Processing large scientific applications generates a huge amount of data, which makes running experiments in the cloud computing environment very expensive and energy-consuming. To find an optimal solution to the workflow scheduling problem, several approaches have been presented for scheduling workflow on cloud resources. However, more efficient approaches are needed to improve cloud service delivery. In this paper, an energy-efficient virtual machine mapping algorithm (EViMA) is proposed to improve resource management in the cloud computing environment to achieve effective scheduling that reduces cloud data center energy consumption, execution makespan, and execution cost. This ensures that the requirements of cloud users are met, and improves the quality of services offered by cloud providers. Our proposed mechanism considers the heterogeneity of scheduling from both cloud users’ and workflow applications’ perspectives. Through simulation experiments on real workflow datasets, the proposed EViMA can provide better solutions for both cloud users and cloud providers by reducing energy consumption, execution makespan, and execution cost better than the state-of-the-art.

Large data centers are currently the mainstream infrastructures for big data processing. As one of the most fundamental tasks in these environments, the efficient execution of distributed data operators (e.g., join and aggregation) are still challenging current data systems, and one of the key performance issues is network communication time. State-of-the-art methods trying to improve that problem focus on either application-layer data locality optimization to reduce network traffic or on network-layer data flow optimization to increase bandwidth utilization. However, the techniques in the two layers are totally independent from each other, and performance gains from a joint optimization perspective have not yet been explored. In this article, we propose a novel approach called NEAL (NEtwork-Aware Locality scheduling) to bridge this gap, and consequently to further reduce communication time for distributed big data operators. We present the detailed design and implementation of NEAL, and our experimental results demonstrate that NEAL always performs better than current approaches for different workloads and network bandwidth configurations.

Processing big data generated in large Internet of Things (IoT) networks is challenging current techniques. To date, a lot of network clustering approaches have been proposed to improve the performance of data collection in IoT. However, most of them focus on partitioning networks with static topologies, and thus they are not optimal in handling the case with moving objects in the networks. Moreover, to the best of our knowledge, none of them has ever considered the performance of computing in edge servers. To solve these problems, we propose a highly efficient IoT network dynamic clustering solution in edge computing using deep reinforcement learning (DRL). Our approach can both fulfill the data communication requirements from IoT networks and load-balancing requirements from edge servers, and thus provide a great opportunity for future high performance IoT data analytics. We implement our approach using a Deep Q-learning Network (DQN) model, and our preliminary experimental results show that the DQN solution can achieve higher scores in cluster partitioning compared with the current static benchmark solution.

Big Data technologies and tools have being used for the past decade to solve several scientific and industry problems, with Hadoop/YARN becoming the “de facto” standard for these applications, although other technologies run on top of it. As any other distributed application, those big data technologies rely heavily on the network infrastructure to read and move data from hundreds or thousands of cluster nodes. Although these technologies are based on reliable and efficient distributed algorithms, there are scenarios and conditions that can generate bottlenecks and inefficiencies, i.e., when a high number of concurrent users creates data access contention. In this paper, we propose a novel network topology called Multi-Layer-Mesh and a path switching algorithm based on SDN, that can increase the performance of a big data cluster while reducing the amount of utilized resources (network equipment), in turn reducing the energy and cooling consumption. A thorough simulation-based evaluation of our algorithms shows an average improvement in performance of 31.77% and an average decrease in resource utilization of 36.03% compared to a traditional Spine-Leaf topology, in the selected test scenarios.

Process models as knowledge graph representation have been widely used in various domains to create products and deliver services. Although different process model discovery approaches have been proposed in recent years, few of them are designed for distributed computing environments. Specifically, none of them has been studied in the emerging edge computing application scenarios. In this paper, based on the requirements of some real-time process services, we propose a system design for learning process models in IoT edge. We present the details of our solution and our preliminary results on a simulated IoT network show that our method can discover real-time process models in less than a second.

Periodic beacon messages are one of the building blocks that enable the operation of vehicular ad hoc networks (VANETs) applications. In vehicular networks environments, congestion and awareness control mechanisms are key for a reliable and efficient functioning of vehicular applications. In order to control the channel load, a reliable mechanism allowing real-time measurements of parameters like the local density of vehicles is a must. These measurements can then serve as an input to perform a fast adaptation of the transmit parameters. In this paper, considerable efforts have been directed in the recent years toward designing flexible yet robust protocols solving this problem; yet, very few have considered a proactive adaptation of the transmit parameters as a preventive measure from channel load peaks. To this end, we take the opportunity to introduce prediction and adaptation algorithm (P&A-A), a new congestion control protocol that performs a joint adaptation of the transmit rate and power, relying on an altruistic short-term prediction algorithm that estimates the vehicular density around a given vehicle within the next short while. Additionally, P&A-A adapts the transmit parameters in a way that guarantees the strict beaconing requirements and satisfies the level of awareness required for the operation of most critical VANET applications. The results of the simulations performed in a realistic scenario justify our theoretical considerations and confirm the efficiency and the effectiveness of our protocol by showing significant improvements in terms of network performance (up to 8% and 14% improvement in collision rate; and up to 10% and 20% increase in busy ratio compared to our previous scheme and the ETSI schemes, respectively) as well as the achieved level of awareness (higher coverage with higher transmission rate and power in dense scenarios, and up to 8% and 55% improvement in density perception accuracy compared to our previous scheme and the ETSI schemes, respectively).

