High Network Latency Research Articles

Adaptation and Evaluation of the Multisplitting-Newton and Waveform Relaxation Methods Over Distributed Volatile Environments

This paper presents new adaptations of two methods that solve large differential equations systems, to the grid context. The first method is based on the Multisplitting concept and the second on the Waveform Relaxation concept. Their adaptations are implemented according to the asynchronous iteration model which is well suited to volatile architectures that suffer from high latency networks. Many experiments were conducted to evaluate and compare the accuracy and performance of both methods while solving the advection-diffusion problem over heterogeneous, distributed and volatile architectures. The JACEP2P-V2 middleware provided the fault tolerant asynchronous environment, required for these experiments.

Communication latency tolerant parallel algorithm for particle swarm optimization

Particle swarm optimization (PSO) algorithm is a population-based algorithm for finding the optimal solution. Because of its simplicity in implementation and fewer adjustable parameters compared to the other global optimization algorithms, PSO is gaining attention in solving complex and large scale problems. However, PSO often requires long execution time to solve those problems. This paper proposes a parallel PSO algorithm, called delayed exchange parallelization, which improves performance of PSO on distributed environment by hiding communication latency efficiently. By overlapping communication with computation, the proposed algorithm extracts parallelism inherent in PSO. The performance of our proposed parallel PSO algorithm was evaluated using several applications. The results of evaluation showed that the proposed parallel algorithm drastically improved the performance of PSO, especially in high-latency network environment.

On an efficient NoC multicasting scheme in support of multiple applications running on irregular sub-networks

When a number of applications simultaneously running on a many-core chip multiprocessor (CMP) chip connected through network-on-chip (NoC), significant amount of on-chip traffic is one-to-many (multicast) in nature. As a matter of fact, when multiple applications are mapped onto an NoC architecture with applicable traffic isolation constraints, the corresponding sub-networks of these applications are mapped onto actually tend to be irregular. In the literature, multicasting for irregular topologies is supported through either multiple unicasting or broadcasting, which, unfortunately, results in overly high power consumption and/or long network latency. To address this problem, a simple, yet efficient hardware-based multicasting scheme is proposed in this paper. First, an irregular oriented multicast strategy is proposed. Literally, following this strategy, an irregular oriented multicast routing algorithm can be designed based on any regular mesh based multicast routing algorithm. One such algorithm, namely, Alternative Recursive Partitioning Multicasting (AL + RPM), is proposed based on RPM, which was designed for regular mesh topology originally. The basic idea of AL + RPM is to find the output directions following the basic RPM algorithm and then decide to replicate the packets to the original output directions or the alternative (AL) output directions based on the shape of the sub-network. The experiment results show that the proposed multicast AL + RPM algorithm can consume, on average, 14% and 20% less power than bLBDR (a broadcasting-based routing algorithm) and the multiple unicast scheme, respectively. In addition, AL + RPM has much lower network latency than the above two approaches. To incorporate AL + RPM into a baseline router to support multicasting, the area overhead is fairly modest, less than 5.5%.

Characterizing, Measuring, and Validating the Temporal Consistency of Live—Virtual—Constructive Environments

A distinguishing characteristic of interactive live—virtual—constructive (LVC) environments is the relaxation of data consistency to improve the performance and scalability of the underlying distributed simulation. Relaxing data consistency improves the interactive performance of the environment because autonomous distributed simulation applications can continue executing and responding to local inputs without waiting for the most current shared data values. Scalability also improves since live and simulated entities from distant geographic locations can be interconnected through relatively high-latency networks. We introduce a temporal consistency model to formally define consistency for the dynamic shared state of a LVC environment for both continuous and discrete data objects. The level of inconsistency tolerated by a LVC is found to be a function of the accuracy and timeliness requirements for the distributed data objects. These requirements are mapped to specific time intervals for which data objects are considered valid. We also develop a real-time algorithm to compute the temporal consistency of individual data objects within the LVC.

Time Synchronization Protocol with Minimum Message Communication for High Latency Networks

Time synchronization is a critical component of the infrastructure of wireless sensor networks (WSN). In a high latency environment such as underwater, traditional approaches to time synchronization have limited accuracy. A new method is describe for time synchronization that takes into account clock skew, clock offset, and also propagation delay. Minimum message communication is used as a performance measure of the quality of this new time synchronization protocol.

