CDC2000 Performance Analysis in Communication Networks

Performance Analysis in Communication Networks
Home Full List of Titles 1: Proceedings of CDC2000 Discrete Event Systems Control in Communication Systems Optimal Control and Applications I Optimisation Approaches and Methods Model Predictive Control Advances in Linear Estimation Stochastic and Uncertain Systems Nonlinear Control and Applications Nonlinear Estimation and Filtering Formation Control and its Applications New Approaches to Fuzzy Control Manufacturing Systems Automotive Applications Stability Issues in Hybrid Control Recent Advances in Stochastic Networks Optimal Control and Applications II Robust Controller Design - mu, L1 and H2 Constrained and Receding Horizon Control Identification and Control around the World Markov Decision Processes Nonlinear Optimisation Observers for Nonlinear Systems Motion Planning Neural / Fuzzy Stability and Control Motor Control Control of Quantum Phenomena I Hybrid Systems Methods Control in Communication Networks Robustness and Optimisation Bumpless Transfer, Antiwindup and Saturation Adaptive Control: Linear Systems Estimation and Closed Loop Identification Control of Markov Processes Nonlinear Filtering and Control Modelling, Identification and Validation of Nonlinear Systems Differential Geometric Control Theory for Mechanical Systems Nonlinear Output Feedback Control Pneumatics and Compression Systems Control of Quantum Phenomena II Stability of Hybrid Systems Performance Analysis in Communication Networks Adaptive Control of Nonlinear Systems LMI Methods in Design Robust Control of Time Delay Systems Subspace Identification Methods Nonlinear Stochastic Filtering and Estimation Bifurcations, Chaos and Control I New Progress in Synthesis of Nonlinear Systems I Implementation Issues of Sliding Mode Control Theory Control of Mixing in Shear Flows Novel Neural Network Control Techniques for Industrial Motion Control Systems Physiological Control Systems Optimal Control of Hybrid Systems Stochastic Models for Communication Networks Control and Stabilisation of Nonlinear Systems New Directions in Robust Control Linear Systems Theory Advanced Topics in Systems Theory Estimation in Action Bifurcations, Chaos and Control II New Progress in Synthesis of Nonlinear Systems II Numerical Design and Analysis Techniques for Nonlinear Systems Analysis and Control of Underactuated Systems Sliding Mode Control I Challenges in the Application of Control to Computer Systems Estimation and Diagnosis of Discrete Event Systems Communications and Games Optimal Control Stochastic Systems Model Reduction Methodologies Identification and Subspace Methods Applications of Nonlinear Adaptive Control Advances in Nonlinear Output Feedback Design The Behavioural Approach to Systems and Control Vision Based Estimation and Control: Recent Advances and Open Problems Agile Control of Military Operations Sliding Mode Control II Model-based Fault Diagnosis of Industrial Processes Discrete Event Systems / Petri Nets System Identification and Confidence Estimation New Approaches to H-Infinity Control I Probabilistic Approaches to Robust Control Time Delay System Stabilisation Identification Methods Controlled Stochastic Processes Output Feedback of Nonlinear Systems Topics in Nonlinear Stabilisation Mobile Robots: Tracking Control Robust Control of Nonlinear Systems Power Systems Stabilisation and Control Disk Drive Control Hybrid Control Applications Discrete Time Systems New Approaches to H-Infinity Control II Linear Systems with Saturating Actuators New Theories in Distributed Parameter Systems Applications of Estimation and Identification Stochastic Control and Tuning Methodologies Control of Nonlinear Systems Iterative Learning and Control Coordinating Robot Systems Nonlinear Time Varying Systems Novel Applications of Neural Networks Aerospace Applications Switched Systems Implicit and Descriptor Systems LQG Periodic Systems and Disturbances New Horizons for Distributed Parameter Systems State Estimation Learning and Neuro-Control Nonlinear Control and Stabilisation I Tracking Vision Servoing Controllability of Nonlinear Systems Control of Flexible Systems Electro-Mechanical Systems Robust Control Methods and Applications Fault Detection and Diagnosis Optimisation and Applications Robust Stability Analysis Numerical Methods in Control Filtering in Continuous Time Stochastic Systems Interplay between Control and Signal Processing Fault Detection and Analysis Nonlinear Dynamical Systems Nonlinear Time Delay Systems Computational Issues in Nonlinear Control Disturbance Rejection Process Control Industry Applications Linear Parameter Varying Systems Linear Control Systems Dynamic and Nonlinear Programming Model Reduction Applications New Techniques for Control and Systems: Numerical Linear Algebra Estimation and Identification using Hidden Markov Models Applications of Stochastic Control Topics in Linear Design Nonlinear Control and Stabilisation II Ambulatory Robot Systems Chaotic and Oscillatory Systems Biomedical System Control Integrated Control and CPU Scheduling Linear Design Techniques Adaptive Disturbance / Noise Compensation Nonlinear Model Predictive Control Sensitivity Design, Analysis and Limitations Analysis of Linear Systems Linear Matrix Inequalities in Design Lyapunov's 2nd Method Robotics: Tracking Control Lagrangian and Hamiltonian Theory Variable Structure Control Machine Vision Signal Processing Methods in Control Applied Nonlinear Control Author Index A B C D E F G H I J K L M N O P Q R S T U V W X Y Z	A Performance Comparison of