As a method of estimating the parameters of statistical models in statistics, maximum likelihood (ML) method has gained much attention and been popularly applied to a variety of fields. In the subjects of information, control, as well as system science and engineering, it is usually exploited to es timate the parameters of diverse linear dynamic systems. The tough task of estimating ML parameters of diverse linear dynamic systems can be attributed to a constrained optimization problem. In an effort to solve the constrained optimization problem effectively, we first use the ML method to address the statistical modelling of ML parameter estimation of general linear dynamic systems. Then we present a novel heuristic particle swarm search algorithm called sliding mode controlling particle swarm optimization (SMCPSO) algorithm. In SMCPSO, we introduce a sliding mode controller (SMC) into standard particle swarm optimization (SPSO). The SMC is intentionally placed between the particle position and the global best particle position so as to improve the particles' position information. It ameliorates the exploration and exploitation search abilities of the particles. Thereafter, SMCPSO is exploited to solve the problem of the ML parameter estimation of a given linear dynamic system together with SPSO, recursive ML generalized least squares (RMLGLS) method, and recursive ML least squares (RMLLS) algorithm. The simulation results demonstrate that SMCPSO is an effective approach and superior to other three approaches in estimating the ML parameters of the linear dynamic system.

Complex software systems often suffer from flaws arising either during the design or the development stages. In many cases due to the lack of formal underpinning it is difficult to assuredly define the model from which the system is generated. In our work, we investigated the suitability of mathematical situation theory as a formal foundation for a computer modelling language. This paper describes our empirical evaluation of this thesis in the real-time domain using Infon Lab, a tool based upon situation theory.

In this paper, we introduce direct back propagation (BP) neural dynamic programming (NDP) into particle swarm optimisation (PSO). Thus, a direct BP NDP inspired PSO algorithm, which we call NDPSO, is proposed. In NDPSO, since direct BP NDP belongs to the class of heuristic dynamic programming algorithms based on model-based adaptive critic designs and often serves as an online learning control paradigm, critic BP neural network is trained to optimise a total reward-to-go objective, namely to balance Bellman's equation, while action BP neural network is used to train the inertia weight, cognitive, and social coefficients so that the critic BP network output can approach an ultimate reward-to-go objective of success. With the collective aid of action-critic BP neural networks, inertia weight, cognitive, and social coefficients become more adaptive. Besides, the NDPSO's mutation mechanism also has greatly improved the dynamic performance of the standard PSO. Empirical experiments are conducted on both unimodal and multimodal benchmark functions. The experimental results demonstrate NDPSO's effectiveness and superiority to many other PSO variants on solving most multimodal problems.

In this paper, we introduce the Moana network infrastructure. It draws on well-adopted practices from the database and software engineering communities to provide a robust and expressive information-sharing service using hypergraph-based network indirection. Our proposal is twofold. First, we argue for the need for additional layers of indirection used in modern information systems to bring the network layer abstraction closer to the developer's world, allowing for expressiveness and flexibility in the creation of future services. Second, we present a modular and extensible design of the network fabric to support incremental architectural evolution and innovation, as well as its initial evaluation.

In recent years, High Performance Computing (HPC) systems have been shifting from expensive massively parallel architectures to clusters of commodity PCs to take advantage of cost and performance benefits. Fault tolerance in such systems is a growing concern for long-running applications. In this paper, we briefly review the failure rates of HPC systems and also survey the fault tolerance approaches for HPC systems and issues with these approaches. Rollback-recovery techniques which are most often used for long-running applications on HPC clusters are discussed because they are widely used for long-running applications on HPC systems. Specifically, the feature requirements of rollback-recovery are discussed and a taxonomy is developed for over twenty popular checkpoint/restart solutions. The intent of this paper is to aid researchers in the domain as well as to facilitate development of new checkpointing solutions.