Big Data platforms are convoluted distributed systems which commonly comprise skill- and labour-intensive solution development to treat inherent Big Data application challenges. Several tools have been proposed to help developers and engineers to overcome the involved complexities in coordinating the execution of plenty processes/threads on multiple machines. However, no work so far has been able to combine both an accurate representation of Big Data jobs and realistic modeling of the behaviour of Big Data platforms at scale, including networking elements and data and job placement. In this paper, we propose BigDataNetSim, the first simulator which models accurately all the main components of the data movements in Big Data platforms (e.g., HDFS, YARN/MapReduce, network topologies, switching/routing protocols) in a large scale system. BigDataNetSim can serve as a valuable tool for engineering Big Data solutions, which includes set-up of systems, prototyping of jobs, and improvement of components/algorithms for Big Data platforms. We also demonstrate that BigDataNetSim can simulate a real Hadoop cluster with a high degree of accuracy in terms of data and job placements, being able to scale up to very large systems.

The identification of workload-dependent performance issues, as well as their root causes, is a time-consuming and complex process which typically requires several iterations of tests (as this type of issues can depend on the input workloads), and heavily relies on human expert knowledge. To improve this process, this paper presents an automated approach to dynamically adapt the workload (used by a performance testing tool) during the test runs. As a result, the performance issues of the tested application can be revealed more quickly; hence, identifying them with less effort and expertise. Our experimental evaluation has assessed the accuracy of the proposed approach and the time savings that it brings to testers. The results have demonstrated the benefits of the approach by achieving a significant decrease in the time invested in performance testing (without compromising the accuracy of the test results), while introducing a low overhead in the testing environment.

Real-time interactive communications, such as videoconferencing, have strong QoS requirements. Adapting such communications to network dynamics needs network management capabilities that traditional networks cannot provide. Recently, SDN has shown to be a potential solution for solving these network adaptation challenges. In a companion paper written by our team, we have proposed a videoconferencing system using SVC multicast trees built over an SDN network. This system minimizes the overall bandwidth consumption and provides the highest possible QoS to the users although it is not adaptable to network dynamics happening over the lifetime of the communication. In this paper, we complement our system, by providing efficient algorithms able to modify the connections between the participants during a videoconference in order to timely react to network changes. The simulation results confirm the efficiency of our solution in terms of time processing and bandwidth savings.

Videoconferencing applications have strong latency requirements and consume large portions of a network's bandwidth. Current videoconferencing solutions are not efficiently implemented as they often rely on a central server and do not leverage network layering services. In this paper, we investigate the impact of using Scalable Video Coding and Software Defined Networking techniques on videoconferencing. Specifically we propose an algorithm capable of reducing the bandwidth consumed by videoconferencing using Software Defined Networking to compute and deploy multicast trees, as well as taking advantage of the Scalable Video Coding layering feature that allows to sacrifice video quality for usability purposes. We show that by smartly dropping video layers at specific locations in the network, the overall bandwidth usage by videoconferencing decreases, and in a separate capacity study we show that our solution allows networks to support more videoconferencing calls as compared to existing solutions.

The global view enabled by a Software Defined Network (SDN) architecture allows us to observe the contribution of each congested router to the overall Mean Opinion Score (MOS) degradation of a VoIP service. Thus, when a number of routers in the path may add to the MOS degradation, the SDN controller can quantify the contribution from each router. This work presents the implementation of a network monitoring system for OpenFlow protocol aware enterprise VoIP networks. The presented solution enables network administrators and developers to access on demand information about intermediate packet loss and MOS for real-time applications. The main contribution of this work is an out-of-the-box architecture to enable users to find the exact location of quality degradation with a small operational cost.

Summary The identification of performance issues and the diagnosis of their root causes are time‐consuming and complex tasks, especially in clustered environments. To simplify these tasks, researchers have been developing tools with built‐in expertise for practitioners. However, various limitations exist in these tools that prevent their efficient usage in the performance testing of clusters (e.g. the need of manually analysing huge volumes of distributed results). In a previous work, we introduced a policy‐based adaptive framework (PHOEBE) that automates the usage of diagnosis tools in the performance testing of clustered systems, in order to improve a tester's productivity, by decreasing the effort and expertise needed to effectively use such tools. This paper extends that work by broadening the set of policies available in PHOEBE, as well as by performing a comprehensive assessment of PHOEBE in terms of its benefits, costs and generality (with respect to the used diagnosis tool). The performed evaluation involved a set of experiments in assessing the different trade‐offs commonly experienced by a tester when using a performance diagnosis tool, as well as the time savings that PHOEBE can bring to the performance testing and analysis processes. Our results have shown that PHOEBE can drastically reduce the effort required by a tester to do performance testing and analysis in a cluster. PHOEBE also exhibited consistent behaviour (i.e. similar time‐savings and resource utilisations), when applied to a set of commonly used diagnosis tools, demonstrating its generality. Finally, PHOEBE proved to be capable of simplifying the configuration of a diagnosis tool. This was achieved by addressing the identified trade‐offs without the need for manual intervention from the tester. Copyright © 2017 John Wiley & Sons, Ltd.

Achieving high-quality voice conference calls over the Internet is a difficult task. Heterogeneous mobile devices and network dynamics must be properly managed by multiparty VoIP systems to ensure a good quality of experience. In this paper, we propose a multiparty VoIP system based on SDN technology that uses both multicast distribution and dynamic stream adaptation to optimize the conference call quality for each participant. We define a tree construction algorithm that includes two modes: a minimum spanning tree mode and a shortest path tree mode. We have simulated conference calls with various parameters and results show that a trade-off can be achieved between path latency and bandwidth consumption when using multicasted streams with adaptive bitrates. Compared to a typical multiparty VoIP system, up to two-third bandwidth consumption reduction is possible with only a small increase of the average path latency. Our solution could be leveraged by other types of real-time applications such as online gaming.

The Software Defined Networking (SDN) paradigm enables quick deployment of software controlled network infrastructures, however new approaches to system monitoring are required to provide network administrators with instant feedback on a network's health. This paper details the deployment of an SDN system architecture featuring the integration of a cloud-based, real-time log-analysis platform. The proposed architecture uses log data collected from host machines, OpenFlow switches and the SDN controllers in a non-invasive style. This work uses a commercially available correlation platform to provide network administrators with a real-time view of the network status and the approach is validated under two scenarios:network overload and a security attack.

