Showing papers on "Network traffic simulation published in 1994"
TL;DR: An overview of discrete event simulation is given and two important modelling issues that are germane to extant and emerging networks: traffic modelling and rare event simulation are singled out.
Abstract: As new communications services evolve, professionals must create better models to predict system performance. The article provides an overview of computer simulation modelling for communication networks, as well as some important related modelling issues. It gives an overview of discrete event simulation and singles out two important modelling issues that are germane to extant and emerging networks: traffic modelling and rare event simulation. Monte Carlo computer simulation is used as a performance prediction tool and Markov models are considered. >
595 citations
TL;DR: A recently completed single-site study has yielded information about how Internet traffic will evolve, as new users discover the Internet and existing users find new ways to incorporate the Internet into their work patterns.
Abstract: A recently completed single-site study has yielded information about how Internet traffic will evolve, as new users discover the Internet and existing users find new ways to incorporate the Internet into their work patterns. The author reviews existing statistics and studies of network growth, which show that network traffic generally grows exponentially with time, at least until the network carrying capacity is reached. He then describes how he captured and reduced the data used in this study. The following points are also addressed: the overall growth in the site's wide-area traffic; the appearance of periodic traffic; the growth in network use by individual computers or users; and the changing geographic profile of the traffic. The implications and limitations of the results are also summarizes. >
128 citations
TL;DR: The authors define the types of simulation software available for network analyses and also discuss desirable features for such products, and present brief descriptions of seven prominent simulation products.
Abstract: In the last five to ten years, simulation has been used with greater frequency to analyze communications networks, due in part to the increased number of networks in existence and to their greater complexity. The authors define the types of simulation software that are available for network analyses and also discuss desirable features for such products. They then present brief descriptions of seven prominent simulation products, emphasizing their approach to building a simulation model. They also give a four-activity approach for selecting a simulation package for an organization or a particular application. A general reference for the simulation concepts discussed in this article is Law and Kelton (1991). >
88 citations
TL;DR: By using the wrapped hexagonal mesh, the inaccuracy of the output measures can be limited even though the number of cells in the simulation is small, and the same statistical accuracy while using significantly less computation power than required for a simulation without cell wrapping.
Abstract: Eliminating the boundary effects is an important issue for a large-scale personal communication service (PCS) network simulation. A PCS network is often modeled by a network of hexagonal cells. The boundary may significantly bias the ouput statistics if the number of hexagonal cells is small in a PCS network simulation. On the other hand, if the simulation is to be completed within a reasonable time on the available computing resources, the number of cells in the simulation cannot be too large. To avoid the inaccuracy caused by the boundary effect for a PCS network simulation with limited computing resources, we propose wrapping the hexagonal mesh into a homogeneous graph (i.e., all nodes in the graph are topologically identical). We show that by using the wrapped hexagonal mesh, the inaccuracy of the output measures can be limited even though the number of cells in the simulation is small. We can thus obtain the same statistical accuracy while using significantly less computation power than required for a simulation without cell wrapping.
84 citations
31 Aug 1994
TL;DR: Network-based model were developed to predict short term future traffic volume based on current traffic, historical average, and upstream traffic and were shown to be capable of producing reliable and accurate forecasts under congested traffic condition.
Abstract: Network-based model were developed to predict short term future traffic volume based on current traffic, historical average, and upstream traffic It is presumed that upstream traffic volume can be used to predict the downstream traffic in a specific time period Three models are developed for traffic flow prediction: a combination of historical average and upstream traffic, a combination of current traffic and upstream traffic, and a combination of all three variables The three models were evaluated using regression analysis The third model is found to provide the best prediction for the analyzed data In order to balance the variables appropriately according to the present traffic condition, a heuristic adaptive weighting system is devised based on the relationships between the beginning period of prediction and the previous periods The developed models were applied to 15-minute freeway data obtained by regular induction loop detectors The prediction models were shown to be capable of producing reliable and accurate forecasts under congested traffic condition The prediction systems perform better in the 15-minute range than in the ranges of 30- to 45-minute It is also found that the combined models usually produce more consistent forecasts than the historical average >
78 citations
11 Dec 1994
TL;DR: This work looks at traditional approaches to distributed quantitative stochastic simulation and proposes a new scenario, Multiple Replications in Parallel Time Streams (MRIP), that solves the problem in an efficient way and shows that MRIP offers linear speedup of simulation.
