Anatomy of UDP and M-VIA for cluster communication

High interprocess communication latency is detrimental to parallel and grid computing. Over the years, the network bandwidth has increased rapidly while the end-to-end latency has not decreased much. This is because the latency is dominated by the protocol software execution time in the kernel instead of the raw transmission time over the link. In this paper, we perform an anatomical analysis of the complete communication path between a sender and a receiver through measurements. We present an in-depth evaluation of various components of the UDP protocol over Fast Ethernet. Virtual Interface Architecture (VIA) protocol has been recently proposed to overcome the software overhead of the TCP/UDP/IP protocol. We analyze M-VIA, a modular VIA implementation for Linux over Ethernet, and compare its performance with UDP. The aim of our experiments is to present the protocol overheads in details rather than to suggest new techniques to reduce overheads.     

End-to-end performance of 10-gigabit Ethernet on commodity systems

Apart form the success in local-area networks (LANs) and system-area networks and anticipated success in metropolitan and wide area networks (MANs and WANs), Ethernet continues to evolve to meet the increasing demands of packet-switched networks. Although the recently ratified 10-Gigabit Ethernet standard differs from earlier Ethernet standards, primarily in that 10GbE operates only over fiber and only in full-duplex mode, the differences are largely superficial. More importantly, l0GbE does not make obsolete current investments in network infrastructure. The 10GbE standard ensures interoperability not only with existing Ethernet but also with other networking technologies such as Sonet, thus paving the way for Ethernets expanded use in MANs and WANs. The world's first host-based 10GbE adapter, officially known as the Intel PRO/10GbE LR server adapter, introduces the benefits of l0GbE connectivity into LAN and system-area network environments, thereby accommodating the growing number of large-scale cluster systems and bandwidth-intensive applications, such as imaging and data mirroring. The 10GbE controller is optimized for servers that use the I/O bus backplanes of the peripheral component interface (PCI) and its higher speed extension, PCI-X.     

基于以太网交换机的Linux协议栈虚拟化技术

网络协议栈虚拟化的目标是在一台网络设备上虚拟出多个协议栈以实现网络设备的一虚多功能。由于 Linux操作系统的网络协议栈本身是不支持这种网络协议栈虚拟化技术的,因此本文通过对Linux操作系统3.11.10版本 IPV4网络协议栈进行改造以实现 Linux网络协议栈的虚拟化,并在设备平台上验证了经过网络协议栈虚拟化改造后的以太网交换机对业务流的隔离功能。     

一种低延时的无线传感器网络MAC协议

针对典型无线传感器网络X-MAC协议在多跳环境下累积延迟问题,提出一种具有低时延的Ex-MAC协议。利用管理邻居节点工作周期和虚通道算法,使得虚通道中的所有节点近似于同步,从而减少数据传输延迟,而且不需要交换同步控制帧。仿真结果与X-MAC协议相比,减少了端到端的延迟和数据包冲突的几率,提高了网络的吞吐量。     

An energy-efficient,transport-controlled MAC protocol for wireless sensor networks

In wireless sensor networks (WSNs), one major cause of wasted energy is that the wireless network interface is always on to accept possible traffic. Many medium access control (MAC) protocols therefore adopted a periodic listen-and-sleep scheme to save energy, at sacrifice of end-to-end latency and throughput. Another cause is packet dropping due to network congestion, necessitating a lightweight transport protocol for WSNs. In this paper, we suggest a transport-controlled MAC protocol (TC-MAC) that combines the transport protocol into the MAC protocol with the aims of achieving high performance as well as energy efficiency in multi-hop forwarding. Although TC-MAC also works through a periodic listen-and-sleep scheme, it lowers end-to-end latency by reserving data forwarding schedules across multi-hop nodes during the listen period and by forwarding data during the sleep period, all while increasing throughput by piggybacking the subsequent data forwarding schedule on current data transmissions and forwarding data consecutively. In addition, TC-MAC gives a fairness-aware lightweight transport control mechanism based on benefits of using the MAC-layer information. The results show that TC-MAC performs as well as an 802.11-like MAC in end-to-end latency and throughput, and is more efficient than S-MAC in energy consumption, with the additional advantage of supporting fairness-aware congestion control.     

