In this paper, we first propose a class of workstation cluster which utilizes optical wavelength-division multiplexing (WDM) technology to connect nodes (work-stations) of the cluster. The optical Bus Computer Cluster...
详细信息
In this paper, we first propose a class of workstation cluster which utilizes optical wavelength-division multiplexing (WDM) technology to connect nodes (work-stations) of the cluster. The optical Bus Computer Cluster (OBCC) falls in the class of cache coherent non-uniform memory access (CC-NUMA) multiprocessors. The basic topology of the cluster is star-shaped with an optical star-coupler in the center to enable one-hop simultaneous broadcasting of information packets from one node to all other cluster nodes. WDM technology not only multiplies by N times the network bandwidth using a single optical fiber, where N is the degree of wavelength multiplexing, but also provides independent communication paths between pairs of cluster nodes by properly assigning wavelengths to inter-node communication. Then we identify the cache subsystem requirements for the OBCC and propose or deferred cache coherence protocol suitable for the OBCC. The basic coherence maintenance scheme is to lazy-evaluate the cache coherence transactions among cluster nodes, utilizing the weak consistency memory model. By deferring the transactions, it is possible to combine multiple transaction issues into one transaction by accumulating modified status bits in the enhanced cache status fields. Since not only the remote memory access but also coherence transaction are costly operations in CC-NUMA systems, the deferred invocation of coherence transactions is particularly useful in CC-NUMA systems such as OBCC. We then give a performance evaluation by simulation that the coherence protocol effectively reduces coherence transactions, particularly in situations where false sharing of longer cache lines becomes noticeable.
Masix is a distributed multi-server operating system based on the Mach microkernel, with multiple personality support. Its main feature is a distributed generic layer (DGL), which offers distributed services to the pe...
详细信息
Masix is a distributed multi-server operating system based on the Mach microkernel, with multiple personality support. Its main feature is a distributed generic layer (DGL), which offers distributed services to the personalities. The distributed multi-server architecture of Masix grants it a high modularity, but also raises many issues, such as transparency, security and performance, which cannot be solved without adequate communication services. To provide total transparency, we extend the traditional Mach communication model to a workstation network by interposing a generic network server (GNS) between the tasks and the microkernel. We defined a global name service, based on a name resolution protocol, which allows any pair of Mach remote tasks to communicate transparently, using the local Mach IPC semantics. Our name server also provides local and remote authentication mechanisms, based on digital signatures and a secret key algorithm. To prevent eavesdropping, all remote communications are transparently encrypted by the GNS, using a public key algorithm. These security measures can be easily merged into the name service, to yield a secure distributed name resolution protocol. Microkernel based-systems are traditionally criticized for their relatively poor performance. As far as network services are concerned, experiments show that a good performance level can be reached, provided that the distinctive features of microkernels are taken into account.
Performance of a network depends primarily on the network topology, switching mechanism, flow control protocol and the underlying routing algorithm. While many routing algorithms have been proposed recently for direct...
详细信息
Performance of a network depends primarily on the network topology, switching mechanism, flow control protocol and the underlying routing algorithm. While many routing algorithms have been proposed recently for direct networks, there is no time efficient technique to evaluate and compare all of them. A conventional routing algorithm simulation of a network on a uniprocessor takes unacceptably large computing time. The simulation can be made very time efficient by parallelizing it and running on a parallel test bed. This research is focussed on designing a parallel routing algorithm simulator for n-dimensional mesh connected networks with wormhole switching and virtual channel flow control. The research addresses partitioning mapping, synchronization issues, and implementation of various routing algorithms for 2-D and 3-D mesh architectures. Experimental results show that the parallel simulator can provide significant speedup compared do a uniprocessor environment.
The concept of an integrated communicationnetwork supporting substation intelligent electronic devices (IED) for protection, control, and data acquisition applications is illustrated. The networkarchitecture is requ...
详细信息
The concept of an integrated communicationnetwork supporting substation intelligent electronic devices (IED) for protection, control, and data acquisition applications is illustrated. The networkarchitecture is required to provide access to IED data for users in all utility-owned facilities including the local substation, other substations, control centres and corporate offices. A secondary but growing class of application is the digital monitoring of power system primary equipment, e.g. switchgear, transformers, lines, buses and feeders and the transfer of protection system data between IEDs not only in the substation but also to external systems. The areas which are to be examined for seeking improved business performance are related to asset management support. Asset maintenance, refurbishment and replacement options are to be examined and decision support systems are to be specified. The paper indicates possibilities to solve the key questions which are to be considered for the implementation of the new technologies: What benefits can be gained in replacing secondary systems (protection and control equipment)? Can the primary plant, circuit breakers, transformers etc. be better utilised? Can primary plant maintenance be optimised? Is it possible to extend the lifetime of primary plant?.
network deployment of voice services based on IS-95 CDMA using both 8 kbps and 13 kbps service options are explored from the standpoint of capacity and coverage. Specifically, issues regarding the various performance ...