Abstract: Quantitative stochastic simulation suffers from the fact that sound simulation studies require very long runlength to obtain the results with sufficient accuracy. We look at traditional approaches to distributed quantitative stochastic simulation and propose a new scenario, Multiple Replications in Parallel Time Streams (MRIP), that solves the problem in an efficient way. An implementation of MRIP in a simulation package AKAROA is also described. AKAROA accepts ordinary (non-parallel), simulation models and creates automatically the environment required for running MRIP on workstations of a local area network. Presented results show that MRIP offers linear speedup of simulation. Limitations of this scenario for running distributed quantitative stochastic simulation are also discussed.
66 citations
TL;DR: A general state-space modeling approach for multiorigin multidestination freeway traffic networks including a variety of connecting routes for each origin-destination pair and a number of well-known control strategies are suggested.
66 citations
Book•
01 Jan 1994
TL;DR: Systems and models simulation common processes in discrete systems the EZSIM environment analysis of simulation inputs and creation of their effectsAnalysis of simulation output applications of simulation simulation tools and the criteria for their selection future directions for simulation.
Abstract: Systems and models simulation common processes in discrete systems the EZSIM environment analysis of simulation inputs and creation of their effects analysis of simulation output applications of simulation simulation tools and the criteria for their selection future directions for simulation.
66 citations
14 Dec 1994
TL;DR: This paper presents a fundamentally different approach for optimal light timing that eliminates the need for an extremely complex open-loop model of the traffic dynamics as a component of the control strategy.
Abstract: A long-standing problem in traffic engineering is to optimize the flow of vehicles through a given road network. Improving the timing of the traffic signals at intersections in the network is generally the most powerful and cost-effective means of achieving this goal. However, because of the many complex aspects of a traffic system-human behavioral considerations, vehicle flow interactions within the network, weather effects, traffic accidents, long-term (e.g., seasonal) variation, etc.-it has been notoriously difficult to determine the optimal signal light timing. This is especially the case on a system-wide (multiple intersection) basis. Much of this difficulty has stemmed from the need to build extremely complex open-loop models of the traffic dynamics as a component of the control strategy. This paper presents a fundamentally different approach for optimal light timing that eliminates the need for such an open-loop model. The approach is based on a neural network (or other function approximator) serving as the basis for the control law, with the weight estimation occurring in closed-loop mode via the simultaneous perturbation stochastic approximation (SPSA) algorithm. Since the SPSA algorithm requires only loss function measurements (no gradients of the loss function), there is no open-loop model required for the weight estimation. The approach is illustrated by simulation on a six-intersection network with moderate congestion and stochastic, nonlinear effects. >
45 citations
01 Jul 1994
TL;DR: A new adaptive protocol for distributed discrete event simulation is proposed that spans the continuum of protocols from conservative to optimistic allowing each process in a distributed simulation to adapt to the specific simulation problem at runtime.
Abstract: A new adaptive protocol for distributed discrete event simulation is proposed. This protocol spans the continuum of protocols from conservative to optimistic allowing each process in a distributed simulation to adapt to the specific simulation problem at runtime. An actual implementation of the protocol has been tested on a network of workstations for a closed queueing system. The results are very favorable and the algorithm has outperformed a conservative and an optimistic protocol in some cases.
41 citations
05 Dec 1994
TL;DR: Experiments with UFNET traffic characteristics show that the costs are such that the proposed method can be employed in actual networks, under moderate load conditions, to achieve traffic neutrality with acceptable overheads.