Cloud computing offers new capacity and flexibility solution to high performance computing (HPC) applications with provisioning of a large number of virtual machines for computational intensive applications. Fault tolerance allows HPC systems on cloud with multiple of nodes to complete execution of computational intensive applications in the present of fault. The most commonly used fault tolerance techniques for HPC is checkpoint/restart. However, checkpoint/restart increases the wall clock time of the execution of applications which increases the execution cost. In this paper we present a fault tolerance framework for high performance computing in Cloud. This framework proposes using process level redundancy (PLR) techniques to reduce the wall clock time of the execution of computational intensive applications.

The Internet has been encouraging and enabling business collaborations via online transactions over the Web. However, managing transactions in a long-running business process across domains still remains a challenge. In this paper, we propose a novel financial-compensation-based transaction management model, fcBTxM, to address this challenge. Unlike classical transaction management, fcBTxM does not attempt to recover data consistency via rollback when a failure occurs. Instead, our model always tries to forward-roll a business process via financial compensation. We use a state machine to capture and describe the states of our transaction model and their relationships. We also develop a set of technologies and protocols for enabling the new transaction management. A real business collaboration example is used to demonstrate the concept, and preliminary testing results are provided to evaluate our technologies.

Particle swarm optimization (PSO) is an optimization approach and has been widely used for a verity of optimization problem in both research and industrial domains. Due to the potential of PSO, several variants of the original PSO algorithms have been developed to improve PSO's efficiency and robustness. This paper proposes another variant of particle swarm optimization algorithm, called N-PωSO. This N-PωSO algorithm is based on classical feedback control theory and topological neighborhood, which offers better search efficiency and convergence stability. As a result, our N-PuωSO method features faster searching from the proportional term without steady-state error. And empirical results show that our N-PωSO algorithm is able to achieve high performance for both unimodal and multimodal optimization problems.

This paper proposes an algorithm for automatic segmentation of video clips into speaker units, with the intention of using the latter as the index units for a video indexing and retrieval system. The algorithm works by pooling together global information of each speaker before detecting the true speaker change locations. The use of stacked scanning can offer better results. The experiment shows that the algorithm is able to boost the discriminating power of BIC resulting in 20% reduction in average detection offset and 5% reduction in average maximum detection offset.

In recent years, we witnessed a range of innovations in the 'service' related technologies, such as Software as a Service, Platform as a Service and Infrastructure as a Service. Along with the Service Oriented Architecture, companies can wrap their technological product as a service, to collaborate with others. Facing the ever-escalating global competition, such collaboration is crucial. The viability of this paradigm highly depends on the compliance and therefore the trustworthiness of all collaborators. However, it is challenging to achieve trustworthiness in such a dynamic cross-domain environment, as each participator may deceit for individual benefits. As a solution, we have proposed to enforce strong accountability to enhance the trustworthiness. With this accountability, incompliance can always be determined in a provable and undeniable way. In this paper, we extend our work by proposing a novel modeling of the collaborative business process. Based on this modeling, we thoroughly analyze the evidence and proving procedure needed for different types of compliance, and evaluate the extent to which those compliance can be indeed proved. We have implemented a demonstrative system to show its effectiveness in real practice.

In recent years, High Performance Computing (HPC) systems have been shifting from expensive massively parallel custom architectures to clusters of commodity personal computers to take advantage of cost and performance benefits. To avoid having to restart an application in case of sudden failure, checkpointing/restart fault tolerance mechanisms are commonly implemented. One drawback to checkpointing/restart is that it creates an overhead which increases the execution time of an application. We present a theoretical analysis of our technique. The results show that the PLR checkpointing/restart can significantly improve the overall reliability of an HPC system.