Dynamic vs. Static Load Balancing Policies in a Mainframe -- Personal Computer Network Model Authors: Hisao Kameda, Said Fathy El-Zoghdy, Inhwan Ryu, Jie Li, Volume: 1, Page 1415 Paper number 1601 Abstract: Distributed computer systems can share job processing in the event of overloads. Load balancing involves the distribution of jobs throughout a networked computer system, thus increasing throughput without having to obtain additional or faster computer hardware. Load balancing policies may be either static or dynamic. Static load balancing policies are generally based on the information about the average behavior of system; transfer decisions are independent of the actual current system state. Dynamic policies, on the other hand, react to the actual current system state in making transfer decisions. This makes dynamic policies necessarily more complex than static ones, and truly optimal dynamic policies are known only for special systems. This study focuses on performance comparison between static and dynamic load balancing policies in a distributed computer system where truly optimal solutions of both dynamic and static policies have been characterized. The system consists of two types of service facilities, a Mainframe node and an unlimited number of Personal Computer nodes. The results suggest that, in the model examined, the dynamic policy outperforms the static one in the mean response time, at most about 30 percent and for the range of parameter values such that the arrival rate is near the processing rate of the Mainframe. CD001601.PDF (From Author) TOP Scheduling Time-Constrained Jobs in the Presence of Background Traffic Authors: Sandjai Bhulai, Ger Koole, Volume: 1, Page 1421 Paper number 1602 Abstract: In this paper we study the scheduling of jobs with a constraint on the average waiting time in the presence of background jobs. The objective is to schedule to s servers such that the throughput of the background traffic is maximized while satisfying the response time constraint on the foreground traffic. The arrivals are determined by a Poisson process and the service times of the jobs are independent exponentially distributed. We consider both the situation where service requirements by both types of jobs are equal and unequal. The first situation is solved to optimality, for the second situation we find the best policy within a certain class of policies. Optimal schedules always keep part of the service capacity free for arriving foreground jobs. Applications of this model can be found in computer systems, communication networks and call centers. CD001602.PDF (From Author) TOP Asymptotics For Polling Models With Limited Service Policies Authors: Woojin Chang, Douglas G. Down, Volume: 1, Page 1427 Paper number 1604 Abstract: In this paper we find exact asymptotic expressions for the event that the total queue length is large for a general limited service, exponential polling model with equal service rates and two classes of customers. It is found that this behaviour divides into two very different regimes, depending on the arrival rates to the system. CD001604.PDF (From Author) TOP A Framework for Simulation-Based Network Control via Hindsight Optimization Authors: Edwin K.P. Chong, Robert L. Givan, Hyeong Soo Chang, Volume: 1, Page 1433 Paper number 1605 Abstract: We describe a novel approach for designing network control algorithms that incorporate traffic models. Traffic models can be viewed as stochastic predictions about the future network state, and can be used to generate traces of potential future network behavior. Our approach is to use such traces to heuristically evaluate candidate control actions using a technique called hindsight optimization. In hindsight optimization, the finite-horizon ``utility'' achievable from a given system state is estimated by averaging estimates obtained from a number of traces starting at the state. For each trace, the utility value of the state is estimated by determining the optimal ``hindsight control''---this is the control that would be applied by an optimal controller that somehow ``knew'' the whole trace beforehand---and then measuring the utility obtained under that control. Averaging over many samples then gives a simulation-based ``hindsight-optimal'' utility for the starting state that upper bounds the true utility value of the state. This technique for estimating state utility can then be used to select the control---simply select the control that gives the highest utility. Our hindsight-optimization approach to designing simulation-based control algorithms can be applied to a wide variety of network decision problems. We present empirical results showing effectiveness for two example control problems---multiclass scheduling and congestion control. CD001605.PDF (From Author) TOP Empirical Methods for Call Admission Control Authors: Raman K. Mehra, Michael Perloff, Volume: 1, Page 1439 Paper number 1606 Abstract: Effective Call Admission Control (CAC) methods are needed to provide Bandwidth on Demand and meet Quality of Service requirements in ATM systems. Current methods are often excessively strict, allowing available bandwidth to go unused, and depend on accurate estimation of traffic description parameters. We formulate and test a simple method for estimating Cell Loss Rate from discretized queue size measurements and using the estimate in Admission decisions. Simulation test results show that our estimation method will improve performance of current methods by correcting for incorrect traffic parameter estimations. CD001606.PDF (From Author) TOP