The continuous growth in user demand for high-quality rich media services puts pressure on Wireless Mesh Network (WMN) resources. Solutions such as those which increase the capacity of the mesh network by equipping mesh routers with additional wireless interfaces provide better Quality of Service (QoS) for video deliveries, but result in higher overall energy consumption for the network. This paper presents LBIS, a distributed solution which combines the benefits of both load-balancing and interface-shifting in order to enhance QoS levels for video services delivered over multi-hop WMNs, while maintaining low energy consumption levels within the network. Simulation-based results show very good performance of our proposed mechanism in terms of QoS metrics (delay, packet loss), Peak Signal-to-Noise Ratio (PSNR) and energy consumption in mesh network topologies, and with varying video traffic loads and distributions. The results demonstrate how LBIS can increase the QoS for video deliveries by more than 30% at the cost of an insignificant increase of the overall network energy consumption compared to the WMN with multiple radio interfaces without the LBIS adaptation.

Summary Nowadays, clustered environments are commonly used in high‐performance computing and enterprise‐level applications to achieve faster response time and higher throughput than single machine environments. Nevertheless, how to effectively manage the workloads in these clusters has become a new challenge. As a load balancer is typically used to distribute the workload among the cluster's nodes, multiple research efforts have concentrated on enhancing the capabilities of load balancers. Our previous work presented a novel adaptive load balancing strategy (TRINI) that improves the performance of a clustered Java system by avoiding the performance impacts of major garbage collection, which is an important cause of performance degradation in Java. The aim of this paper is to strengthen the validation of TRINI by extending its experimental evaluation in terms of generality, scalability and reliability. Our results have shown that TRINI can achieve significant performance improvements, as well as a consistent behaviour, when it is applied to a set of commonly used load balancing algorithms, demonstrating its generality. TRINI also proved to be scalable across different cluster sizes, as its performance improvements did not noticeably degrade when increasing the cluster size. Finally, TRINI exhibited reliable behaviour over extended time periods, introducing only a small overhead to the cluster in such conditions. These results offer practitioners a valuable reference regarding the benefits that a load balancing strategy, based on garbage collection, can bring to a clustered Java system. Copyright © 2016 John Wiley & Sons, Ltd.

Summary As part of the process to test a new release of an application, the performance testing team need to confirm that the existing functionalities do not perform worse than those in the previous release, a problem known as performance regression anomaly. Most existing approaches to analyse performance regression testing data vary according to the applied workload, which usually leads to the need for an extra performance testing run. To ease such lengthy tasks, we propose a new workload‐independent, automated technique to detect anomalies in performance regression testing data using the concept known as transaction profile (TP). The TP is inferred from the performance regression testing data along with the queueing network model of the testing system. Based on a case study conducted against two web applications, one open source and one industrial, we have been able to automatically generate the ‘TP run report’ and verify that it can be used to uncover performance regression anomalies caused by software updates. In particular, the report helped us to isolate the real anomaly issues from those caused by workload changes with an average F1 measure of 85% for the open source application and 90% for the industrial application. Such results support our proposal to use the TP as a more efficient technique in identifying performance regression anomalies than the state of the art industry and research techniques. Copyright © 2015 John Wiley & Sons, Ltd.

Channel congestion is a well-known problem in wireless networks in general and Vehicular Ad Hoc Networks (VANETs) in particular. Literature solutions propose to alleviate this problem by controlling the network load based on parameters like vehicle density or packet collision rate. In other words, each vehicle will observe the density of vehicles (or the packet collision rate) around itself in a certain time interval, and use this information to adjust its transmit parameters i.e. transmit rate and/or power, the next time it has a beacon to transmit (in the following time window). However, the information collected in the current time window might not still be valid in the next one. In fact, in a highly dynamic network like VANETs, vehicle density, and consequently the busy ratio and the collision rate, might vary a great deal even in the smallest time intervals. To cope with this newly identified problem, we propose a novel vehicle-centric short-term density prediction scheme that estimates the vehicular density around a given vehicle within the next time window allowing each vehicle to adapt its transmit parameters based on the current state of the network (as opposed to the previous state). The accuracy and the efficiency of our proposed scheme is demonstrated in a proof-of-concept case study, showing a significant improvement in terms of network performance.

Due to the limited capacity of road networks and sporadic on-route events, road traffic congestions are posing serious problems in most big cities worldwide and resulting in considerable number of casualties and financial losses. In order to deal efficiently with these problems and alleviate their impact on individuals, environment, and economic activities, advanced traffic monitoring and control tools (e.g., SCATS and SCOOT) are being widely used in hundreds of major cities in the world. However, due to increasing road traffic and dynamic spatio-temporal events, additional proactive mechanisms remain needed to prevent traffic congestions. Within this context, we argue that the emergent V2X communication technologies, and especially V2I (Vehicle to Infrastructure), would be of great help. To this end, we investigate in this paper the opportunities that could be offered by V2I technology in improving commuters' journey duration and mitigating the above irritating and frequent problems. We then propose an approach where road-side facilities (e.g. traffic light controllers at road intersections) communicate traffic light cycle information to approaching vehicles. Based on this information, the vehicles collaboratively determine their optimal speeds and other appropriate actions to undertake in order to cross road intersections with minimum delays while ultimately avoiding stoppings. The obtained evaluation results show that our approach achieves a significant gain in terms of the commuters' average travel time reduction.