A Parallel Algorithm To Solve Large Stiff ODE Systems On Grid Systems

This paper introduces a parallel algorithm to solve large stiff ODE systems on distributed clusters, with computing nodes geographically distant from each other. This algorithm is based on the waveform relaxation method coupled with a sequential solver for differential equations systems. With respect to the standard PVODE algorithm (Parallel Variable-coefficient Ordinary Differential Equations solver; Byrne, George, and Hindmars 1999), it drastically reduces the number of messages exchanged between nodes which makes it less sensitive to slow communications. Thus, it is a coarse-grained algorithm well suited for grid environments connected via high latency networks. In this paper, we present various experiments which compare the PVODE solver and our algorithm and which show the benefits brought by this work.

Investigating Performance of XML Web Services in Real-Time Business Systems

Service-Oriented Architecture (SOA) is being one of the widely accepted methodologies in software market for building and integrating different kinds of software systems. This acceptance comes from the extreme benefits that it offers to their adopters including agility, dynamicity, and loose-coupling. These benefits are usually missed in traditional software terminologies and practices. XML Web Services is the most used technology for applying SOA because it is easy to use and it allows high interoperability between different systems due to its dependency on standards that are widely accepted and supported by almost all large software vendors. However, XML Web Services suffer from a number of drawbacks such as low performance, bad utilization for hardware resources, and high network latency. These pitfalls may prevent some adopters from utilizing SOA in large and complex systems. So, these issues should be first addressed and resolved before leveraging it into real-time systems. This paper presents an experimental evaluation for the performance of XML Web Services in real-time business systems. Furthermore, it offers some tactics and strategies that might be used to enhance the overall performance of XML Web Services.

Hand Motion Prediction for Distributed Virtual Environments

We use our hands to manipulate objects in our daily life. The hand is capable of accomplishing diverse tasks such as pointing, gripping, twisting and tearing. However, there is not much work that considers using the hand as input in distributed virtual environments (DVEs), in particular over the Internet. The main reasons are that the Internet suffers from high network latency, which affects interaction, and the hand has many degrees of freedom, which adds additional challenges to synchronizing the collaboration. In this paper, we propose a prediction method specifically designed for human hand motion to address the network latency problem in DVEs. Through a thorough analysis of finger motion, we have identified various finger motion constraints and we propose a constraint-based motion prediction method for hand motion. To reduce the average prediction error under high network latency, e.g., over the Internet, we further propose a revised dead reckoning scheme here. Our performance results show that the proposed prediction method produces a lower prediction error than some popular methods while the revised dead reckoning scheme produces a lower average prediction error than the traditional dead reckoning scheme, in particular at high network latency.

Web Service versus Distributed Objects

Web services are promoted as a new model for distributed systems, yet many skeptics see them as simply a poor implementation of traditional remote procedure calls (RPC) or distributed objects. Previous comparisons, which show Web services to be significantly slower than RPC, are biased because they use a RPC communication style and ignore the document-oriented style that is recommended for Web services. We compare the straightforward design and implementation of a small file server application implemented using proxy-based RMI and document-oriented Web services. We show that Web services outperform RMI when accessing multiple/deeply nested files, especially over high-latency networks. However, the automatically generated Web service interfaces are awkward to use, so we develop a technique for wrapping the Web service to make it as easy to use as the distributed object implementation. This case study provides a more detailed comparison of the relationship between Web services and distributed objects.

Simplifying Ajax-Style Web Development

Conventional browser-based Web applications require the user to submit a request to the server, wait for the server to process the request and generate a response, and then wait for the browser to update the interface with the results. This request-wait-response-wait pattern is extremely disruptive and lowers productivity. High network latency and interface complexity and slow server responsiveness can further impair the user experience, resulting in decreased customer satisfaction, shorter and less frequent Web site visits, and, ultimately, reduced revenue to e-businesses. In this article the author describes asynchronous JavaScript and XML, a standards-based programming technique designed to make Web-based applications more responsive, interactive, and customizable - in short, to recreate the seamless user experience of most other desktop applications.

A direct Schur–Fourier decomposition for the efficient solution of high‐order Poisson equations on loosely coupled parallel computers

AbstractIn this paper a parallel direct Schur–Fourier decomposition (DSFD) algorithm for the direct solution of arbitrary order discrete Poisson equations on parallel computers is proposed. It is based on a combination of a Direct Schur method and a Fourier decomposition and allows to solve each Poisson equation almost to machine accuracy using only one communication episode. Thus, it is well suited for loosely coupled parallel computers, that have a high network latency compared with the CPU performance. Several three‐dimensional direct numerical simulations (DNS) of wall‐bounded turbulent incompressible flows have been carried out using the DSFD algorithm. Numerical examples illustrating the robustness and scalability of the method on a PC cluster with a conventional 100 Mbits/s network are also presented. Copyright © 2005 John Wiley & Sons, Ltd.