Abstract: We provide cost estimates for achieving spatial neutrality under realistic network traffic conditions using two methods. Measurements done on the University of Florida campus wide backbone network (UFNET) provide us with considerable experience to model an actual network better. Simulation results show that the algorithm's improvement over padding alone is greater for a sparse traffic matrix than for a uniform random traffic matrix. It accomplishes this by smoothing the traffic matrix by rerouting, reducing the padding overhead required to achieve a neutral traffic matrix. On the other hand, a sparse traffic matrix leads to increased costs over uniform random traffic matrix for both padding alone and for padding with rerouting. Experiments done with UFNET traffic characteristics show that the costs are such that the proposed method can be employed in actual networks, under moderate load conditions, to achieve traffic neutrality with acceptable overheads. >
08 Jun 1994
TL;DR: It is shown that under non-uniform traffic loading a better grade-of-service can be provided to network subscribers than by fixed channel assignment, and the network performance measures used in this paper are the average and maximum probability of blocking across the network.
Abstract: We examine several methods to allocate radio frequency channels across a cellular network. In this paper we consider centralized channel assignment strategies that utilize global network traffic information. The channel assignment methods considered are linear programming, quadratic programming, and a heuristic algorithm to assign resources given the network traffic loading. Each method attempts to assign resources, under frequency reuse constraints, to achieve a desired target performance measure for the network. The network performance measures used in this paper are the average and maximum probability of blocking across the network. We compare the performance of a linear highway network and a planar network using our assignment methods to those of uniform fixed channel allocation (FCA) for non-uniform network traffic loading. We show that under non-uniform traffic loading a better grade-of-service can be provided to network subscribers than by fixed channel assignment. >
TL;DR: The approach is aimed at accurately reproducing dynamic network behavior at a level similar to the models described by Hutchinson and Patten (1986), and Unger and Bidulock (1982) .
Abstract: Simulating trunk and signaling networks allows planners to examine network survivability, capacity, and the impact of new service introductions. The challenge is to accurately predict network performance and behavior for a rapidly changing environment, and for the very large networks that are encountered in the public telephone system. This paper describes the discrete event modeling of SS7 signaling networks that enable dynamic network analysis for very large, highly realistic, network models. The primary distinction between the authors approach and related work is the development of a very detailed model, one that incorporates most SS7 message-handling and network-management procedures. Their approach is aimed at accurately reproducing dynamic network behavior at a level similar to the models described by Hutchinson and Patten (1986), and Unger and Bidulock (1982) . They define network simulation objectives and the SS7 model components and functions, and then give a brief overview of how a specific network model is created, including a description of the model's implementation and how it was validated. They also discuss a model and simulation experiments for a large part of the Ameritech network. >
TL;DR: This paper discusses the application of the network traffic simulation model INTEGRATION to a 35-km section of Highway 401 in Toronto, Canada, and results in a correlation coefficient of estimated and observed link flows of 97.23% are presented.
Abstract: This paper discusses the application of the network traffic simulation model INTEGRATION to a 35-km section of Highway 401 in Toronto, Canada. Results for the eastbound direction from 4 a.m. to 12 ...
02 Oct 1994
TL;DR: Sectoring is offered as basis for building an efficient methodology for autonomous agents in discrete-event simulation and identifying the underlying factors and issues hindering further development are identified.
Abstract: Implementing independent, continuously evolving components (i.e. autonomous agents) in discrete-event simulation while providing a useful and efficient broadbased modeling methodology has proven a challenge to the simulation community. Reasons for these difficulties are innate to the implementation of current accepted simulation modeling practices and theories of clock advances, modeling world view, and network modeling paradigms. Meeting the challenge of implementing autonomous agents requires understanding the underlying factors and issues hindering further development. This paper identifies these factors and offers sectoring as basis for building an efficient methodology for autonomous agents in discrete-event simulation. >
Book•
01 Apr 1994
TL;DR: Directions in teletraffic research ATM multiplexer performance signalling and intelligent network performance analysis of queueing systems B-ISDN traffic descriptors switch overload controls video and data traffic characteristics performance of new network services.