Data aggregation techniques have emerged as promising solutions for extending Wireless Sensor Networks (WSNs) lifetime. However, this approach suffers from a design issue in delivering the strict requirements needed by some monitoring applications. Carefully balancing Energy, Delay and Accuracy is essential for achieving these requirements. In this work, we focus on distributed data aggregation, where a sensor estimates the network information by the exchange of readings with different priority levels. We then propose an optimal decision policy for scheduling the transmission of the aggregated data at the node level. To model the investigated problem, we first adopt Markov Decision Process (MDP) whereby we define the reward function. Then, we apply a Genetic Algorithm (GA) to find a set of optimal decisions that ensures the best trade-off between energy saving, delay and accuracy of the received data based on their priority level. The simulation results yield excellent performance and our optimization shows a significant enhancement up to 20% compared to the other policies.

Large amounts of data often require expensive and time-consuming analysis. Therefore, highly scalable and efficient techniques are necessary to process, analyze and discover useful information. Database sampling has proven to be a powerful method to surpass these limitations. Using only a sample of the original large database brings the benefit of obtaining useful information faster, at the potential expense of lower accuracy. In this paper, we demonstrate \vfds, a novel fast database sampling system that maintains the referential integrity of the data. The system is developed over the open-source database management system, MySQL. We present various scenarios to demonstrate the effectiveness of VFDS in approximate query answering, sample size, and execution time, on both real and synthetic databases.

Data center optimization, mainly through virtual machine (VM) placement, has received considerable attention in the past years. A lot of heuristics have been proposed to give quick and reasonably good solutions to this problem. However it is difficult to compare them as they use different datasets, while the distribution of resources in the datasets has a big impact on the results. In this paper we propose the first benchmark for VM placement heuristics and we define a novel heuristic. Our benchmark is inspired from a real data center and explores different possible demographics of data centers, which makes it suitable when comparing the behaviour of heuristics. Our new algorithm, RBP, outperforms the state-of-the-art heuristics and provides close to optimal results quickly.

Modern computer applications, especially at enterprise-level, are commonly deployed with a big number of clustered instances to achieve a higher system performance, in which case single machine based solutions are less cost-effective. However, how to effectively manage these clustered applications has become a new challenge. A common approach is to deploy a front-end load balancer to optimise the workload distribution between each clustered application. Since then, many research efforts have been carried out to study effective load balancing algorithms which can control the workload based on various resource usages such as CPU and memory. The aim of this paper is to propose a new load balancing approach to improve the overall distributed system performance by avoiding potential performance impacts caused by Major Java Garbage Collection. The experimental results have shown that the proposed load balancing algorithm can achieve a significant higher throughput and lower response time compared to the round-robin approach. In addition, the proposed solution only has a small overhead introduced to the distributed system, where unused resources are available to enable other load balancing algorithms together to achieve a better system performance.

Server consolidation is an important problem in any enterprise, where capital allocators (CAs) must approve any cost saving plans involving the acquisition or allocation of new assets and the decommissioning of inefficient assets. Our paper describes iVMp an interactive VM placement algorithm, that allows CAs to become 'agile' capital allocators that can interactively propose and update constraints and preferences as placements are recommended by the system. To the best of our knowledge this is the first time that this interactive VM placement recommendation problem has been addressed in the academic literature. Our results show that the proposed algorithm finds near optimal solutions in a highly efficient manner.

More and more drivers use on-board units to help them navigate in the increasing urbanised environment they live and work in. These system (e.g., routing applications on smart phones) are now very often on-line, and use information from the traffic situation (e.g., accidents, congestion) to get the best route. We can now envisage a world where all trips are assigned and updated by such an on-line system, making the best routing decisions based on traffic conditions. The problem is that current systems consider only 'local' elements (e.g., driver preference and current traffic condition) and do not make routing decisions from a global perspective. This can lead to a lot of similar routing assignments that could lead to further traffic congestion. The objective of the next generation on-line navigation systems is then to come up with a 'smart', real-time route assignment, which balances the load between the different road segments and offers the best quality to the drivers. However, every routing decision made has an impact on the traffic conditions (one more vehicle on the road segments selected) and computing the load induced by the trips is a computationally heavy problem. This paper addresses this question of real-time on-line traffic assignment, and shows that under certain conditions it is possible to have (i) an accurate estimation of the load and travel time on every road segment and (ii) an optimised traffic assignment that adapts to divergence and evolutions (e.g., accidents) of the system.

Wireless Mesh Networks (WMNs) are becoming increasingly popular and user demand for high-quality rich media services is continuously growing. Despite the fact that WMNs offer significant flexibility, they suffer in respect to Quality of Service (QoS) provisioning. This paper proposes a novel mechanism for providing enhanced QoS support to video services in multi-hop WMNs. The mechanism makes use of an innovative hybrid hierarchical architecture which combines centralized and distributed approaches. The proposed solution relies on performance monitoring at WMN nodes and performs load balancing by off-loading traffic from the highest loaded nodes to less loaded neighbours. Simulation-based results presented outline the performance of our proposed mechanism in terms of QoS metrics (delay, throughput, packet losses and PSNR) in different network load scenarios. The results clearly demonstrate how our proposed mechanism outperforms the traditional OLSR protocol in terms of QoS performance.

Bugs in a project, at any stage of Software life cycle development are costly and difficult to find and fix.Moreover, the later a bug is found, the more expensive it is to fix.There are static analysis tools to ease the process of finding bugs, but their results are not easy to filter out critical errors and is time consuming to analyze.To solve this problem we used two steps: first to enhance the bugs severity and second is to estimate the code quality, by Weighted Error Code Density metric.Our experiment on 10 widely used open-source Java applications automatically shows their code quality estimated using our objective metric.We also enhance the error ranking of FindBugs, and provide a clear view on the critical errors to fix as well as low priority ones to potentially ignore.

