High Network Latency Research Articles

Object-NoSQL Database Mappers: a benchmark study on the performance overhead

In recent years, the hegemony of traditional relational database management systems (RDBMSs) has declined in favour of non-relational databases (NoSQL). These database technologies are better adapted to meet the requirements of large-scale (web) infrastructures handling Big Data by providing elastic and horizontal scalability. Each NoSQL technology however is suited for specific use cases and data models. As a consequence, NoSQL adopters are faced with tremendous heterogeneity in terms of data models, database capabilities and application programming interfaces (APIs). Opting for a specific NoSQL database poses the immediate problem of vendor or technology lock-in. A solution has been proposed in the shape of Object-NoSQL Database Mappers (ONDMs), which provide a uniform abstraction interface for different NoSQL technologies.Such ONDMs however come at a cost of increased performance overhead, which may have a significant economic impact, especially in large distributed setups involving massive volumes of data.In this paper, we present a benchmark study quantifying and comparing the performance overhead introduced by Object-NoSQL Database Mappers, for create, read, update and search operations. Our benchmarks involve five of the most promising and industry-ready ONDMs: Impetus Kundera, Apache Gora, EclipseLink, DataNucleus and Hibernate OGM, and are executed both on a single node and a 9-node cluster setup.Our main findings are summarised as follows: (i) the introduced overhead is substantial for database operations in-memory, however on-disk operations and high network latency result in a negligible overhead, (ii) we found fundamental mismatches between standardised ONDM APIs and the technical capabilities of the NoSQL database, (iii) search performance overhead increases linearly with the number of results, (iv) DataNucleus and Hibernate OGM’s search overhead is exceptionally high in comparison to the other ONDMs.

A TestBed for Deployment of Datacenter Switches for Training Purposes

Today’s network industry needs highly qualified engineers to understand, configure, develop and upgrade switches that can build scalable high performance and ultra-low latency networks. This paper describes a new networking laboratory that has been set up to provide hands-on experience on datacenter network switches, such as ARISTA, programming and monitoring the traffic. This paper explains the networking laboratory coursework that students carry out during the course of a semester. The laboratory consists of six hands on experiments on the latest datacenter switches from Arista Networks, two programming assignments that teaches some of the protocols used at data link layer and routing protocols used at the network layer of OSI reference model. The final two laboratory experiments use the Wireshark software tool for traffic monitoring and peeking into details of some of the protocols in TCP/IP such as ARP protocol, Ethernet and so forth. The coursework details include the switch basics, switch hardware, Extensible Operating System (EOS), configuration of Link Aggregation Control Protocol, Multi-chassis Link Aggregation Protocol, Access Control Lists, and Open Short Path First version 2 on Arista switches 7050T-64 and 7048T-4S. The programming tasks cover High-level Data Link Control (HDLC) and Routing Algorithms.

Implementation of the Algorithm for Congestion control in the Dynamic Circuit Network (DCN)

Transport Control Protocol (TCP) incast congestion happens when a number of senders work in parallel with the same server where the high bandwidth and low latency network problem occurs. For many data center network applications such as a search engine, heavy traffic is present on such a server. Incast congestion degrades the entire performance as packets are lost at a server side due to buffer overflow, and as a result, the response time becomes longer. In this work, we focus on TCP throughput, round-trip time (RTT), receive window and retransmission. Our method is based on the proactive adjust of the TCP receive window before the packet loss occurs. We aim to avoid the wastage of the bandwidth by adjusting its size as per the number of packets. To avoid the packet loss, the ICTCP algorithm has been implemented in the data center network (ToR).

