In order to to guarantee continuous delivery of a video stream in an on-demand video server environment, a collection of resources (referred a logical channel) are reserved in advance. To conserve server resources, multiple client requests for the same video can be batched together and served by a single channel. Increasing the batching interval results in larger savings in server capacity, however, it also increases the reneging probability of a client. A complication introduced by batching is that if a batched client pauses, a new stream (which may not be immediately available) needs to be started when the client resumes. To provide short response time to resume requests, some channels are set aside and are referred to as contingency channels. To further improve resource utilization, even when a non-batched client pauses, the channel is released and reacquired upon resume. In this paper, we first develop an analytical model that predicts the reneging probability and expected resume d...

Approximating the joint data distribution of a multi-dimensional data set through a compact and accurate histogram synopsis is a fundamental problem arising in numerous practical scenarios, including query optimization and approximate query answering. Existing solutions either rely on simplistic independence assumptions or try to directly approximate the full joint data distribution over the complete set of attributes. Unfortunately, both approaches are doomed to fail for high-dimensional data sets with complex correlation patterns between attributes. In this paper, we propose a novel approach to histogram-based synopses that employs the solid foundation of statistical interaction models to explicitly identify and exploit the statistical characteristics of the data. Abstractly, our key idea is to break the synopsis into (1) a statistical interaction model that accurately captures significant correlation and independence patterns in data, and (2) a collection of histograms on low-dimensional marginals that, based on the model, can provide accurate approximations of the overall joint data distribution. Extensive experimental results with several real-life data sets verify the effectiveness of our approach. An important aspect of our general, model-based methodology is that it can be used to enhance the performance of other synopsis techniques that are based on data-space partitioning (e.g., wavelets) by providing an effective tool to deal with the “dimensionality curse”. 1.

For a video-on-demand computer system we propose a scheme which balances the load on the disks, thereby helping to solve a performance problem crucial to achieving maximal video throughput. Our load balancing scheme consists of two stages. The static stage determines good assignments of videos to groups of striped disks. The dynamic phase uses these assignments, and features a DASD dancing algorithm which performs real-time disk scheduling in an effective manner. Our scheme works synergistically with disk striping. We examine the performance of the DASD dancing algorithm via simulation experiments.

Effective support for XML query languages is becoming increasingly important with the emergence of new applications that access large volumes of XML data. All existing proposals for querying XML (e.g., XQuery) rely on a pattern-specification language that allows path navigation and branching through the XML data graph in order to reach the desired data elements. Optimizing such queries depends crucially on the existence of concise synopsis structures that enable accurate compile-time selectivity estimates for complex path expressions over graph-structured XML data. In this paper, we introduce a novel approach to building and using statistical summaries of large XML data graphs for effective path-expression selectivity estimation. Our proposed graph-synopsis model (termed XSKETCH) exploits localized graph stability to accurately approximate (in limited space) the path and branching distribution in the data graph. To estimate the selectivities of complex path expressions over concise XSKETCH synopses, we develop an estimation framework that relies on appropriate statistical (uniformity and independence) assumptions to compensate for the lack of detailed distribution information. Given our estimation framework, we demonstrate that the problem of building an accuracy-optimal XSKETCH for a given amount of space is ### -hard, and propose an efficient heuristic algorithm based on greedy forward selection. Briefly, our algorithm constructs an XSKETCH synopsis by successive refinements of the label-split graph, the coarsest summary of the XML data graph. Our refinement operations act locally and attempt to capture important statistical correlations between data paths. Extensive experimental results with synthetic as well as real-life data sets verify the effectiveness of our app...

For a video-on-demand computer system we propose a scheme which balances the load on the disks, thereby helping to solve a performance problem crucial to achieving maximal video throughput. Our load balancing scheme consists of two components. The static component determines good assignments of videos to groups of striped disks. The dynamic component uses these assignments, and features a "DASD dancing" algorithm which performs realtime disk scheduling in an effective manner. Our scheme works synergistically with disk striping. We examine the performance of the proposed algorithm via simulation experiments. 1 Introduction Consider a video-on-demand (VOD) computer system consisting of a central processor and a collection of shared disks, sometimes known as direct access storage devices (DASDs). VOD computer systems must be able to "play" multiple streams of many different videos simultaneously, based on customer demand. Most videos will be stored most cost effectively on disk. (A few ...