Nowadays, enterprise software systems store a large amount of operational information in logs. Manually analysing these data can be time-consuming and error-prone. Although a static knowledge database eases the task to capture recurring problems, maintaining such a knowledge repository requires periodic knowledge updates by domain experts. Moreover, as the repository grows, the problem of memory efficiency will also arise.

Deployed software applications use log files to keep a record of system events. Log analysis provides support for system administrators to gain the knowledge of system health and behavior. As a result, the ability to efficiently search for patterns in historical events has become a major requirement for timely analysis. Enterprise systems today produce high volumes of log data, regularly in the order of thousands of events per second, which requires to build inverted indexes for quick data retrieval. However, current inverted indexing techniques are rarely designed to handle high volumes of dynamic stream data and often resource consuming. We propose an efficient indexing solution, which reduces the necessary resources by employing bloom filter techniques. The solution builds a generic indexing engine for the Run Time Correlation Engine logging framework to achieve efficient monitoring in the Cloud. In particular, our solution is able to deliver significant performance improvement over existing indexing engines.

A wireless mesh network is characterized by dynamicity. It needs to be monitored permanently to make sure its properties remain within certain limits in order to provide Quality-of-Service to the end users or to identify possible faults. To establish in every moment what is the appropriate reporting interval of the measured information and the way it is disseminated are important tasks. It has to achieve information quickly enough to solve any issue but excessive as to affect the data traffic. The problem that arises is that the monitoring information needs to travel in the network along with the user traffic and thus, potentially causing congestion. Considering that a wireless mesh network has highly dynamic characteristics there is a need for a good understanding of the influences of disseminating monitoring information in the network along with user traffic. In this paper we provide an evaluation of the network performance while monitoring information is collected from network nodes. We study how different monitoring packet sizes and different reporting frequency of the information can impact the user traffic and compare these values to the case in which only user data travels across the network.

Understanding system utilization is currently a difficult challenge for industry. Current monitoring tools tend to focus on monitoring critical servers and databases within a narrow technical context, and have not been designed to to manage extremely heterogeneous IT infrastructure such as desktops, laptops, and servers, where the number of devices can be in the order of tens of thousands. This is an issue for many different domains (organizations with large IT infrastructures, cloud computing providers, or software as a service providers) where an understanding of how computer hardware is being utilized is essential for understanding business cost, workload migrations and future investment requirements. Furthermore, organizations find it difficult to understand the raw metrics collected by current monitoring tools, in particular when trying to understand to what degree their systems are being utilized in the context of different business purposes. This paper presents different techniques for the extraction of meaningful resource utilization information from raw monitoring data, a utilization scoring algorithm, and then subsequently outlines a profile-based method for tracking the utilization of IT assets (systems) in large heterogeneous IT environments. We intend to determine how efficiently system resources are utilized considering their business use. We will provide to the end-user an assessment of the system utilization together with additional information to perform remedial action.

Composing large enterprise applications from reusable software components has become a major software development technique. Components are considered as black boxes and are composed together by their provided and required services. In comparison to other software development approaches, this allows for rapid development, clean and explicit architectures, and easy component reuse. Currently, there are several systems that support development and composition of components: service component architecture and Fractal are two well-known ones. At first sight, both have the same goals and provide very similar means to developers. In this paper, we briefly present both systems and then we focus on an in-depth comparison and evaluation of them.

The emerging Media Independent Handover (MIH) standard proposes to support session continuity during handover between heterogeneous networks. One of the critical features provided by MIH is an Event Service which includes predictive network degradation events, such as Link_Going_Down (LGD), which are triggered based on link layer metrics. Our results highlight the reactivness of media stream quality to network degradation. The point of degradation however, is specific to the characteristics of the class of media streaming service. Many existing event algorithms utilize static performance thresholds which are unresponsive to the requirements of individual application service classes. In this paper we propose a Service Oriented Link Triggering Algorithm (SOLTA) which triggers the LGD and Link_Down (LD) events based on link layer metrics but subject to the performance characteristics of the supported class of service. SOLTA illustrates that for 802.11, it is necessary to have aggressive service class specific, link event triggering. SOLTA also illustrates how a soft path handover approach such as Stream Control Transmission Protocols Concurrent Multi-path Transfer (SCTP-CMT) variant is necessary to support seamless session migration.

In the Future Internet, free exchange of information between enterprise applications and networking systems promotes the personalization of services and enables many different types of end-user applications and management operations optimizing the network performance. As result of this free information exchange, we need to facilitate the federation of information between these applications, harmonizing the differences between operation, management data and information models in heterogeneous networks, and application management systems. This paper concentrates on identifying the research challenges that we have to address in the areas of software engineering and network management supporting federation. In particular, we describe research challenges for federated management systems highlighting the flexibility these systems can offer in the Future Internet communications, when they are able to support value-added federated end-to-end services. As part of our approach, semantic techniques are cited to represent networking information governing technology and/or network protocols offering a wide diversity of end-to-end services as a result of this transparent information sharing process.

In this paper, the system capacity of a multi-cell IEEE 802.16j system operating in transparent mode is investigated. A previous published analytical model is used and incorporates interference from neighbouring cells. The model can be used to determine downlink performance under max-min fairness constraints for both sectored and omnidirectional systems with different amounts of relays. The study provides guidelines on the design of 802.16j systems operating in transparent mode specifically, how to maximise throughput for such systems. It is shown that inter-cell interference from relays is limited in a suburban environment and that relays can deliver significant gains over single hop systems: gains of 30% and 60% can be achieved without and with spatial reuse in an omnidirectional system while 15% and 25% in a sectorised system.