A non-blocking wavelength routing ONoC based on two-dimension bus architecture

With the improvement of silicon-based optical devices and on-chip optical technologies, optical network-on-chip (ONoC) is becoming a significant interconnection solution for its high bandwidth, low network latency and efficient energy utilization. Some bus-based ONoCs face the problems of high bus congestion, low network utilization, which leads to high network latency and an extra overhead in power dissipation. In this paper, a non-blocking wavelength routing ONoC based on two-dimension bus architecture (2DWR-bus) is proposed to solve the problem face by previous bus-based ONoCs, realize multiple IP cores communicating with the same destination IP core simultaneously. The network simulation is carried out for the 16 cores and 64 cores ONoC under synthetic traffics. The end-to-end (ETE) delay and saturation throughput performance are evaluated and compared between 2DWR-bus and similarly-configured ONoCs. Netrace is used in the simulation to evaluate the network performance under realistic scientific application benchmarks. The insertion loss and required laser power for 2DWR-bus is calculated and made a comparison. The evaluation result shows that 2DWR-bus ONoC has better network performance when compared with other equivalent ONoCs, especially under high network offered load.

Parallelizing query optimization on shared-nothing architectures

Data processing systems offer an ever increasing degree of parallelism on the levels of cores, CPUs, and processing nodes. Query optimization must exploit high degrees of parallelism in order not to gradually become the bottleneck of query evaluation. We show how to parallelize query optimization at a massive scale. We present algorithms for parallel query optimization in left-deep and bushy plan spaces. At optimization start, we divide the plan space for a given query into partitions of equal size that are explored in parallel by worker nodes. At the end of optimization, each worker returns the optimal plan in its partition to the master which determines the globally optimal plan from the partition-optimal plans. No synchronization or data exchange is required during the actual optimization phase. The amount of data sent over the network, at the start and at the end of optimization, as well as the complexity of serial steps within our algorithms increase only linearly in the number of workers and in the query size. The time and space complexity of optimization within one partition decreases uniformly in the number of workers. We parallelize single- and multi-objective query optimization over a cluster with 100 nodes in our experiments, using more than 250 concurrent worker threads (Spark executors). Despite high network latency and task assignment overheads, parallelization yields speedups of up to one order of magnitude for large queries whose optimization takes minutes on a single node.

An Architecture of IoT Service Delegation and Resource Allocation Based on Collaboration between Fog and Cloud Computing

Despite the wide utilization of cloud computing (e.g., services, applications, and resources), some of the services, applications, and smart devices are not able to fully benefit from this attractive cloud computing paradigm due to the following issues: (1) smart devices might be lacking in their capacity (e.g., processing, memory, storage, battery, and resource allocation), (2) they might be lacking in their network resources, and (3) the high network latency to centralized server in cloud might not be efficient for delay-sensitive application, services, and resource allocations requests. Fog computing is promising paradigm that can extend cloud resources to edge of network, solving the abovementioned issue. As a result, in this work, we propose an architecture of IoT service delegation and resource allocation based on collaboration between fog and cloud computing. We provide new algorithm that is decision rules of linearized decision tree based on three conditions (services size, completion time, and VMs capacity) for managing and delegating user request in order to balance workload. Moreover, we propose algorithm to allocate resources to meet service level agreement (SLA) and quality of services (QoS) as well as optimizing big data distribution in fog and cloud computing. Our simulation result shows that our proposed approach can efficiently balance workload, improve resource allocation efficiently, optimize big data distribution, and show better performance than other existing methods.

An integrated solution for remote data access

Data management constitutes one of the major challenges that a geographically- distributed e-Infrastructure has to face, especially when remote data access is involved. We discuss an integrated solution which enables transparent and efficient access to on-line and near-line data through high latency networks. The solution is based on the joint use of the General Parallel File System (GPFS) and of the Tivoli Storage Manager (TSM). Both products, developed by IBM, are well known and extensively used in the HEP computing community. Owing to a new feature introduced in GPFS 3.5, so-called Active File Management (AFM), the definition of a single, geographically-distributed namespace, characterised by automated data flow management between different locations, becomes possible. As a practical example, we present the implementation of AFM-based remote data access between two data centres located in Bologna and Rome, demonstrating the validity of the solution for the use case of the AMS experiment, an astro-particle experiment supported by the INFN CNAF data centre with the large disk space requirements (more than 1.5 PB).