Abstract: Directions in teletraffic research ATM multiplexer performance (1) signalling and intelligent network performance analysis of queueing systems (1) B-ISDN traffic descriptors switch overload controls video and data traffic characteristics performance of new network services B-ISDN traffic and resource allocation ATM multiplexer performance (2) teletraffic analysis of mobile telecommunications performance of networks of queues worst case ATM traffic ATM switch design dynamic routing analysis of queueing systems (2) simulation techniques performance of data networks advances in teletraffic modelling B-ISDN network management and design performance of local and metropolitan area networks analysis of queueing systems adaptive connection admission control in B-ISDN B-ISDN network architecture traffic management and network dimensioning multirate traffic performance congestion control in B-ISDN design of circuit switched networks B-ISDN design issues peak rate policing in B-ISDN robust transmission network design data network protocols and routing.
31 Aug 1994
TL;DR: The model results indicate that METACOR follows the time evolution condition in the considered network, which implies that the macroscopic corridor model is able to cope with different traffic conditions at an acceptable level of accuracy.
Abstract: The paper presents the macroscopic corridor model METACOR, as well as its calibration and validation. METACOR, which was developed through the extension and the integration of two macroscopic traffic flow models, has been applied to the corridor peripherique test site to be used for model validation. The model results (regarding the output trajectories of the 3 traffic variables namely volumes, occupancy rates and speed), indicate that METACOR follows the time evolution condition in the considered network. This implies that METACOR is able to cope with different traffic conditions (fluid, dense and congested) at an acceptable level of accuracy. METACOR is a generally applicable macroscopic simulation tool for corridor traffic. Compared to other available tools, its main advantage is its moderate computational effort. This advantage is the result of a systematic macroscopic approach with destination-specific subflows that avoids iterations for dynamic traffic assignment. >
Patent•
12 Oct 1994
TL;DR: In this article, a facility is provided for enhancing an operation support system so that, based on data generated as a result of an event occurring in an associated telecommunications network, the operations support system can predict the likelihood of the event occurring again in the network.
Abstract: A facility is provided for enhancing an operations support system so that, based on data generated as a result of an event occurring in an associated telecommunications network, the operations support system can predict the likelihood of the event occurring again in the network.
11 Dec 1994
TL;DR: The approach discussed in this paper is the implementation of interval time specifications in the discrete-event model and the construction of an interval time clock for the qualitative simulation model.
Abstract: In this paper, we will discuss implementation issues of a simulation modeling methodology that combines discrete-event simulation with qualitative simulation. Our main reason for doing so is to extend the application of discrete-event simulation to systems found in business for which precise quantitative information is lacking. The approach discussed in this paper is the implementation of interval time specifications in the discrete-event model and the construction of an interval time clock for the qualitative simulation model.
TL;DR: Three traffic network simulation models implemented on massively parallel computer, the Connection Machine CM-2, suggest that parallel computing architecture offers a very promising alternative for the implementation of real-time traffic simulation.
11 Dec 1994
TL;DR: A modular concept of distributed simulation modules and user interface can be found out without much knowledge of the underlying simulation details to identify the optimum network design parameters for a heterogeneous network environment.
Abstract: This paper presents a simulation tool to help the network designer to identify the optimum network design parameters for a heterogeneous network environment. The tool consists of two parts: simulation modules and a user interface. The simulation modules are created using enhancements of an asynchronous, event-driven C-function library to enable the simulation of heterogeneous scenarios. In the future, this module will be distributed to speed-up the simulation runtime. The user interface helps the user to create the design of the network to be simulated. By this modular concept of distributed simulation modules and user interface the optimum network can be found out without much knowledge of the underlying simulation details.
04 Oct 1994
TL;DR: A fuzzy logic self-learning model and adaptive predictor of traffic now is presented, and the results of an application of this predictor on real world telephone links, are shown, and are compared to applications of a math-model and a Kalman-filter predictor.
Abstract: Network management is a real-time observation and control action to optimize the grade of service of a network. More emphasis is being placed on the intelligent automatic exchange for stored program controlled (SPC) systems that are capable of real-time network management. The modeling of traffic characteristics and the prediction of future traffic flow are the first steps to efficient network control and management. This paper presents a fuzzy logic self-learning model and adaptive predictor of traffic now. The results of an application of this predictor on real world telephone links, are shown, and are compared to applications of a math-model and a Kalman-filter predictor.