In this paper, the system capacity of IEEE 802.16j systems operating in transparent mode is investigated under varying numbers of relays and associated transmit power. The study is based on an extended variant of an analytical model defined in previous work and used to determine the throughput gain that can be achieved under a max-min fairness constraint. The study finds that significant overall throughput gains can be achieved over 802.16e systems: 125% and 55% gain with and without spatial reuse, respectively. However, the overall throughput gain drops as the transmit power at the relays (RSs) decreases despite the fact that a greater number of concurrent transmissions can be achieved due to less interference among the RSs' transmissions. Decreasing the transmit power at the RSs eventually leads to a significant reduction of the achievable throughput gain even though a large number of RSs are deployed in scenarios with or without spatial reuse.

Ethernet passive optical networks (EPONs) have attracted considerable attention from the industry, as they offer a simple, highly flexible and cost-effective solution to the problem of providing broadband access to a customer. The authors present a new approach to bandwidth allocation in EPONs where the optical line terminator (OLT) has full control over the access mechanism. This results in much simpler optical network unit (ONU) architecture. It is shown that such an OLT-centric architecture offers full support for differentiated services and makes enforcement of service level agreements possible. Extensive simulation experiments show that bandwidth allocation algorithms deployed in such a centralised environment can deliver good performance in terms of average and maximum packet delay. The authors introduce two new algorithms that target SLA-aware EPONs and provide a good protection of offered quality of service against interference from other sources.

Existing terminal oriented handover mechanisms capable of meeting the strict delay bounds of real time applications such as VoIP do not consider the QoS of candidate handover networks. In this paper ECHO - a QoS based handover solution for VoIP- is proposed. ECHO is endpoint centric and does not require any network support; it leverages the SCTP transport protocol. ECHO incorporates network metrics that directly affect VoIP quality into the handover decision process. A dynamic variant of the ITU-T E-Model is used to calculate how the network metrics map to a user perceived voice quality metric known as the MOS. The MOS value is then used to make handover decisions between each of the available access networks. The results show that the addition of the QoS capabilities significantly improves the handover decisions that are made.

The performance of multi-homed transport protocols tolerant of network failure is studied. It evaluates the performance of different retransmission policies combined with path failure detection thresholds, infinite or finite receive buffers for various path bandwidths, delays and loss rate conditions through stream control transmission protocol simulation. The results show that retransmission policies perform differently with different path failure detection threshold configurations. It identifies that retransmission of all data on an alternate path with the path failure detection threshold set to zero performs the best in symmetric path conditions but its performance degrades acutely in asymmetric path conditions even when the alternate path delay is shorter than the primary path delay. It illustrates that retransmission of all data on the same path with the path failure detection threshold set to one or zero gives the most stable performance in all path configurations.

Monitoring, analysing and understanding component based enterprise software systems are challenging tasks. These tasks are essential in solving and preventing performance and quality problems. Obtaining component level interactions which show the relationships between different software entities is a necessary prerequisite for such efforts. This paper focuses on component based Java applications, currently widely used by industry. They pose specific challenges while raising interesting opportunities for component level interaction extraction tools. We present a range of representative approaches for dynamically obtaining and using component interactions. For each approach we detail the needs it addresses, and the technical requirements for building an implementation of the approach. We also take a critical look at the different available implementations of the various techniques presented. We give performance and functional considerations and contrast them against each other by outlining their relative advantages and disadvantages. Based on this data, developers and system integrators can better understand the current state-of-the-art and the implications of choosing or implementing different dynamic interaction extraction techniques.

Future Quality of Service (QoS) aware networks, such as those based on the Differentiated Services (DiffServ) architecture, will provide users with the opportunity to assign a priority to the traffic they receive and generate. Different priorities will entail different per-byte or per-packet charges. To aid users operating within such an environment, a priority selection strategy is required. In this paper, we propose a priority selection strategy based on the concept of Consumer Surplus (CS), which is the positive difference between utility or the user’s willingness-to-pay and cost. We then evaluate this strategy in a networking scenario where a user is involved in a multiplayer Counterstrike game, in which different priorities may be assigned to the downstream traffic. A utility curve, applicable to a First Person Shooter (FPS) game like Counterstrike, is presented. Using simulation results we show that for varied traffic loads, the CS strategy offers comparable application performance to that of always selecting the highest priority, but at a reduced cost to the user.

In future networks based on the differentiated services (DiffServ) architecture, users may be provided with the opportunity to choose their assured forwarding (AF) class. Users will choose the most inexpensive AF class that satisfies their QoS requirements. Previously the authors have proposed an algorithm known as consumer surplus maximization (CSM) [1]. Although this algorithm can be beneficial when employed by a single user, widespread adoption of the algorithm could cause network instability. This can be avoided by replacing static charges, with a lightweight congestion-only charging (CoC) pricing mechanism. Simulation results are presented from a networking scenario where users connect across a single DiffServ domain to engage in a multiplayer first person shooter (FPS) game.

First person shooter (FPS) games are a popular online gaming genre played predominately over wired networks. FPS games are highly interactive and have stringent delay requirements. However, the explosive growth in wireless LAN (WLAN) deployment has seen an increase in the use of such networks for gaming purposes. The varied performance of the IEEE 802.11 MAC Distributed Coordination Function (DCF) has tended to make WLANs unsuitable for FPS games. The IEEE 802.11e standard introduces quality of service (QoS) mechanisms including enhanced distributed channel access (EDCA) allowing prioritization of competing flows. Using the NS-2 simulator, we evaluate the capability of 802.11g and 802.11e WLANs to support Quake IV games traffic in the presence of web traffic. We compare the results achieved using EDCA with those achieved by a non-elevated differentiated services scheduler known as Best Effort with loss trade-off (BELT) that differentiates at the IP layer. We find that the BELT scheduler compares favorably with EDCA in this context.