基于跨层设计的无线传感器网络MAC协议

针对占空比MAC协议存在端到端传输延迟问题,提出一种新的占空比MAC协议——PRMAC。PRMAC通过跨层路由信息帧的提前传送可以调度数据包在一个周期内多跳传输,从而降低网络延迟,提高能量有效性。NS-2仿真结果表明,PRMAC在没有牺牲能量有效性的情况下,改进了传统占空比协议的端到端传输延迟,并能提高网络吞吐量。     

uIP协议栈在基于DSP以太网通信系统中的应用

文章通过对uIP协议的简单分析介绍,给出了以太网通信系统整体硬件平台、设计思路和软件工作流程。完成了以太网控制器硬件驱动的开发和嵌入式uIP协议栈的移植,并给出了面向DSP的优化编程方法。实现了以TMS320F2812数字信号处理器和CS8900A以太网控制器为核心的局域网络DSP端的以太网通信。     

An Open Generic Corba-Programmable Videoconferencing Platform

In this paper we address the design and implementation of a generic terminal architecture for videoconferencing based on the principle of programmable protocol stacks. In this architecture, there exist two types of software components. A first type are the multimedia engines, which are responsible for handling the multimedia stream (execution layer). Secondly, the managing components handle the multimedia engines' lifecycles and provide a set of network- and codec-independent CORBA-IDL interfaces which make building the desired protocol stack straightforward (middleware based control layer). As a result, the developed platform can be easily integrated in existing multimedia control frameworks such as the TINA architecture or the H.323 terminal framework.     

数字化变电站网络性能的测试方法研究

针对数字化变电站中以太网络交换机的测试需求,制作了基于WinPcap和WinSNMP的SWITCHTEST软件测试包.使用WinPcap,在高精度时间间隔中向交换机发送模拟网络报文并且回收,通过比较收发报文而判断交换机的性能.使用WinSNMP,可以对交换机进行网络监控管理,对交换机进行基本性能测试:吞吐量、时延、帧丢失率、拥塞;高级性能测试:MAC地址容量、RSTP恢复、广播风暴、VLAN、ARP欺骗、SNTP对时;网络管理协议测试:SNMP、RMON、TRAP、IGMP Snooping 、GMRP.由以上各项测试可以得到以太网络交换机的性能参数,测试结果自动生成WORD报告和实时曲线,对于提高电力信息以太网络的安全可靠运行有重要参考意义.说明了使用WinPcap和WinSNMP可以较好地分析以太网络交换机性能.     

一种无中心无线网络协议栈的设计及实现

本文介绍了一种无中心无线网络系统的空中接口高层协议栈的设计及软硬件实现参考。本高层协议栈是依据现代通信协议的分层思想对传统协议栈进行重新设计,和传统协议栈相比协议层次更清晰,提高了组网的灵活性、增大了网络容量,并使得设备更加易于配置和管理。     

一种基于Linux的网络协议栈的设计与实现

实现网络协议栈是嵌入式系统开发中的重要内容,并且具有一定的难度。文章利用Linux内核中基于AF_INET套接字类型的TCP/IP网络协议栈的实现机制,在不影响Linux系统原有网络协议栈的情况下创建了一种新的套接字类型的网络协议栈,并应用于实际路由器软件的实现,通过在原有的Linux内核中添加新的服务,实现了对路由器硬件的操作。     

井下气体监测系统的嵌入式网关设计与实现

为满足矿井下大范围空间的气体监测的以太网通信要求,实现监测数据由井下到井上的远距离传输,设计了一套基于JN5148单片机和W5500网络芯片的ZigBee-以太网网关设计方案,阐述了网关的硬件以及通信软件的设计.利用W5500内部集成的TCP/IP协议栈实现了以太网的通信,省去了在主控单片机上的TCP/IP协议栈移植.该方案具有开发周期短和数据传输稳定可靠等优点,能够简单灵活地实现ZigBee协议与TCP/IP协议的切换.设计网关的测试结果表明,可以实现ZigBee与以太网的数据传输和共享.     

基于双缓冲队列的RS485串口／以太网转换器软件设计

为了将分散于工业现场的串口设备接入以太网,本文设计一种基于STM32F107VC微处理器的嵌入式RS485串口/以太网的数据转换器。移植嵌入式TCP/IP协议栈LwIP到ARM平台,构造双缓冲队列,保证串行通信的效率,通过软件编程实现串行通信协议和以太网的TCP/IP协议的相互映射,使得串口数据和以太网数据相互转换。实验表明,串口/以太网数据转换器能够正常工作,数据转换稳定、可靠。     

A Speculative and Adaptive MPI Rendezvous Protocol Over RDMA-enabled Interconnects

Overlapping computation with communication is a key technique to conceal the effect of communication latency on the performance of parallel applications. Message Passing Interface (MPI) is a widely used message passing standard for high performance computing. One of the most important factors in achieving a good level of overlap is the MPI ability to make progress on outstanding communication operations. In this paper, we propose a novel speculative MPI Rendezvous protocol that uses RDMA Read and RDMA Write to effectively improve communication progress and consequently the overlap ability. Performance results based on a modified MPICH2 implementation over 10-Gigabit iWARP Ethernet reveal a significant (80–100%) improvement in receiver side overlap and progress ability. We have also observed up to 30% improvement in application wait time for some NPB applications as well as the RADIX application. For applications that do not benefit from this protocol, an adaptation mechanism is used to stop the speculation to effectively reduce the protocol overhead.     