Darkcube: A k-Hypercube based P2P VoIP protocol

We propose a new peer-to-peer voice-over-IP (VoIP) protocol, called Darkcube, which builds upon a k-hypercube topology presenting interesting properties regarding multi-party VoIP conferencing. This new topology is based on multiple hypercubes isolating the speaker nodes into a main high speed hypercube and efficiently retransmitting to the remaining listener nodes guaranteeing a good conversational voice communication quality among all users and achieving a great overall user experience. Standard hypercube topologies can present certain performance problems when the number of participants is high; our proposal isolates speakers within a main, low delay, hypercube retransmitting pre-mixed audio to the remaining listening hypercubes improving overall performance and supporting more nodes with better quality than Skype under similar conditions. It is tolerant to common problems on current wireless networks such as high network latencies or temporary network outages, as well as malfunctioning or malicious nodes, making it well suited for its operation on mobile devices.

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.

The web-based user interface for EAST plasma control system

The plasma control system (PCS) plays a vital role at EAST for fusion science experiments. Its software application consists of two main parts: an IDL graphical user interface for setting a large number of plasma parameters to specify each discharge, several programs for performing the real-time feedback control and managing the whole control system. The PCS user interface can be used from any X11 Windows client with privileged access to the PCS computer system. However, remote access to the PCS system via the IDL user interface becomes an extreme inconvenience due to the high network latency to draw or operate the interfaces. In order to realize lower latency for remote access to the PCS system, a web-based system has been developed for EAST recently. The setup data are retrieved from the PCS system and client-side JavaScript draws the interfaces into the user's browser. The user settings are also sent back to the PCS system for controlling discharges. These technologies allow the web-based user interface to be viewed by authorized users with a web browser and have it communicate with PCS server processes directly. It works together with the IDL interface and provides a new way to aid remote participation.

Scalable Cloudlet-based Mobile Computing Model

Mobile Cloud Computing (MCC) has been introduced as a feasible solution to the inherited limitations of mobile computing. These limitations include battery lifetime, processing power and storage capacity. By using MCC, the processing and the storage of intensive mobile device jobs will take place in the cloud system and the results will be returned to the mobile device. This reduces the required power and time for completing such intensive jobs. However, connecting mobile devices with the cloud, suffers from the high network latency and the huge transmission power consumption especially when using 3G/LTE connections. On the other hand, multimedia applications are the most common applications in today's mobile devices; such applications require high computing resources. In this paper, a Cloudlet-based MCC system is introduced, aiming at reducing the power consumption and the network delay of multimedia applications while using MCC. The MCC concepts with the proposed Cloudlet framework are integrated and a new scalable framework for the MCC model is proposed. Our practical experimental results using multimedia applications show that using the proposed model reduces the power consumption of the mobile devices as well as reducing the communication latency when the mobile device requests a job to be performed remotely while satisfying the high quality of service requirements.

A novel resource selection framework to improve QoS in computational grid

The computational grid is organised as a collection of heterogeneous resources to solve complex computational problems in scientific domains. The fundamental issue to be addressed here is the selection of appropriate compute and network resources for every submitted job. We present a novel resource selection framework, which selects trusted resources across multiple resource sites based on the computation and network capacity of the resource. We developed a new scheme for efficient data transfer in high latency wide area networks. To efficiently utilise the bandwidth of the connected resources, an improvement over the existing TCP is proposed. TCP striping aims at minimising the communication time and maximising the network utilisation. We have solved the problem of finding the optimal number of stripes considering the available bandwidth and data file size. Simulation results have shown that the proposed work completes the submitted jobs with a minimum communication time and execution time. Also, there is a significant improvement of bandwidth utilisation in high bandwidth high latency networks.