详细信息
network deployment of voice services based on IS-95 CDMA using both 8 kbps and 13 kbps service options are explored from the standpoint of capacity and coverage. Specifically, issues regarding the various performance tradeoffs associated with operation at 8 kbps and 13 kbps are examined. Significant differences in the system performance exist between these service options, particularly on the forward link. System processing gain, E/sub b//N/sub t/ link performance, forward link power control, along with voice activity and Erlang variations, all contribute to system performance differences. The impact of these and other deployment-related variables on the delivered end-to-end quality of voice services in a large network are addressed. The functionality of a CDMA radio network computer model, TELE*design, which can be used for RF planning of CDMA-based cellular and PCS systems is also discussed.
Personal communicationnetwork (PCN) technology has become the most important topic in next-generation telecommunicationsystems. This technology is going to deliver a set of communication services to a hand-held unit...
详细信息
Personal communicationnetwork (PCN) technology has become the most important topic in next-generation telecommunicationsystems. This technology is going to deliver a set of communication services to a hand-held unit at any location. In order to support all these services, asynchronous transfer mode (ATM) technology has been proposed as the choice for mixed data, voice and video communications over a single network. It has been recognised that multimedia wireless networks have become a natural way of providing wireless extensions for future broadband networks. These networks, based on an ATM transport architecture, are designed to support a mix of broadband ISDN services including constant bit-rate (CBR), variable bit-rate (VBR) and packet data. MDR-TDMA is one of the proposed protocols for the medium access control (MAC) sublayer of such a network. We analyze the performance of this protocol for a multimedia scenario. Simulation results of quality-of-service (QoS) and other parameters are presented for this system.
Parallel and distributed computer architectures are increasingly being considered for application in a wide variety of computationally intensive embedded systems. Many such applications impose highly dynamic demands f...
详细信息
Parallel and distributed computer architectures are increasingly being considered for application in a wide variety of computationally intensive embedded systems. Many such applications impose highly dynamic demands for resources (processors, memory, and communicationnetwork), because their computations are data-dependent, or because the applications must constantly interact with a rapidly changing physical environment, or because the applications themselves are adaptive. This paper presents a set of dynamic resource allocation techniques aimed at maintaining high levels of application performance in the presence of varying resource demands. It focuses on a class of applications structured as multiple pipelines of data-parallel stages, as this structure is common to many sensor-based applications. We discuss the issues involved in resource management for such applications, and present preliminary results from our implementations on Intel Paragon. Our approach uses feedback control-a real-time monitoring system is used to detect significant performance shortfalls, and resources are reallocated among the application components in an attempt to improve performance. The main contribution of this work is that it combines real-time monitoring of an application's performance with dynamic resource allocation, and focuses on practical implementations rather than simulation and analysis.
The proceedings contain 33 papers. The topics discussed include: characterization of an all-optical clock recovery figure-eight laser for systems applications;simplified tuned optical receivers for use in lightwave op...
The proceedings contain 33 papers. The topics discussed include: characterization of an all-optical clock recovery figure-eight laser for systems applications;simplified tuned optical receivers for use in lightwave optical subcarrier multiplex systems;improved formulation for optically preamplified PPM receivers incorporating timing jitter effects;photonic switch size limitations imposed by phase noise from semiconductor optical amplifiers;widely tunable source generating subpicosecond pulses at 1550 nm;description of all-opticalnetwork test bed and applications;digital pulse interval and width modulation for optical fiber communications;design and realization of a suboptimum presynchronization filtering strategy for optical receivers;controlled switching of solitons in fiber nonlinear directional couplers;and comparison and classification of all-optical CDMA systems for future telecommunicationnetworks.
The proceedings contains 33 papers. Some of the specific topics discussed are: widely tunable source generating picosecond pulses at 1550 nm;description of all-opticalnetwork test bed and applications;asymmetry of fr...
详细信息
ISBN:
(纸本)0819419788
The proceedings contains 33 papers. Some of the specific topics discussed are: widely tunable source generating picosecond pulses at 1550 nm;description of all-opticalnetwork test bed and applications;asymmetry of free space optical links;compressed HDTV over the optical fiber network using PPM CDMA;Kautz topologies for all-optical self-routing networks;and pulse-time modulation for subcarrier-multiplexed systems.
We describe the design and performance of a wide-area all-opticalnetwork. An improved model of the benefits of wavelength changers in all-optical mesh networks is presented and comparisons are made to simulations. Th...
详细信息
ISBN:
(纸本)0819419788
We describe the design and performance of a wide-area all-opticalnetwork. An improved model of the benefits of wavelength changers in all-optical mesh networks is presented and comparisons are made to simulations. The model is shown to be accurate for a moderate number of wavelengths and a small blocking probability for a variety of networks.
暂无评论