First Person Shooter (FPS) games are a popular online gaming genre played predominately over wired networks. FPS games have stringent delay requirements. However, the explosive growth in wireless LAN (WLAN) deployment has seen an increase in the use of such networks for gaming purposes. The varied performance of the IEEE 802.11 standard has tended to make WLANs unsuitable for FPS games. The IEEE 802.11e enhancement introduces Quality of Service (QoS) mechanisms including Enhanced Distributed Channel Access (EDCA) allowing prioritization of competing flows. Using the NS-2 simulator, we evaluate the capability of 802.11 WLANs to support Quake IV games traffic in the presence of web traffic. We compare the results achieved using EDCA with those achieved by a non-elevated differentiated services scheduler known as Best Effort with Loss Trade-off (BELT). We find that the BELT scheduler compares favorably with EDCA in this context.

In networks based on the Differentiated Services (DiffServ) architecture, users may be presented the opportunity to choose their Assured Forwarding (AF) class. Users will attempt to maximize their Consumer Surplus (CS) in such an environment. They can increase their CS by choosing the lowest cost AF class that satisfies their Quality of Service (QoS) requirements. Previously the authors have proposed an algorithm known as Consumer Surplus Maximization (CSM) [2]. Although this algorithm improves CS for users, widespread adoption of the algorithm can provoke network instability. If many users switch classes at once, degradation on one AF class can easily propagate. The likelihood of such an occurrence can be reduced by replacing fixed per-byte or per-packet charges associated with individual AF classes with a lightweight Congestion-only Charging (CoC) pricing mechanism. An additional charge is only applied to packets that receive discernible benefit from belonging to a particular class. We present simulation results from a networking scenario where users connect across a single DiffServ domain to engage in a multiplayer First Person Shooter (FPS) game.

In this paper a transport layer handover mechanism for Voice over Internet Protocol (VoIP) using the Stream Control Transmission Protocol (SCTP) is proposed. This handover mechanism utilizes the multi-homing feature of SCTP to allow connection to multiple wireless networks. Each available wireless network is probed with simulated VoIP data to obtain network performance metrics. Handover decisions are made based on the Mean Opinion Score (MOS) calculated from the ITU-T E-Model for voice quality assessment using the obtained measurements. Implementation and simulations were carried out using the NS2 network simulator. The results show the MOS obtained using the proposed handover mechanism on IEEE 802.11b WLAN access points (APs) under varying load conditions. High correlation is shown between the MOS from the proposed handover scheme and the MOS experienced by a VoIP node present in the network. A simulation showing an SCTP endpoint handover between AP's is presented.

In this paper a transport layer handover mechanism for Voice over Internet Protocol (VoIP) in 802.11b using the Stream Control Transmission Protocol (SCTP) is proposed. The multi-homing feature of SCTP is used to allow connections to several 802.11b Access Points (AP's). Probing packets that model VoIP encoded data at various transmission rates are used to obtain quality metrics from each of the available networks. Handover decisions are made based on the Mean Opinion Score (MOS) calculated from the ITU-T E-Model for voice quality assessment using the obtained measurements. The handover mechanism is shown to operate in 802.11b networks with heterogeneous transmission rates using multiple VoIP codecs. The results show a high correlation between the MOS predicted by the proposed mechanism and the MOS experienced by a VoIP call present in the network. Results verify the accurate operation of the scheme using multiple VoIP codecs at various transmission rates. A simulation showing an SCTP endpoint handover between heterogeneous transmission rate AP's is presented.

Development in wireless access technologies and multihomed personal user devices is driving the way towards a heterogeneous wireless access network environment. Success in this arena will be reliant on the ability to offer an enhanced user experience. Users will plan to take advantage of the competition and always connect to the network which can best service their preferences for the current application. They will rely on intelligent network selection decision strategies to aid them in their choice. The contribution of this paper is to propose an intelligent utility-based strategy for network selection in this multi-access network scenario. A number of utility functions are examined which explore different user attitudes to risk for money and delay preferences related to their current application. For example we show that risk takers who are willing to pay more money get a better service.

Web service compositions are often used to realise service-based enterpriseapplications. These enterprise systems are built from many existing discreteapplications, often legacy applications exposed using Web service interfaces.Acceptance of these systems is often constrained by non-functional aspects,such as Quality of Service (QoS). A number of factors affect the QoS of anenterprise system, including availability, scalability and performance. Thereare a number of architectural configurations or distribution patterns, whichexpress how a composed system is to be deployed. These distribution patternshave a direct impact upon the QoS of the composition. However, the amount ofcode required to realise these distribution patterns is considerable. Additionally, there is an increased deployment time associated with setting updifferent distribution patterns. We therefore propose a novel approach whichcombines a Model Driven Architecture using UML 2.0 for modeling andsubsequently generating Web service compositions, with a method for achieving dynamic decentralised interaction amongst services with reduced deployment overheads. These approaches combined provide for the generation of dynamic Web service compositions driven by a distribution pattern model.

Abstract — This paper presents how SCTP can stall in multihomed scenarios during failover and under certain circumstances. A stall is where an SCTP end-point ceases to communicate for an extended period of time, but does not report any error to the upper layer. This paper presents two different sets of circumstances where a stall can occur: firstly when there is an underestimation of the Retransmission Time-Out (RTO) value for a redundant network path; and secondly when a network error occurs that causes only SACKs to be lost, which confuses the SCTP sender about which network path is operational. Solutions to both of these stalls are presented that include modifying the RTO value, applying Karn’s algorithm to path monitoring and ensuring the destination address selection policy for SACKs is changed. This paper also presents a mechanism to de-couple data acknowledgement and path monitoring when using multi-homed transport protocols, which should remove the ambiguity about path monitoring and offers a universal solution to the stall. I.

