Using observation ageing to improve Markovian model learning in QoS engineering

Markovian models are widely used to analyse quality-of-service properties of both system designs and deployed systems. Thanks to the emergence of probabilistic model checkers, this analysis can be performed with high accuracy. However, its usefulness is heavily dependent on how well the model captures the actual behaviour of the analysed system. Our work addresses this problem for a class of Markovian models termed discrete-time Markov chains (DTMCs). We propose a new Bayesian technique for learning the state transition probabilities of DTMCs based on observations of the modelled system. Unlike existing approaches, our technique weighs observations based on their age, to account for the fact that older observations are less relevant than more recent ones. A case study from the area of bioinformatics workflows demonstrates the effectiveness of the technique in scenarios where the model parameters change over time.

Description and training of neural network dynamics

Attractor properties of a popular discrete-time neural network model are illustrated through numerical simulations. The most complex dynamics is found to occur within particular ranges of parameters controlling the symmetry and magnitude of the weight matrix. A small network model is observed to produce fixed points, limit cycles, mode-locking, the Ruelle-Takens route to chaos, and the period-doubling route to chaos. Training algorithms for tuning this dynamical behaviour are discussed. Training can be an easy or difficult task, depending whether the problem requires the use of temporal information distributed over long time intervals. Such problems require training algorithms which can handle hidden nodes. The most prominent of these algorithms, back propagation through time, solves the temporal credit assignment problem in a way which can work only if the relevant information is distributed locally in time. The Moving Targets algorithm works for the more general case, but is computationally intensive, and prone to local minima.

The 'moving targets' training algorithm

A simple method for training the dynamical behavior of a neural network is derived. It is applicable to any training problem in discrete-time networks with arbitrary feedback. The algorithm resembles back-propagation in that an error function is minimized using a gradient-based method, but the optimization is carried out in the hidden part of state space either instead of, or in addition to weight space. Computational results are presented for some simple dynamical training problems, one of which requires response to a signal 100 time steps in the past.

The 'moving targets' training algorithm

A simple method for training the dynamical behavior of a neural network is derived. It is applicable to any training problem in discrete-time networks with arbitrary feedback. The method resembles back-propagation in that it is a least-squares, gradient-based optimization method, but the optimization is carried out in the hidden part of state space instead of weight space. A straightforward adaptation of this method to feedforward networks offers an alternative to training by conventional back-propagation. Computational results are presented for simple dynamical training problems, with varied success. The failures appear to arise when the method converges to a chaotic attractor. A patch-up for this problem is proposed. The patch-up involves a technique for implementing inequality constraints which may be of interest in its own right.

Investigation of a cross-layer link adaptation algorithm for IEEE 802.11n networks

Link adaptation is a critical component of IEEE 802.11 systems, which adapts transmission rates to dynamic wireless channel conditions. In this paper we investigate a general cross-layer link adaptation algorithm which jointly considers the physical layer link quality and random channel access at the MAC layer. An analytic model is proposed for the link adaptation algorithm. The underlying wireless channel is modeled with a multiple state discrete time Markov chain. Compared with the pure link quality based link adaptation algorithm, the proposed cross-layer algorithm can achieve considerable performance gains of up to 20%.

Variational mean-field algorithm for efficient inference in large systems of stochastic differential equations

This work introduces a Gaussian variational mean-field approximation for inference in dynamical systems which can be modeled by ordinary stochastic differential equations. This new approach allows one to express the variational free energy as a functional of the marginal moments of the approximating Gaussian process. A restriction of the moment equations to piecewise polynomial functions, over time, dramatically reduces the complexity of approximate inference for stochastic differential equation models and makes it comparable to that of discrete time hidden Markov models. The algorithm is demonstrated on state and parameter estimation for nonlinear problems with up to 1000 dimensional state vectors and compares the results empirically with various well-known inference methodologies.

