺图编码设计了一种分级环互连的路由结构以及寻径方法,在均匀通信模式测试了不同的分级环级联链路缓冲区设置方法下网络的性能,详细分析了按照等比序列设置分级环级联链路缓冲区时分级环互连方式的动态网络特性,最后根据互连性能与Mesh等二维片上互连方式比较的结果,给出了分级环互连方式的使用场合.实验结果表明,虽然在较小规模网络中性能较差,但是分级环互连方式能以较低的成本、较高的性能实现大规模、超大规模片上网络的互连,其中单环分级互连方式在较低网络负载下综合性能更好,而双环分级互连方式则具有更大的网络负载能力,在较高网络负载下性能更好.     

高层次数据流的多电压低功耗综合方法

为了有效地进行算法行为描述到寄存器传输级结构描述的转换,提出一种同时考虑功能单元功耗、互连功耗和电压转换功耗的模型和基于网络流的低功耗设计方法.首先对给定的数据流图进行单电压高层次综合,然后对单电压综合结果迭代地进行多电压调整;提取每次迭代时需要调整的网络流子图,对该子图运行最小费用最大流增量算法.该方法充分利用前面迭代中得到的优化解,避免了对整个网络流的重复计算.实验结果表明,文中方法在互连功耗、电压转换功耗和总功耗等方面均有较大优化.     

Exploring virtual network selection algorithms in DSM cache coherence protocols

Distributed shared memory (DSM) multiprocessors typically require disjoint networks for deadlock-free execution of cache coherence protocols. This is normally achieved by implementing virtual networks with the help of virtual channels or virtual lanes multiplexed on a single physical network. To keep the coherence protocol simple, messages are usually assigned to virtual lanes in a predefined static manner based on a cycle-free lane assignment dependence graph. However, this static split of virtual networks (such as request and reply networks) may lead to underutilization of certain virtual networks while saturating the other networks. In this paper, we explore different static and dynamic schemes to select the virtual lanes for outgoing messages and mix the load among them without restricting any particular type of message to be carried only by a particular virtual network. We achieve this by exposing the selection algorithms to the coherence protocol itself, so that it can inject messages into selected virtual lanes based on some local information, and still enjoy deadlock-freedom. Our execution-driven simulation on five applications from the SPLASH-2 suite shows that as the system scales, the virtual network selection algorithms play an important role. For 128-node systems, our dynamic selection algorithm speeds up parallel execution by as much as 22 percent over an optimized baseline system running a modified SGI Origin 2000 protocol. We also explore how network latency, the number of message buffers per virtual lane, and the depth of network interface output queues affect the relative performance of various virtual lane selection algorithms.     

Scale-free network models with accelerating growth

Complex networks are everywhere. A typical example is software network. Basing on analyzing evolutive structure of the software networks, we consider accelerating growth of network as power-law growth, which can be more easily generalized to real systems than linear growth. For accelerating growth via a power law and scale-free state with preferential linking, we focus on exploring the generic property of complex networks. Generally, two scenarios are possible. In one of them, the links are undirected. In the other scenario, the links are directed. We propose two models that can predict the emergence of power-law growth and scale-free state in good agreement with these two scenarios and can simulate much more real systems than existing scale-free network models. Moreover, we use the obtained predictions to fit accelerating growth and the connectivity distribution of software networks describing scale-free structure. The combined analytical and numerical results indicate the emergence of a novel set of models that considerably enhance our ability to understand and characterize complex networks, whose applicability reaches far beyond the quoted examples.     

Scale-free network modles with accelerating growth

Complex networks are everywhere. A typical example is software network. Basing on analyzing evolutive structure of the software networks, we consider accelerating growth of network as power-law growth, which can be more easily generalized to real systems than linear growth. For accelerating growth via a power law and scale-free state with preferential linking, we focus on exploring the generic property of complex networks. Generally, two scenarios are possible. In one of them, the links are undirected. In the other scenario, the links are directed. We propose two models that can predict the emergence of power-law growth and scale-free state in good agreement with these two scenarios and can simulate much more real systems than existing scale-free network models. Moreover, we use the obtained predictions to fit accelerating growth and the connectivity distribution of software networks describing scale-free structure. The combined analytical and numerical results indicate the emergence of a novel set of models that considerably enhance our ability to understand and characterize complex networks, whose applicability reaches far beyond the quoted examples.     

一种安全的多源网络编码签名算法

网络编码已被证明在提高网络传输速率、减少网络拥塞、增强网络可靠性以及降低结点能耗等方面有着巨大优势,因此可被广泛用于计算机网络、无线传感器网络以及p2p系统中。但是,网络编码也面临着多方面的安全威胁,其中污染攻击是其最主要的安全威胁之一。目前,绝大部分的签名方案均只能适用于对单源网络编码中的污染攻击,无法满足普遍存在的多源网络编码的安全要求。对多源网络编码中的污染攻击进行了更深入的分析,给出了安全的多源网络编码所需要满足的条件,提出了一个安全高效的多源网络编码签名算法来预防网络中的污染攻击,该算法可以满足多源网络编码的全部安全要求。     