This paper proposes a dynamic cache partitioning method for simultaneous multithreading systems. We present ageneral partitioning scheme that can be applied to setassociative caches at any partition granularity. Further-more, in our scheme threads can have overlapping partitions, which provides more degrees of freedom when par-titioning caches with low associativity. Since memory reference characteristics of threads canchange very quickly, our method collects the miss-rate characteristics of simultaneously executing threads at run-time, and partitions the cache among the executing threads. Partition sizes are varied dynamically to improve hit rates.Trace-driven simulation results show a relative improvement in the L2 hit-rate of up to 40.5 % over those gener-ated by the standard least recently used replacement policy, and IPC improvements of up to 17%. Our results show thatsmart cache management and scheduling is important for SMT systems to achieve high performance.

Our Continuous Adaptive Monitoring (CAM) system provides responses for continuous queries by monitoring and extracting information scattered across the web. Continuous queries are the queries for which responses given to users must be continuously updated, as the sources of interest get updated. Such queries occur, for instance, during on-line decision making, e.g., traffic flow control, weather monitoring etc. Whereas push-based techniques may be able to refresh query results meeting user requirements, they do not scale well. With the pull based approach, the problem of keeping the responses current reduces to the problem of deciding how often to visit a source to determine if and how it has been modified so that a user response can be updated accordingly. As should be evident, periodical monitoring is not scalable. Also, it can lead to huge wastage of monitoring resources. Hence CAM employs a multiphase approach. In the tracking phase, changes to an initially identified set of relevant pages, are tracked. From the observed change characteristics of these pages, a probabilistic model of their change behaviour is formulated and weights are assigned to pages to denote their importance for the current queries. Based on these statistics, during the next phase, the Resource Allocation phase, resources, needed to continuously monitor these pages for changes, are allocated. Given these resource allocations, the Scheduling phase produces an optimal achievable schedule of monitoring. An experimental evaluation of our approach compared to prior approaches on synthetic data for crawling dynamic web pages shows the effectiveness of our approach to monitoring dynamic changes. For example, by monitoring just 5 % of the possible change instances, CAM is able to return 90 % of the changed information to the users. In this demonstration, we show how CAM keeps users up-to-date with respect to a set of ongoing sports related events. 1.

We discuss new theoretical and experimental results on the Dynamic Rate Shaping (DRS) approach for transcoding compressed video bitstreams (MPEG-1, MPEG-2, MPEG-4, H.261 as well as JPEG). A set of low complexity algorithms for both constrained and unconstrained DRS are presented. We present the first extensive experimental study on the various DRS algorithms (causally optimal, memoryless, and rate-based) both in their constrained and generalized forms. The study proves the computational viability of the DRS approach to transcoding and identifies a range of rate shaping ratios for which it is better than requantization, both complexity-wise as well as in performance. We then substantiate the almost-optimal experimental performance of the memoryless algorithm by analyzing the behavior of the DRS problem assuming a first order Autoregressive source. By deriving the statistical and rate-distortion characteristics of different components of the inter-frame rate shaping problem, we offer an explanation as to why the set of optimal breakpoint values for any frame is somewhat invariant to the accumulated motion compensated shaping error from past frames. This result is significant as it opens up the way to construct much simpler memoryless algorithms that give minimal penalty in achieved quality, not just for this, but possibly other types of algorithms. Of equal, if not more, importance is the very first use of matrix perturbation theory for tracking the spectral behavior of the auto-correlation matrix of the source signal and the motion residual it yields. 1 1

The availability of large scale multitasked parallel architectures introduces the following processor assignment problem: we are given a long sequence of data sets, each of which is to undergo processing by a collection of tasks whose inter-task data dependencies form a series-parallel partial order. Each individual task is potentially parallelizable, with a known experimentally determined execution signature. Recognizing that data sets can be pipelined through the task structure, the problem is to find a "good" assignment of processors to tasks. Two objectives interest us: minimal response time per data set given a throughput requirement, and maximal throughput given a response time requirement. Our approach is to decompose a series-parallel task system into its essential "serial" and "parallel" components; our problem admits the independent solution and recomposition of each such component. We provide algorithms for the series analysis, and use an algorithm due to Krishnamurti and Ma...

The mutual information of independent parallel Gaussian-noise channels is maximized, under an average power constraint, by independent Gaussian inputs whose power is allocated according to the waterfilling policy. In practice, discrete signaling constellations with limited peak-to-average ratios (m-PSK, m-QAM, etc.) are used in lieu of the ideal Gaussian signals. This paper gives the power allocation policy that maximizes the mutual information over parallel channels with arbitrary input distributions. Such policy admits a graphical interpretation, referred to as mercury/waterfilling, which generalizes the waterfilling solution and allows retaining some of its intuition. The relationship between mutual information of Gaussian channels and nonlinear minimum mean-square error (MMSE) proves key to solving the power allocation problem.