Broadband Traffic Modeling - Essay Example

Network Design Assignment Submitted by Abdl Table of Contents Part I: A report on Broadband Traffic Modeling 3 Objective 3 2. Broadband Traffic Modeling 3 2.1 Introduction 3 2.2 Background 3 2.3 Influence of Traffic Modeling on Network Performance and Design 4 2.3.1 Window sizes 4 2.3.2 Multi-Domain 4 2.3.3 Frequent Reconfigurations 5 2.4 Traffic Modelling Approach 5 2.4.1 Markovian Modulated Poison Process 5 2.4.2 M/Pareto 6 2.4.3 Regression 6 2.5. Broadband Characteristics 6 2.5.1 Multimedia 6 2.5.2 Multi-Point 7 2.5 Conclusion 7 References 8 Part 1: A Report on Broadband Traffic Modeling. 1. Objective: The central objective underlying this report is to explore and review the efficacy of broadband traffic modeling based on the works present by Addie, Zukerman and Neame (1998) in (Broadband traffic modeling: simple solutions to hard problems) as well as some other relevant papers. 2. Broadband Traffic Modeling: 2.1 Introduction: Broadband traffic modeling is a technology of integrated communication network, capable of transmitting diversified communication services in a single network. This supports a high-speed, effective transmission of data, voice, images and videoconferencing across the network in a real time simulation with zero waiting states. Broadband also enhances fair computation, reliable storage and easy manipulation of data responsible for maintaining an acceptable quality of service (QoS) level that is deliverable by the network. Background: Communication infrastructure modes, dates back approximately two thousand years ago. This is the time when Beacon systems couriers and advanced communication networks were engineered Owing to the monumental economical demands of the modern world. Traditional networks have been overridden by the technological renaissance that enhances effective Asynchronous Transfer Mode (ATM) scheme, across the network. Traditional infrastructures were developed on separate networks with unnecessary detractors. Voice in a telephone network for example is too noisy and inefficient for burst data communication. Concurrently data networks, that stores and forward massages using computers have a limited bandwidth connectivity that cannot transmit digital images and voice simultaneously. 2.3. Influence of Traffic Modeling on Network Performance and Design. Fibre Optical Transport Network, based on Wavelength Division, Multiplexing and Optical Cross-Connects technology, offer a tremendous transportation capacity. It calls for a frequent manual intervention for updated management. Due to the inconsistency of the Traffic pattern on an integrated system. Network operators are advised to adopt an optical network a smart, automatic real-time system (ION) or Automatic Switched Optical network, for tractability of the changing traffic load information. 2.3.1 Window sizes A smaller observation window harms the network stability; while a too large observation window worsens the network reliability. Research shows that a suitable traffic observation window size improves the offered Quality of Service (QoS) by reconfiguring the logical layer network at the right time and in the right way. Therefore undesirable observation window size might give less accurate information and trigger the operation at the wrong 2.3.2 Multi-Domain. The performance of multi-domain networks, such as differentiated services (DiffServ), guarantees quality of service across the network. Surprisingly network providers hesitate to use it, since enabling and maintaining it on their production networks requires a lot of new know-how and policy management, and they doubt that it will significantly improve the performance and reliability of their over-provisioned networks, or enable new applications. Hence the proposition, to simulate a model of a network connection, between two computers over a fixed multi-domain route in which the network connection passes different service providers. It corresponds to a real network connection that is used for an existing application. To reduce the complexity of the simulation, Routers are put in place to model traffic on all the connections along the fixed path instead of modeling the whole system. 2.3.2 Frequent Reconfigurations. Observations made above in the multi-layer traffic engineering. Clearly shows that the time traffic information is monitored has a great impact on the network performance. Consequently when the observation window is small, the network topology is unstable, thus worsening grooming stability. However, a too large observation window might make the network topology quite stable, but is lazy in reacting to the traffic bursts and gills. 