Nonlinear pulse combining and pulse compression in multi-core fibers

We demonstrate light pulse combining and pulse compression using a continuous-discrete nonlinear system implemented in a multi-core fiber (MCF). It is shown that the pulses initially injected into all of the cores of a ring MCF are combined by nonlinearity into a small number of cores with simultaneous pulse compression. We demonstrate the combining of 77% of the energy into one core with pulse compression over 14× in a 20-core MCF. We also demonstrate that a suggested scheme is insensitive to the phase perturbations. Nonlinear spatio-temporal pulse manipulation in multi-core fibers can be exploited for various applications, including pulse compression, switching, and combining.

Nonlinear combining and compression in multicore fibers

We demonstrate numerically light-pulse combining and pulse compression using wave-collapse (self-focusing) energy-localization dynamics in a continuous-discrete nonlinear system, as implemented in a multicore fiber (MCF) using one-dimensional (1D) and 2D core distribution designs. Large-scale numerical simulations were performed to determine the conditions of the most efficient coherent combining and compression of pulses injected into the considered MCFs. We demonstrate the possibility of combining in a single core 90% of the total energy of pulses initially injected into all cores of a 7-core MCF with a hexagonal lattice. A pulse compression factor of about 720 can be obtained with a 19-core ring MCF.

Transition from time-continuous systems to discrete mappings

Detailed transport studies in plasmas require the solution of the time evolution of many different initial positions of test particles in the phase space of the systems to be investigated. To reduce this amount of numerical work, one would like to replace the integration of the time-continues system with a mapping.

Using system dynamics in a discrete-event simulation study of a manufacturing plant

Purpose - To provide an example of the use of system dynamics within the context of a discrete-event simulation study. Design/methodology/approach - A discrete-event simulation study of a production-planning facility in a gas cylinder-manufacturing plant is presented. The case study evidence incorporates questionnaire responses from sales managers involved in the order-scheduling process. Findings - As the project progressed it became clear that, although the discrete-event simulation would meet the objectives of the study in a technical sense, the organizational problem of "delivery performance" would not be solved by the discrete-event simulation study alone. The case shows how the qualitative outcomes of the discrete-event simulation study led to an analysis using the system dynamics technique. The system dynamics technique was able to model the decision-makers in the sales and production process and provide a deeper understanding of the performance of the system. Research limitations/implications - The case study describes a traditional discrete-event simulation study which incorporated an unplanned investigation using system dynamics. Further, case studies using a planned approach to showing consideration of organizational issues in discrete-event simulation studies are required. Then the role of both qualitative data in a discrete-event simulation study and the use of supplementary tools which incorporate organizational aspects may help generate a methodology for discrete-event simulation that incorporates human aspects and so improve its relevance for decision making. Practical implications - It is argued that system dynamics can provide a useful addition to the toolkit of the discrete-event simulation practitioner in helping them incorporate a human aspect in their analysis. Originality/value - Helps decision makers gain a broader perspective on the tools available to them by showing the use of system dynamics to supplement the use of discrete-event simulation. © Emerald Group Publishing Limited.

A simulation of a workflow management system

Suggests that simulation of the workflow component of a computer supported co-operative work (CSCW) system has the potential to reduce the costs of system implementation, while at the same time improving the quality of the delivered system. Demonstrates the value of being able to assess the frequency and volume of workflow transactions using a case study of CSCW software developed for estate agency co-workers in which a model was produced based on a discrete-event simulation approach with implementation on a spreadsheet platform.

A flexible, extensible object oriented real-time near photorealistic visualization system:the system framework design

In this paper we describe a novel, extensible visualization system currently under development at Aston University. We introduce modern programming methods, such as the use of data driven programming, design patterns, and the careful definition of interfaces to allow easy extension using plug-ins, to 3D landscape visualization software. We combine this with modern developments in computer graphics, such as vertex and fragment shaders, to create an extremely flexible, extensible real-time near photorealistic visualization system. In this paper we show the design of the system and the main sub-components. We stress the role of modern programming practices and illustrate the benefits these bring to 3D visualization. © 2006 Springer-Verlag Berlin Heidelberg.