Atmospheric sounding with CHAMP: GPS ground station data for occultation processing

The double difference technique is the standard processing method of challenging minisatellite payload (CHAMPs), global positioning system (GPS) occultation data to correct for satellite clock errors. In order to apply this technique, the implementation of a global fiducial GPS ground network is required. This network (“High Rate and Low Latency Network”) was installed jointly by GeoForschungsZentrum Potsdam (GFZ) and Jet Propulsion Laboratory (JPL) in preparation of the CHAMP occultation experiment and is operated jointly by both institutes. Presently (June 2001) it consists of 28 stations (18 funded and operated by JPL, 10 by GFZ). Aspects of utilizing the ground station data for the GPS occultation processing are discussed. The network configuration allows for occultation data processing with about ∼3.5 ground stations per occultation event. The global distribution of that redundancy is found to be irregular. The network fulfills low latency requirements imposed by numerical weather prediction (NWP) systems. For the first time reduced acquisition rates of 1/5, 1/10 and 1/30 Hz compared to the current standard 1 Hz were applied to GPS occultation data processing. For the three resulting sets of 1400 vertical dry temperature profiles (using 1/5, 1/10 and 1/30 Hz, respectively) it was shown that the mean and standard deviation of the dry temperatures in relation to corresponding meteorological analyses are almost identical in relation to those of a referencing 1 Hz data set.

Grid-enabled particle physics event analysis: experiences using a 10 Gb, high-latency network for a high-energy physics application

This paper examines issues encountered attempting to exploit a high-bandwidth, high-latency link in support of a high-energy physics (HEP) analysis application. The primary issue was that the TCP additive increase/multiplicative decrease (AIMD) algorithm is not suitable for “long fat networks”. While this is a known problem, the magnitude of the impact on application performance was much greater than anticipated. We were able to overcome much of the impact, by altering the AIMD coefficients. Such an approach, of course, is non-TCP compliant, and there was insufficient time to test the network friendliness of these modifications.

High-level trigger system for the LHC ALICE experiment

Abstract The central detectors of the ALICE experiment at LHC will produce a data size of up to 75 MB / event at an event rate ⩽200 Hz resulting in a data rate of ∼15 GB / s . Online processing of the data is necessary in order to select interesting (sub)events (“High Level Trigger”), or to compress data efficiently by modeling techniques. Processing this data requires a massive parallel computing system (High Level Trigger System). The system will consist of a farm of clustered SMP-nodes based on off-the-shelf PCs connected with a high bandwidth low latency network.

BuddyCache

Collaborative applications provide a shared work environment for groups of networked clients collaborating on a common task. They require strong consistency for shared persistent data and efficient access to fine-grained objects. These properties are difficult to provide in wide area networks because of high network latency. BuddyCache is a new transactional caching approach that improves the latency of access to shared persistent objects for collaborative strong-consistency applications in high-latency network environments. The challenge is to improve performance while providing the correctness and availability properties of a transactional caching protocol in the presence of node failures and slow peers.We have implemented a BuddyCache prototype and evaluated its performance. Analytical results, confirmed by measurements of the BuddyCache prototype using the multi-user 007 benchmark indicate that for typical Internet latencies, e.g. ranging from 40 to 80 milliseconds round trip time to the storage server, peers using BuddyCache can reduce by up to 50% the latency of access to shared objects compared to accessing the remote servers directly.

Traffic management techniques to face the effects of intrinsic delays in geostationary satellite networks

A new scheme for the management of real-time traffic over high-latency broadband satellite networks is described. Early studies on this topic have been reported by Iera, Molinaro and Marano (see IEEE J. Select Areas Commun.,, vol.18, p.2393-2403, 2000), with reference to an integrated terrestrial-satellite platform. In this paper, further enhancement is introduced into the traffic management scheme with a view to both the reduction of intrinsic impairments caused by the adverse operational environment and the achievement of better performance levels and QoS guarantees. A real-time traffic handling strategy, including distributed connection admission control (CAC) and traffic resource management (TRM) schemes, is harmonized with an in-band signaling technique for burst-based bandwidth request and with an effective policy for the allocation of radio resources. Furthermore, the impact of traffic reshaping at the satellite terminal on the CAC-TRM technique is tested. Added features reduce the adverse effects of the long propagation delay across the satellite link and show outstanding effectiveness in improving CAC-TRM performance and network resource efficiency, while matching service quality requirements. The main reason for the improved performance of the overall management scheme is the adaptability to different traffic profiles conveyed over satellite links.