Service Oriented Robotic Architecture for Space Robotics: Design, Testing, and Lessons Learned

This paper presents the lessons learned from six years of experiments with planetary rover prototypes running the Service Oriented Robotic Architecture (SORA) developed by the Intelligent Robotics Group (IRG) at the NASA Ames Research Center. SORA relies on proven software engineering methods and technologies applied to space robotics. Based on a service oriented architecture and robust middleware, SORA encompasses onboard robot control and a full suite of software tools necessary for remotely operated exploration missions. SORA has been field‐tested in numerous scenarios of robotic lunar and planetary exploration. The experiments conducted by IRG with SORA exercise a large set of the constraints encountered in space applications: remote robotic assets, flight‐relevant science instruments, distributed operations, high network latencies, and unreliable or intermittent communication links. In this paper, we present the results of these field tests in regard to the developed architecture, and we discuss its benefits and limitations.

Highly available transactions

To minimize network latency and remain online during server failures and network partitions, many modern distributed data storage systems eschew transactional functionality, which provides strong semantic guarantees for groups of multiple operations over multiple data items. In this work, we consider the problem of providing Highly Available Transactions (HATs): transactional guarantees that do not suffer unavailability during system partitions or incur high network latency. We introduce a taxonomy of highly available systems and analyze existing ACID isolation and distributed data consistency guarantees to identify which can and cannot be achieved in HAT systems. This unifies the literature on weak transactional isolation, replica consistency, and highly available systems. We analytically and experimentally quantify the availability and performance benefits of HATs---often two to three orders of magnitude over wide-area networks---and discuss their necessary semantic compromises.

An efficient secure authentication scheme with user anonymity for roaming user in ubiquitous networks

Recently the advances in a variety of hardware/software technologies and wireless networking have brought explosive growth of mobile network/device involved services and it raised a number of security concerns such as user authentication/authorization, user information privacy and secure data transmission between service providers and consumers using mobile devices through mobile network. Therefore, it is essential that legitimate user authentication as well as user anonymity for privacy is ensured between those entities (service providers and consumers). As compared to stationary devices, due to high network latency and low computing performance of mobile network/devices, the authentication scheme in wireless and mobile networks requires a mechanism to achieve low communication and computation costs. Thus, in this paper we propose a lightweight authentication scheme. Also, the scheme ensures user’s anonymity and provides secure password update, session key agreement, and mutual authentication between entities to resist to possible attacks in ubiquitous networks. And we present security and performance analysis of our proposed scheme, and experimental results to show the efficiency of the scheme as well.

An interactive pervasive whiteboard based on MVC architecture for ubiquitous collaboration

Collaboration is about interactions among multiple users by sharing resources. With the advent of new generation of mobile access devices such as smartphone and tablet PC, we address this for ubiquitous collaboration--the capability of multiple users to link together with disparate access devices in anytime and in anywhere. The implementation and experiments of mobile applications for ubiquitous collaboration were challenge since cell phone (also referred to feature phone) prior to smartphone had high network latency and low computing performance. However, even though mobile device technologies and wireless networking are becoming more advanced with time, the research on the impact of ubiquitous collaboration is still immature. This paper extends our prior work with interactive multimedia services--whiteboard application on 3G and Wi-Fi Android platform based mobile devices, based on MVC (Model View Controller) architecture for ubiquitous collaboration. Then, we present interaction and usability study with multiple users using (wire, wireless) small-sized devices (smartphones), mid-sized devices (tablet PCs), and large-sized device (desktops) to show how disparate access devices and networking have an effect on software design of shared applications in ubiquitous collaboration system.

Efficient and Secure Web Services by using Multi Agents

Mobile agents are an excellent technology for implementing Web services. Within a set of federated Web services, mobile agents can reduce bandwidth requirements and mitigate the effects of high-latency network connections. This paper presents a model for implementing Web services with mobile agents where agents are free to move between cooperating Web servers to implement the service functionality. Also for increasing security of web services, we illustrate a novel distributed protocol for multi agent environments. In this approach, the encrypted private key and the message are broken into different parts carrying by different agents, which make it difficult for malicious entities to mine the private key for message encryption, while the private key for the encrypted key is allocated on the predetermined destination nodes. On the other hand, all of the previously proposed encryption algorithms can be applied in the proposed approach that deteriorates the key discovery process. To improve the overall security, the paper makes use of Advanced Encryption Standard (AES) as the encryption base for message encryption. Our mobile agent Web services present typical WSDL interfaces, so mobile agent functionality can be consumed from legacy clients, and federated services can be gradually migrated to a mobile agent implementation. General Terms Multi agent, Cryptographic Protocol, Mobile Agents, Security, and Web services et. al.