In this paper, a novel evolutionary strategy-based particle swarm optimisation (PSO) approach is presented. The evolutionary strategy is dependent on a BP neural proportional integral derivative (PID) controller, which is intentionally placed between the position and the global best position in a closed loop of their relationship. The BP neural PID controller is designed to aid the position to well track the global best position. Furthermore, it can be utilised to improve the evolutionary dynamics of the PSO algorithm. The experiments for performance evaluation are tried on both analytical benchmark functions and an automatic control system of pulp consistency in pulping and paper-making engineering. The experimental results illustrate that the proposed approach enhances the diversity of swarms, considerably improves the global convergence efficiency and outperforms the PSO algorithm. In addition, it is also easily used in real industrial practice and offers a novel and convenient solution to engineering optimisation design of industrial systems.

Web services have been widely accepted as a platform-independent services-oriented technology. On the other hand, ubiquitous technologies are getting popular in a variety of domain applications. In particular, hosting web services from mobile devices became a way of extending knowledge sharing for teaching and learning purposes. This paper presents our design and implementation of a mobile web services framework for syndromic surveillance diagnosis and learning. This framework can assist farmers and veterinary students to study surveillance and diagnosis of farm animal diseases in the field. In this paper, we also present a performance study of hosting web services on mobile devices by evaluating the mobile learning application.

The enormous amount of data generated in daily operations and the increasing demands for data accessibility across organizations are pushing individuals and organizations to outsource their data storage to cloud storage services. However, the security and the privacy of the outsourced data goes beyond the data owners' control. In this paper, we propose a service composition approach to preserve privacy for data stored in untrusted storage service. A virtual file system, called Trust Store, is prototyped to demonstrate this concept. It allows users utilize untrusted storage service provider with confidentiality and integrity of the data preserved. We deployed the prototype with Amazon S3 and evaluate its performance.

As hand-held devices become popular in society, the demand for mobility is extended to teaching and learning purposes. This paper presents the design and implementation of a mobile learning system for syndromic surveillance and diagnosis. This system can assist farmers and veterinary students to study surveillance and diagnosis of farm animal diseases in the field. In this paper, we present a mobile problem-based learning method used for designing the diagnosis learning system. We also describe our solution to the limitation of storage capacity of hand-held devices. We also present how the design is implemented as portable software that can be deployed onto a large range of mobile phones and other hand-held devices.

Reinforcement Learning (RL) based approaches are now widely applied to building autonomous management applications with self-optimizing features. In this paper, we engineer an autonomous QoS controller based on a combination of fuzzy logic and RL to guarantee differentiated average response times for distributed services. We argue that there are two advantages of the fuzzy/RL QoS controller: the ability to autonomously tune if-then rules, and efficient modeling of the system state through fuzzy logic. Experiments conducted on a prototype show that the online learning fuzzy/RL controller can optimize its fuzzy rules effectively and efficiently and achieve better performance than manually tuned fuzzy controllers and the Self-Tuning Fuzzy Controller (STFC) used in.

Summary form only given. Accountability in Service Oriented Architecture (SOA) is a capability of making business processes across all participants (services, applications and people) accountable in terms of both business logic and Quality of Services (QoS). While accountability is a critical mechanism to enhance trust between collaborative services, there is the lack of standard accountability support in the current SOA infrastructure. For example, it is difficult with the existing technologies/infrastructure to resolve a dispute between two (web) services if some interactions between the two services go wrong; there is also little existing accountability support for a service consumer to collect quantity evidences to complain a service provider, who fails to meet its Service Level Agreement (SLA). As the increasing real-world activities are performed through the Internet connected services, we envision that there will be growing requirements for making the behaviors of both service providers and consumers accountable. In the business world, one may be reluctant to transact directly with a stranger. But a mutually trusted middleman can be used to facilitate transactions and resolve possible disputes. In this tutorial, we will share our observations and research results on building accountability into SOA. First, we will review related work on accountability in traditional distributed systems, ranging from Internet protocols and network file systems to outsourced database management systems. We will examine what methods embodied in these work can fit service computing in Internet scale and what cannot. Then we will present our research work on middleman-based approach to delivering accountability as a service, including our recent research results. This tutorial will focus on the major technical challenges of enabling SOA accountable and our solutions to these challenges. Finally, we will demonstrate our solutions using a collaborative services scenario deployed in Amazon EC^2 cloud. The goal of this tutorial is to provide detailed understanding of accountability issues and related technologies in SOA with in-depth related work discussions, recent research outcomes and a deployed accountability service prototype.