Optimizing Remote File Access for Parallel and Distributed Network Applications

This paper presents a paradigm for remote file access called Smart File Objects (SFOs). The SFO is a realization of the ELFS (Extensible File Systems) concept of files as typed objects, but it is applied to wide-area networks (J. Karpovich et al., in “Proceedings of the 9th OOPLSA,” 1994). The SFO is an object-oriented application- specific file access paradigm designed to address the bottleneck imposed by high latency, low bandwidth, unpredictable, and unreliable networks such as the current Internet. Newly emerging network applications such as multimedia, metacomputing, and collaboratories will have different sensitivities to these network “features.” These applications will require a more flexible file access mechanism than what is provided by conventional distributed file systems. The SFO uses application and network information to adaptively prefetch and cache needed data in parallel with the execution of the application to mitigate the impact of the network. Preliminary results indicate that the SFO can provide substantial performance gains for network applications.

More efficient serialization and RMI for Java

In current Java implementations, Remote Method Invocation (RMI) is too slow, especially for high-performance computing. RMI is designed for wide-area and high-latency networks, it is based on a slow object serialization, and it does not support high-performance communication networks. The paper demonstrates that a much faster drop-in RMI and an efficient drop-in serialization can be designed and implemented completely in Java without any native code. Moreover, the re-designed RMI supports non-TCP/IP communication networks, even with heterogeneous transport protocols. We demonstrate that for high-performance computing some of the official serialization's generality can and should be traded for speed. As a by-product, a benchmark collection for RMI is presented. On PCs connected through Ethernet, the better serialization and the improved RMI save a median of 45% (maximum of 71%) of the runtime for some set of arguments. On our Myrinet-based ParaStation network (a cluster of DEC Alphas) we save a median of 85% (maximum of 96%), compared to standard RMI, standard serialization, and Fast Ethernet; a remote method invocation runs as fast as 80 μs round trip time, compared with about 1.5 ms. Copyright © 2000 John Wiley & Sons, Ltd.

More efficient serialization and RMI for Java

In current Java implementations, Remote Method Invocation (RMI) is too slow, especially for high-performance computing. RMI is designed for wide-area and high-latency networks, it is based on a slow object serialization, and it does not support high-performance communication networks. The paper demonstrates that a much faster drop-in RMI and an efficient drop-in serialization can be designed and implemented completely in Java without any native code. Moreover, the re-designed RMI supports non-TCP/IP communication networks, even with heterogeneous transport protocols. We demonstrate that for high-performance computing some of the official serialization's generality can and should be traded for speed. As a by-product, a benchmark collection for RMI is presented. On PCs connected through Ethernet, the better serialization and the improved RMI save a median of 45% (maximum of 71%) of the runtime for some set of arguments. On our Myrinet-based ParaStation network (a cluster of DEC Alphas) we save a median of 85% (maximum of 96%), compared to standard RMI, standard serialization, and Fast Ethernet; a remote method invocation runs as fast as 80 μs round trip time, compared with about 1.5 ms. Copyright © 2000 John Wiley & Sons, Ltd.

NuMesh: An architecture optimized for scheduled communication

The NuMesh system defines a high-speed communication substrate optimized for off-line routing. By determining possible communication paths at compile time, highly efficient hardware and software constructs can be exploited to yield superior network performance. These communication paths can be independently tuned to allow more utilized paths greater bandwidth. Although communication paths are scheduled, data need not be sent during every scheduled cycle. Flow-control protocols allow for empty communication cycles as well as for data backup in the network. Limited gate delays between NuMesh registers as well as single-cycle message transfers allow for a high clock frequency and low network latency. A highly pipelined architecture for this communication is presented and a mechanism for efficient flow-controlled communication is discussed. A unique communication protocol is presented and shown to provide single-cycle transfers between nodes. An overview of the necessary compiler support is also provided. Preliminary results and a description of the current hardware and software status are listed.