18 Apr 1994
TL;DR: It will be demonstrated that, in order to fully exploit the potential of the simulation methods, the interconnection of various design teams with each other and remote supercomputers via a broadband network is a prerequisit.
Abstract: Aeronautical industry depends on heavy numerical simulation for performance optimization of aircraft beeing designed. In Europe, aircraft conception more and more depends upon effective cooperation between a number of industry and research partners. An effective way to enhance global competitivity is to enable activation of the best simulation tools existing across the community, for definition of a given product. In the scope of this paper we describe the impact that numerical flow simulation already has on the design process of civil transports at DA. It will be demonstrated that, in order to fully exploit the potential of the simulation methods, the interconnection of various design teams with each other and remote supercomputers via a broadband network is a prerequisit. The PAGEIN soft- and hardware environment, that allows international cooperative working in the field of CFD based on international networking, will be presented
27 Jun 1994
TL;DR: The experimental results indicate that the HNN architecture for providing intelligent control of complex urban traffic networks has strong adaptive properties and it can be built with little knowledge about the signal operations.
Abstract: The goal of this work is to develop a hierarchical neural network (HNN) architecture for providing intelligent control of complex urban traffic networks which are usually nonlinear and hard to model mathematically. Two types of neural networks, such as a global planning network and local control networks, are employed for traffic modeling and control. The experimental results indicate that the control scheme has strong adaptive properties and it can be built with little knowledge about the signal operations. >
07 Dec 1994
TL;DR: In this paper, a hybrid modeling/simulation framework based on the Zeigler's DEVS formalism and associated system theory is presented, in which both accuracy in models and speed in simulation experimentations are obtained.
Abstract: This paper presents a hybrid modeling/simulation framework within which both accuracy in models and speed in simulation experimentations are obtained. Based on the Zeigler's DEVS formalism and associated system theory, the framework is based on the transformation of selected DEVS model into equivalent analytic ones to simulate both analytic and simulation models within a single environment. For high-speed hybrid simulation, the authors extended DEVSim++ which is a realization of the DEVS formalism in C++. To exemplify the proposed approach, the authors demonstrate performance modeling and simulation of a simple communication network. >
01 Dec 1994
TL;DR: The DRACULA traffic simulator can be applied in many areas of urban traffic control and management, such as detailed evaluation of traffic signal control strategies, environmental issues such as air pollution due to emission from vehicles in idling, accelerating, decelerating or cruising, and analyses of the effects of variable demand and supply upon the performance of a network.
Abstract: The DRACULA traffic simulator is a microscopic model in that the vehicles are individually represented. The movement of vehicles in the network are represented continuously and updated every one second.
The network is modelled as a set of nodes and links which represent junctions and streets respectively. Vehicles are generated at their origins with a random headway distribution and are assigned a set of driver/vehicle characteristics (according to user-specified probabilities) and a fixed route. The movement of the vehicles on a network is governed by a car-following law, the gap acceptance rules and the traffic regulations at intersections. They can join a queue, change lane, discharge to another link or exit from the system. The traffic regulation at an intersection is actuated by traffic lights or right-of-way rules.
The inputs to the simulation are network data, trip matrix, fixed-time signal plans, gap-acceptance and car-following parameters. Outputs are in forms of animated graphics and statistical measures of network performance.
The program is written in C-language. All types of vehicle attributes are represented as one entity using the structure data type which provides a flexibility in storing and modifying various types of data. Attributes of nodes, links and lanes are also represented as structures. The large number of variables associated with vehicles and the network imply that the performance of the simulation depends on the size of the network and the total number of vehicles within the network at one time.
The simulator can be applied in many areas of urban traffic control and management, such as detailed evaluation of traffic signal control strategies, environmental issues such as air pollution due to emission from vehicles in idling, accelerating, decelerating or cruising, and analyses of the effects of variable demand and supply upon the performance of a network.
TL;DR: It is found that the traffic characterization, when used to represent the output traffic from a token-ring network, provides significantly more accurate results for the mean packet waiting time at the bridge than when modeling theoutput traffic by a Poisson process.
CNET1