Traffic models are at the heart of any performance evaluation of telecommunications networks. An accurate estimation of network performance is critical for the success of broadband networks. Such networks need to guarantee an acceptable quality of service (QoS) level to the users. Therefore, traffic models need to be accurate and able to capture the statistical characteristics of the actual traffic. In this article we survey and examine traffic models that are currently used in the literature. Traditional short-range and non-traditional long-range dependent traffic models are presented. Number of parameters needed, parameter estimation, analytical tractability, and ability of traffic models to capture marginal distribution and auto-correlation structure of actual traffic are discussed. n Figure 1. Finite state model for voice. This research was supported in part by the National Science Foundation under grant NCR-9396299. This article is based on Georgia Tech technical report G...

This paper describes the basis for techniques such as stochastic subdivision in the theory of random processes and estimation theory. The popular stochastic subdivision construction is then generalized to provide control of the autocorrelation and spectral properties of the synthesized random functions. The generalized construction is suitable for generating a variety of perceptually distinct high-quality random functions, including those with non-fractal spectra and directional or oscillatory characteristics. It is argued that a spectral modeling approach provides a more powerful and somewhat more intuitive perceptual characterization of random processes than does the fractal model. Synthetic textures and terrains are presented as a means of visually evaluating the generalized subdivision technique. Categories and Subject Descriptors: I.3.3 [Computer Graphics]: Picture/Image Generation; I.3.7 [Computer Graphics]: Three Dimensional Graphics and Realism -<F11.

We examine whether the U.S. rate of price inflation has become harder to forecast and, to the extent that it has, what changes in the inflation process have made it so. The main finding is that the univariate inflation process is well described by an unobserved component trend-cycle model with stochastic volatility or, equivalently, an integrated moving average process with time-varying parameters. This model explains a variety of recent univariate inflation forecasting puzzles and begins to explain some multivariate inflation forecasting puzzles as well. Key words: Phillips curve, trend-cycle model, moving average, great moderation JEL codes: C53, E37 *We thank Jonas Fisher for bringing several of the issues discussed in this paper to our attention in a 1999 conversation, Luca Benati for (more recent) helpful suggestions, and Matthew Shapiro, Robert Gordon, and two anonymous referees for helpful comments on an earlier draft. Replications files for the results in this paper can be downloaded from

In order to evaluate di#erent scheduling strategies for parallel computers, simulations are often executed. As the scheduling quality highly depends on the workload that is served on the parallel machine, a representative workload model is required. Common approaches such as using a probability distribution model can capture the static feature of real workloads, but they do not consider the temporal relation in the traces. In this paper, a workload model is presented which uses Markov chains for modeling job parameters. In order to consider the interdependence of individual parameters without requiring large scale Markov chains, a novel method for transforming the states in di#erent Markov chains is presented. The results show that the model yields closer results to the real workloads than other common approaches.

Being able to deal with time-warped sequences is crucial for a large number of tasks autonomous agents can be faced with in real-world environments, where robustness concerning natural temporal variability is required, and similar sequences of events should automatically be treated in a similar way. Such tasks can easily be dealt with by natural animals, but equipping an animat with this capability is rather difficult. The presented experiments show how this problem can be solved with a neural network by ensuring slow state changes. An animat equipped with such a network not only adapts to the environment by learning from a number of examples, but also generalizes to yet unseen time-warped sequences. 1 Introduction For numerous tasks, autonomous agents have to be able to deal with time-warped sequences of events. Sequential patterns of variable length are common in realworld environments, and the number of available training examples are usually relatively small. Animats should not on...

The Kalman filter algorithm can be used for many data assimilation problems. For large systems, that arise from discretizing partial differential equations, the standard algorithm has huge computational and storage requirements. This makes direct use infeasible for many applications. In addition numerical difficulties may arise if due to finite precision computations or approximations of the error covariance the requirement that the error covariance should be positive semi-definite is violated. In this paper an approximation to the Kalman filter algorithm is suggested that solves these problems for many applications. The algorithm is based on a reduced rank approximation of the error covariance using a square root factorization. The use of the factorization ensures that the error covariance matrix remains positive semi-definite at all times, while the smaller rank reduces the number of computations and storage requirements. The number of computations and storage required depend on the ...

This paper presents PRESTO, a novel two-tier sensor data management architecture comprising proxies and sensors that cooperate with one another for acquiring data and processing queries. PRESTO proxies construct time-series models of observed trends in the sensor data and transmit the parameters of the model to sensors. Sensors check sensed data with model-predicted values and transmit only deviations from the predictions back to the proxy. Such a model-driven push approach is energyefficient, while ensuring that anomalous data trends are never missed. In addition to supporting queries on current data, PRESTO also supports queries on historical data using interpolation and local archival at sensors. PRESTO can adapt model and system parameters to data and query dynamics to further extract energy savings. We have implemented PRESTO on a sensor testbed comprising Intel Stargates and Telos Motes. Our experiments show that in a temperature monitoring application, PRESTO yields one to two orders of magnitude reduction in energy requirements over on-demand, proactive or model-driven pull approaches. PRESTO also results in an order of magnitude reduction in query latency in a 1 % duty-cycled five hop sensor network over a system that forwards all queries to remote sensor nodes.

An enormous amount of analytical literature has recently appeared on the topic of “unit roots ” in macroeconomic time series. Indeed, tests for the presence of unit roots and techniques for dealing with them have together comprised one of the most active areas, over the past decade, in the entire field of macroeconomics. The issues at hand have involved substantive questions about the nature of macroeconomic growth and fluctuations in developed economies and technical questions about model formulation and estimation in systems that include unit-root variables. The present paper attempts to describe several of the main issues and to evaluate alternative positions. It does not pretend to be a comprehensive survey of the literature or to provide an “even-handed ” treatment of issues, however. 1 Instead, it attempts to develop a convincing perspective on the topic, one that is consistent with the views of many active researchers in the area but that may nevertheless be somewhat idiosyncratic. The exposition that is presented below is designed to be predominantly nontechnical in nature. Indeed, it takes a rather old-fashioned approach to

Applying techniques from Machine Learning to real-world domains and problems often requires considerable processing of the input data, to both remove noise and to augment the amount and type of information present. We describe our work in the task of situation assessment in the domain of US Army training exercises involving hundreds of agents interacting in real-time over the course of several days. In particular, we describe techniques we have developed to process this data and draw general conclusions on the types of information required in order to apply various Machine Learning algorithms and how this information may be extracted in real-world situations where it is not directly represented.

This paper proposes an approach for solving multivariate control problems with a fuzzy controller. Instead of using selected input variables, statistical indices are extracted to feed the fuzzy controller. The original input space is transformed into an eigenspace. If a sequence of training data are sampled in a local context, a small number of eigenvectors which possess larger eigenvalues provide a good summary of all the original variables. Fuzzy controllers can be trained for mapping the input projection in the eigenspace to the outputs. Implementations with the prediction of time series and vision-based robot control validate the concept. 1. Multivariate Problems in Modelling and Control It is well-known that general fuzzy rule descriptions of systems with a large number of input variables suffer from the problem of the "curse of dimensionality". In many real applications, it is difficult to identify the correct influential factors and reduce their number to the minimum. A genera...