Abstract The accurate prediction of cost profiles of future job tasks on the nodes of clusters or grids helps users to select the cost performance profile that best suits their budgets. The current software designed to deploy job tasks uses forecasts of the target computers' future load profiles. Once the job task's future load profile is known, converting it to its cost profile needs further analysis as the cost depends on other factors such as its delays and variable background loads. This paper introduces a new charging mechanism for charging future job tasks, called the Cost Profiling Model. The model has been implemented in a Software Framework and evaluated for its prediction accuracy. Copyright © 2009 John Wiley & Sons, Ltd.

In this paper, we propose to combine linear prediction coefficients from the autoregressive model (AR) and the time series itself as features for the clustering algorithm. The purpose of the use of the AR model is to realize the importance of dynamic modeling of microarray time series data. We define the distance among the time series profiles using the autoregressive model and use the hierarchical clustering and the k-means clustering methods for comparison. The results show that the performance of the clustering DNA microarray time course profile is increased with the linear prediction coefficients in addition to the time series itself used as features.

The capsule-centric object-oriented (OO) UML-RT framework is being widely applied in modeling large-scale complex systems. Like general UML, it is important to have software patterns to cope with the complexity in the software structure and the constant changes in requirements. However, there are not yet adequate and explicit discussions on the application of software patterns in UML-RT, and it is difficult and unnecessary to develop UML-RT patterns from the scratch. In this paper, we present some suggestions to transform the widely used UML design patterns into UML-RT compliant patterns while keeping as much of the applicability, functionality and flexibility of the original patterns as possible.

While Web Services Security (WSS) enhances the security of web services, it may also introduce additional performance overheads to standard web services due to additional CPU processing and larger messages transferred. In this paper, we aim at clarifying this concern by conducting a quantitative performance evaluation of WSS overhead. Based on the evaluation, we extend our existing web services performance model by taking the extra WSS overheads into account. The extended performance model is validated on different environments with different messages sizes and WSS security policies.

UML-RT has been widely used in modeling complex embedded/real-time systems. Designers of such complex systems, however, also need a sound, systematic, and practical software process that can guide the whole development phase. In this paper, we describe a mapping between COMET and UML-RT that effectively maps COMET objects and models into UML-RT prototypes and software structures. Furthermore, based on the mapping relationship, we present a new software process that derives from COMET to address the issues of designing complex systems specifically for UML-RT. Finally, a practical example that applies this process using the UML-RT compliant CASE tool, Rational Rose Real Time, is fully discussed in order to demonstrate the applicability of the mapping and the software process.

Web services security (WSS) has been approved as a standard by OASIS and widely adopted in the industry as a solution for enhancing the security of Web services. However, the performance of WSS remains a concern due to the additional security contents added to SOAP message and the extra service time for processing these security contents. This paper aims at clarifying this concern by conducting a performance evaluation of WSS. A simple Web service is designed and used for performance testing with a variety of WSS polices and message sizes. The test results are categorized, compared and analyzed to work out the overheads for individual security setting. This work is expected to provide an overview and guidance for WSS performance overhead

In this paper, we propose spectral analysis method to identify periodically expressed genes in microarray data using a forward and backward linear prediction (FBLP) model and the singular value decomposition (SVD) (FBLP-SVD) algorithm. The spectrum-mean-subtraction method is employed prior to this analysis as a pre-filtering procedure. The combination of the spectrum-mean-subtraction and FBLP-SVD algorithm offers a effective tool for periodicity identification. Using our technique, more genes have been successful identified as periodic genes in the genome of Saccharomyces cerevisiae.