PocketWeb

The high network latencies and limited battery life of mobile phones can make mobile web browsing a frustrating experience. In prior work, we proposed trading memory capacity for lower web access latency and a more convenient data transfer schedule from an energy perspective by prefetching slowly-changing data (search queries and results) nightly, when the phone is charging. However, most web content is intrinsically much more dynamic and may be updated multiple times a day, thus eliminating the effectiveness of periodic updates. This paper addresses the challenge of prefetching dynamic web content in a timely fashion, giving the user an instant web browsing experience but without aggravating the battery lifetime issue. We start by analyzing the web access traces of 8,000 users, and observe that mobile web browsing exhibits a strong spatiotemporal signature, which is different for every user. We propose to use a machine learning approach based on stochastic gradient boosting techniques to efficiently model this signature on a per user basis. The machine learning model is capable of accurately predicting future web accesses and prefetching the content in a timely manner. Our experimental evaluation with 48,000 models trained on real user datasets shows that we can accurately prefetch 60% of the URLs for about 80-90% of the users within 2 minutes before the request. The system prototype we built not only provides more than 80% lower web access time for more than 80% of the users, but it also achieves the same or lower radio energy dissipation by more than 50% for the majority of mobile users.

PocketWeb

The high network latencies and limited battery life of mobile phones can make mobile web browsing a frustrating experience. In prior work, we proposed trading memory capacity for lower web access latency and a more convenient data transfer schedule from an energy perspective by prefetching slowly-changing data (search queries and results) nightly, when the phone is charging. However, most web content is intrinsically much more dynamic and may be updated multiple times a day, thus eliminating the effectiveness of periodic updates. This paper addresses the challenge of prefetching dynamic web content in a timely fashion, giving the user an instant web browsing experience but without aggravating the battery lifetime issue. We start by analyzing the web access traces of 8,000 users, and observe that mobile web browsing exhibits a strong spatiotemporal signature, which is different for every user. We propose to use a machine learning approach based on stochastic gradient boosting techniques to efficiently model this signature on a per user basis. The machine learning model is capable of accurately predicting future web accesses and prefetching the content in a timely manner. Our experimental evaluation with 48,000 models trained on real user datasets shows that we can accurately prefetch 60% of the URLs for about 80-90% of the users within 2 minutes before the request. The system prototype we built not only provides more than 80% lower web access time for more than 80% of the users, but it also achieves the same or lower radio energy dissipation by more than 50% for the majority of mobile users.

Game-on-demand:

In recent years, online gaming has become very popular. In contrast to stand-alone games, online games tend to be large-scale and typically support interactions among users. However, due to the high network latency of the Internet, smooth interactions among the users are often difficult. The huge and dynamic geometry data sets also make it difficult for some machines, such as handheld devices, to run those games. These constraints have stimulated some research interests on online gaming, which may be broadly categorized into two areas: technological support and user-perceived visual quality . Technological support concerns the performance issues while user-perceived visual quality concerns the presentation quality and accuracy of the game. In this article, we propose a game-on-demand engine that addresses both research areas. The engine distributes game content progressively to each client based on the player's location in the game scene. It comprises a two-level content management scheme and a prioritized content delivery scheme to help identify and deliver relevant game content at appropriate quality to each client dynamically. To improve the effectiveness of the prioritized content delivery scheme, it also includes a synchronization scheme to minimize the location discrepancy of avatars (game players). We demonstrate the performance of the proposed engine through numerous experiments.