IPv6在矿井工业以太网中的应用研究

介绍了IPv6产生的背景和发展现状,阐述了IPv6应用于矿井工业以太网的发展趋势和研究方向,指出对于所有基于矿井工业以太网的硬件和软件的开发都应带有双协议栈,随着IPv6技术的成熟和普及再研发出只有IPv6协议栈的适用于煤矿的网络设备,如支持IPv6的防爆网络设备、组态软件以及相应的数据采集驱动设备、控制硬件等。     

Phoebus: A system for high throughput data movement

Phoebus is an infrastructure for improving end-to-end throughput in high-bandwidth, long-distance networks by using a “session layer” protocol and “gateways” in the network. Phoebus has the ability to dynamically allocate network resources and to use segment-specific transport protocols between gateways, as well as to apply other performance-improving techniques on behalf of the user. We have developed interfaces to Phoebus to allow its use in various real applications and data movement services. This paper extends our earlier work with tests of Phoebus-enabled applications on both real-world networks as well as over configurable network testbeds that allow us to modify latency and loss rates. We demonstrate that Phoebus improves the performance of bulk data transfer in a variety of network configurations and conditions.     

基于Ethernet的嵌入式测控系统的设计

在分析工业控制网络发展趋势、嵌入式以太网接入Internet网络数据传输需要的协议的基础上,针对现场总线通信方式的不足,设计了一个基于Ethernet的嵌入式测控系统。文章阐述了系统的设计方案,在硬件电路设计上,重点介绍了测控系统数据采集节点设计和以太网通信接口电路的设计;在软件系统设计上,提出测控终端软件设计方案,并实现了基于TCP／IP协议栈的网络通信设计。实际使用表明,该系统运行灵活、可靠、稳定,可直接使用企业内部的Intranet信息网,也可直接接人现行的公共Internet网络。     

Will 10-Gigabit Ethernet have a bright future?

Researchers, engineers, and vendors are readying the newest Ethernet standard which promises not only faster speeds than previous versions but also important new features. The proposed 10-Gigabit Ethernet (10 GbE) standard, known as IEEE 802.3ae, would not only provide data rates of 10 Gbits per second but would also work with metropolitan and wide area networks (MANs and WANs), where ATM (asynchronous transfer mode) and Sonet (synchronous optical network) currently dominate. Ethernet and Fast Ethernet were designed to work primarily with local area networks (LANs), where they are the dominant technologies. The IEEE expects to formally adopt the 802.3ae standard this year     

Long-haul secure data transfer using hardware-assisted GridFTP

Extreme-scale scientific collaborations require high-performance wide-area end-to-end data transports to enable fast and secure transfer of high data volumes among collaborating institutions. GridFTP is the de facto protocol for large-scale data transfer in science environments. Existing predominant network transport protocols such as TCP have serious limitations that consume significant CPU power and prevent GridFTP from achieving high throughput on long-haul networks with high latency and potential packet loss, reordering and jitter. On the other hand, protocols such as UDT that address some of the TCP shortcomings demand high computing resources on data transfer nodes. These limitations have caused underutilization of existing high-bandwidth links in scientific and collaborative grids. To address this situation, we have enhanced Globus GridFTP, the most widely used GridFTP implementation, by developing transport offload engines such as UDT and iWARP on SmartNIC, a programmable 10GbE network interface card (NIC). Our results show significant reduction in server utilization and full line-rate sustained bandwidth in high-latency networks, as measured for up to 100 ms of network latency. In our work, we also offload OpenSSL on SmartNIC to reduce host utilization for secure file transfers. The offload engine can provide line-rate data channel encryption/decryption on top of UDT offload without consuming additional host CPU resources. Lower CPU utilization leads to increased server capacity, which allows data transfer nodes to support higher network and data-processing rates. Alternatively, smaller or fewer DTNs can be used for a particular data rate requirement.     

万兆以太网技术及应用

万兆以太网（10GbE)是一种高速以太网技术，这种以太网采用IEEE802.3以太网媒体访问控制（MAC）协议、帧格式和帧长度。它的优点是减少网络的复杂性，兼容现有的局域网技术，降低系统费用，并提供更快、更新的数据业务。10GbE与弹性分组数据环（RPR）结合产生的10GRPR技术。吸收了10GbE的经济性、光纤环网的高带宽效率和可靠性，是一种非常有潜力的宽带城域网技术。