This paper describes and formalises the service oriented heterogeneous wireless network environment (SOHWNE), the future service provision and delivery environment that supports ubiquitous user access anywhere at any time from diverse devices to a broad range of services. These services can be offered by third parties and can be accessed via one of many available networks. This paper also proposes and describes a novel algorithm for intelligent cost-oriented and performance-aware selection between available networks. This user-centric strategy focuses on the maximisation of consumer surplus when selecting the best available connection for transferring non real-time data, with user specified time constraints, in a user-centric SOHWNE

Research to date on composite Web services (CWS) has focused on selection of servers and precedence between requests as opposed to the requirements of the clients. This paper proposes that CWSs should have greater emphasis on both the individual customer requirements and the resultant impact on the performance of the overall system. The system described here employs the use of Web service level agreements and hard-and-soft criteria to develop an efficient Web service manager.

Component technologies are increasingly being used for building enterprise systems, as they can address complex functionality and flexibility problems and reduce development and maintenance costs. Nonetheless, current component technologies provide little support for predicting and controlling the emerging performance of software systems that are assembled from distinct components.This paper presents a framework for automating the performance management of complex, component-based systems. The adopted approach is based on the alternate usage of multiple component variants with equivalent functional characteristics, each one optimized for a different running environment. A fully-automated framework prototype for J2EE is presented, along with results from managing a sample enterprise application on JBoss. A mechanism that uses monitoring data to learn and automatically improve the framework's management behaviour is proposed. The framework imposes no extra requirements on component providers, or on the component technologies.

In future wireless and mobile environments it is likely that users will have access to multiple networks at the same time. Therefore there is a need to have mechanisms in place to decide which network is the most suitable for each user at each moment in time for every application that the user requires. We propose a user-centric solution where users choose the radio access network which meets their data transfer terms best. Naturally, each user wants timely quality data delivery at a low cost. In the radio environment data rates can never be guaranteed due to the unreliable nature of the radio links. The user network selection algorithm needs to predict the data rate on offer in each of the available networks and make the decision based on those predictions. This paper allows the user to select the network that will maximise the consumer surplus for non real-time data, while taking into account the delays

A novel technique for estimating the average delay experienced by an IP packet in cellular data networks with an SR-ARQ loop is presented. This technique uses the following input data: a statistical description of the radio channel, ARQ loop design parameters and the size of a transported IP packet. An analytical model is derived to enable a closed form mathematical estimation of this delay. To validate this model, a computer based simulator was built and tests showed good agreement between the simulation results and the model. This new model is of particular interest in predicting the packet delay for conversational traffic such as that used for VoIP applications.

The report summarizes the results of the Workshop on Middleware Benchmarking held during OOPSLA 2003. The goal of the workshop was to help advance the current practice of gathering performance characteristics of middleware implementations through benchmarking. The participants of the workshop have focused on identifying requirements of and obstacles to middleware benchmarking and forming a position on the related issues. Selected requirements and obstacles are presented, together with guidelines to adhere to when benchmarking, open issues of current practice, and perspectives on further research. Copyright © 2005 John Wiley & Sons, Ltd.

Passive optical networks (PON) are thought to be the next step in the development of access networks and providing broadband access in the "last mile" area. Ethernet PON (EPON) are gaining the most attention from industry as they offer a highly flexible, cost effective solution. In this paper, we present a new approach to the problem of optimal bandwidth allocation in EPON. We show that by moving all the access control functionality to the optical line terminator, a flexible solution could be achieved which offers full support for service level agreements and removes the burden of managing the configuration of every optical network unit. We include results of simulations that show that such an approach can deliver good performance in terms of average and maximum packet delay. We show that a novel Grant multiplexing scheme can significantly reduce jitter and end-to-end delay experienced by the high priority classes of traffic. To prove these points we present results of detailed experiments that were run on a C++ event driven simulator, that we have developed.

A novel methodology is proposed for the analysis of the IP packet delay performance of SR-ARQ mechanisms in a generalized wireless system. A simulation model of the system including a novel channel model is described and results are obtained for a range of IP packet size. To demonstrated the efficacy of the methodology, the test scenario is tailored to the transmission of small packets containing real-time data carried over an EGPRS system. The results show that the use of a mean value for IP packet delay estimation is of limited use when small packets are considered.

Developers often lack the time or knowledge to profoundly understand the performance issues in largescale component-oriented enterprise applications. This situation is further complicated by the fact that such applications are often built using a mix of in-house and commercial-off-the-shelf (COTS) components. This paper presents a methodology for understanding and predicting the performance of component-oriented distributed systems both during development and after they have been built. The methodology is based on three conceptually separate parts: monitoring, modelling and performance prediction. Performance predictions are based on UML models created dynamically by monitoring-and-analysing a live or under-development system. The system is monitored using non-intrusive methods and run-time data is collected. In addition, static data is obtained by analysing the deployment configuration of the target application. UML models enhanced with performance indicators are created based on both static and dynamic data, showing performance hot spots. To facilitate the understanding of the system, the generated models are traversable both horizontally at the same abstraction level between transactions, and vertically between different layers of abstraction using the concepts defined by the Model Driven Architecture. The system performance is predicted and performance-related issues are identified in different scenarios by generating workloads and simulating the performance models. Work is under way to implement a framework for the presented methodology with the current focus on the Enterprise Java Beans technology.

We believe that the typical hospital computing environment is especially fruitful as a domain for investigating the problems of interoperability and cooperation. We state this belief as hospital computing environments consist of a heterogeneous collection of autonomous information systems. These systems range from centralised hospital-wide systems, such as patient administration systems, to departmental systems such as pharmacy stock-control, laboratory information systems, accident and emergency systems and so on. Many of these are legacy systems which have been operating for many years and are difficult to integrate and virtually impossible to rewrite. In this article we discuss and assess the work of the Jupiter Project and its successor LIOM (Legacy system Interoperability using Object-oriented Methods).