基于神经网络的配网系统光伏输出功率控制分析

随着目前世界上的能源需求愈发扩大,光伏发电凭借其储备量大,且清洁无污染的特性,逐步成为目前新能源发电的主流,但是,由于光伏发电效率受到环境光照强度的影响,因此,其输出功率时时发生变化,所以,目前光伏电池的最大功率跟踪(Most Power Point Trace,MPPT)与控制技术已经成为了业界最为关注的问题。针对该问题,本文利用BP神经网络技术对光伏电池的最大输出功率进行检测以及控制,通过对光伏系统以及人工神经网络的基本原理进行介绍,引入了BP神经网络的基本概念,最后搭建了基于BP神经网络的配网光伏输出功率控制系统,通过仿真,证明了其理论的可行性与正确性,能够为我国光伏产业提供一定帮助。     

Exploring the Design Space of Self-Regulating Power-Aware On/Off Interconnection Networks

With power consumption becoming increasingly critical in interconnected systems, power-aware networks become part-and-parcel of many single-chip and multichip systems. As communication links consume significant power regardless of utilization, a mechanism to realize such power-aware networks is on/off links-network links that can be turned on/off as a function of traffic. In this paper, we investigate and propose self-regulating power-aware interconnection networks that turn their links on/off in response to bursts and dips in traffic in a distributed fashion. We explore the design space of such on/off networks, outlining a 5-step design methodology along with various building block solutions at each step that can be effectively assembled to develop various on/off network designs. We applied our methodology to the design of two classes of on/off networks with links that possess substantially different on/off delays, an on-chip network as well as a chip-to-chip network, and show that our designs are able to adapt dynamically to variations in network traffic. Three specific network designs are then constructed, presented, and evaluated. Our simulations show that link power consumption can be reduced by up to 54.4 percent, with a modest increase in network latency     

Alleviating consumption channel bottleneck in wormhole routed k-aryn-cube systems

This paper identifies performance degradation in wormhole routed k-ary n-cube networks due to limited number of router-to-processor consumption channels at each node. Many recent research in wormhole routing have advocated the advantages of adaptive routing and virtual channel flow control schemes to deliver better network performance. This paper indicates that the advantages associated with these schemes cannot be realized with limited consumption capacity. To alleviate such performance bottlenecks, a new network interface design using multiple consumption channels is proposed. To match virtual multiplexing on network channels, we also propose each consumption channel to support multiple virtual consumption channels. The impact of message arrival rate at a node on the required number of consumption channels is studied analytically. It is shown that wormhole networks with higher routing adaptivity, dimensionality, degree of hot-spot traffic, and number of virtual lanes have to take advantage of multiple consumption channels to deliver better performance. The interplay between system topology, routing algorithm, number of virtual lanes, messaging overheads, and communication traffic is studied through simulation to derive the effective number of consumption channels required in a system. Using the ongoing technological trend, it is shown that wormhole-routed systems can use up to two-four consumption channels per node to deliver better system performance     

基于聚类算法和转换网络的光伏短期功率预测

精准的光伏功率预测对优化光伏电站的运行和管理以及提高光伏发电的效率具有重要的作用。本文提出了一种基于聚类算法和转换网络的光伏短期功率预测方法。该方法首先基于自编码器的无监督聚类算法对光伏短期功率数据进行了预处理,以降低光伏出力数据本身的不稳定性对功率预测的影响。之后,该方法使用具有自注意力机制和多头注意力机制的转换网络进行光伏短期功率的预测。转换网络由编码器和解码器组成。转换网络相比传统的循环神经网络(RNN)更善于挖掘时序之间的关系。注意力机制使得转换网络具有并行计算的能力,可以加快网络训练的速度。最后,在澳大利亚光伏功率与气象数据中心 (DKASC)的光伏数据集上验证了本文提出的光伏短期功率预测方法。实验结果表明,本文提出的方法具有令人满意的预测精度。     

基于深度Q网络的电力工控网络异常检测系统

电力是指以电能作为动力的能源,完整的电力系统包括发电、输电、变电、配电和用电等环节。电力是关系国计民生的基础产业,电力供应和安全事关国家安全战略,事关经济社会发展全局。工业自动化和控制系统（简称“工控”）作为电力的感官和中枢神经系统,确保其网络安全,使其始终处于稳定可靠运行状态,对于保障电力安全运营至关重要。由于大部分网络都是高度互联的,因此都易受到网络攻击的威胁。虽然基于网络的入侵检测系统可以将入侵警告和安全响应进行很好的结合,但是随着技术的不断发展,攻击变得越来越普遍且难以检测,其中逃逸技术就是这类技术的一个代表,它可以通过伪装修改网络数据流以此来逃避入侵检测系统的检测。结合所学知识和电力工控网络的特点,提出一种基于深度强化学习的电力工控网络入侵检测系统,深度强化学习的算法融合神经网络和Q-learning的方法来对网络中的异常现象进行训练,通过训练使系统能及时地检测出入侵行为并发出警告。     

Minimization of Energy Losses in Electric Power Distribution Systems by Intelligent Search Strategies

Energy is continuously dissipated in electric power systems due to electrical resistance in transmission and distribution lines. This paper addresses the problem of obtaining a network topology with minimum energy losses for electric power distribution systems. As distribution networks must operate radially, the problem can be formulated as a generalization of the minimum spanning tree problem. The generalization is due to variation in costs as network configuration changes. Nonlinear network flow techniques are teamed with search strategies borrowed from the field of artificial intelligence to overcome computation intractability.