2.4 Traffic Modelling Approach. Traffic Model approach is the process used to predict a real traffic stream flowing through a network. This helps in determining the analytical behaviour of the network and also in finding the hardware expositions of the attributes shown by traffic in quantified data, prompting a qualitative output execution. Markovian Modulated Poison Process. Markovian is a traditional short-range simple model that mimics the real behaviour behind the interaction between users, protocols and the network. It estimates the Long Range Dependence (LRD). Synchronizing the characteristic of the Internet traffic exiting from the Internet through an Edge Router.i.e average load, packet flow size etc. Based on the notions of sessions, flows and packets, typical of real network applications. As a result it has the capabilities to predict the network behaviour as the traffic characteristics change. M/Pareto Model: Fractal traffic is characterized by significant long bursts [1]. Such bursts are caused by downloading large files, or long periods of high levels of VBR video activities, intensive bursts of database activity, and so on. It is therefore appealing to model fractal traffic by a model involving such long bursts. Regression. Regression models are used in simulations to define the next random variable in the sequence by an explicit function of previous random variables. Hence, they are used to model sequences that do not vary much between successive observations, e.g. number of bits/frame of teleconferencing VBR video. AR, ARMA, DAR, and TES Regression processes can only model stationary processes, while ARIMA regression process can model both stationary and some non-stationary processes. Regression querying is intractable. Broadband Characteristics The ability for the integrated systems, to enhance digitalized voice, data transfer, video images and facelift the economical implementation for diversity of services in a congested system. 2.4.1 Multimedia Communicating of audio, data, still images or motion video or a combination of these media. Based on the bandwidth, signal latency within the network and signal fidelity upon delivery by the network. Consequently, information content of each medium may affect the information generated by other media. For example, voice could be transcribed into data via voice recognition and data commands may control the way voice and video is presented. These interactions most often occur at the communication terminals, but may also occur within the network. 2.4.2 Multi-Point: This involves the set-up of connections between more than two people. These connections can be multi-media. They can be one way or two-way communications. These connections may be reconfigured many times within the duration of a call. A few examples will be used to contrast point-to-point communications versus multi-point communications. Traditional voice calls are predominantly two party calls, requiring a point-to-point connection using only the voice medium. To access pictorial information in a remote database would require a point-to-point connection that sends low bit-rate queries to the Database, and high bit-rate video from the database. Entertainment video applications are largely point-to-multi-point connections, requiring one-way communication of full motion video and audio from the program source to the viewers. Video teleconferencing involves connections among many parties, communicating voice, video, as well as data. 2.5 Conclusion: The exponential growth of integrated multiservice networking has dealt a blow on the conventional mode on communication. Taking a close projection on the future leaves no doubt, about the embracing of a high-speed video conferencing information exchange technology that supports Quality of Service (QoS) with a strong network topology. Multimedia attributes with real time simulation are seen as important facets to be considered in designing the transmission systems of the world to come. The change in the paradigm shift from the traditional long distance information transfer to embracing digitalized Asynchronous Transfer mode (ATM) is a milestone improvement that is promising in the economical basis. References: Addie R.G., Neame T.D, & Zukerman M., (June 1998) "Applying multiplexing Characterization to VBR video traffic," in Proceedings of ITC 16, pp. 847-856. Andersen A. T., Nielsen B. F., (1998) "A markovian approach for modeling packet traffic with long-range dependence," IEEE journal on selected areas in communications, Vol.16, No.5, pp. 719-732, 1 Crawley, E. et al., "A Framework for QoS-based Routing in the Internet", RFC 2386, Aug. 1998. Feldman. A., Gilbert .A. C., Kurtz T.G & Wilinger. W. (April.1998) "The changing nature of network traffic: Scaling phenomena," Computer Communication Review, vol. 28, no. 2, pp. 5-29. IEEE Communications Magazine, (Aug 1998) Volume: 36 Issue: 8, Pages: 88 -95. Read More