Latest Module Specifications
Current Academic Year 2025 - 2026
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Description The aim of this module is to introduce the theory and practice of mathematical network analysis and optimisation methods as they apply to the problems of performance analysis of communications protocols, network dimensioning and capacity planning, network architecture design and traffic analysis in modern large-scale data networks, such as optically switched metro and access networks, datacenter and high performance computing interconnects, and femto-macro cell wireless network architectures. Network analysis is essential to understanding and evaluating the fundamental performance properties (e.g. latency, jitter, throughput, packet-drop rate) of complex network architectures and communications protocols. Network dimensioning methods are essential to planning and deploying large-scale networks under given capacity and cost constraints. This module will cover fundamental theory in probability, stochastic processes, queuing theory, graph theory and optimisation methods and apply them to solving various data network design and performance management problems. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Learning Outcomes 1. Derive key results in queuing and teletraffic theory, as apply to the study of communication network performance analysis. 2. Apply methods from probability and queuing theory to modelling of performance-related behaviour of a range of packet-switched and circuit-switched systems and networks. 3. Apply queuing theory equations to calculate system performance measures (e.g. latency, throughput, packet loss) and to perform basic dimensioning of network resources to meet required performance targets. 4. Develop a number of different probabilistic traffic models and determine their applicability to representing different network traffic types. 5. Formulate a range of different network flow and resource dimensioning problems as mathematical optimisation problems. 6. Apply optimisation theory to solving network flow, routing and resource allocation problems. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
All module information is indicative and subject to change. For further information,students are advised to refer to the University's Marks and Standards and Programme Specific Regulations at: http://www.dcu.ie/registry/examinations/index.shtml |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Indicative Content and Learning Activities
Course Introduction The what and why of network analysis and dimensioning. Typical questions answered by network analysis methods. Typical network design problems solved using dimensioning methods. Overview of the methods and the required mathematical background and tools. Review of Probability, Stochastic Processes and Markovian Queuing Systems Probability spaces, random variables, distribution functions, moment generation functions and transform methods, renewal processes, the Poisson process, continuous-time Markov Chains and Markovian queuing systems. Loss Systems and Applications to Blocking Network Analysis and Design The Erlang-B and Engset systems. Blocking in non-Markovian queues, Equivalent Random Theory (ERT), networks with blocking and the reduced load approximation. Applications to performance analysis of wavelength division multiplexed (WDM) optically-switched networks and hierarchical cellular networks. Quasi-Markovian/Non-Markovian Queuing Models Semi-Markov processes, mean delay and the delay distribution in the M/G/1 queue. Mean delay in G/M/1 and GI/GI/1 queues. Application to analysis of polling networks and Passive Optical Network (PON) performance. Network Traffic Modelling Interrupted Poisson Process (IPP), Markov Modulated Poisson Process (MMPP). Traffic autocorrelation, self-similar traffic, heavy tails and the Pareto distribution. Application to modelling of Internet, circuit-switched and transport traffic. Network Optimisation Theory Linear Programming (LP), Integer Linear Programming (ILP). LP and ILP solution methods and software tools. Problems on graphs, network flow problems, link-path and node-link formulations. Network Design and Dimensioning Problems Network dimensioning metrics, constraints and objectives. Uncapacitated and capacitated flow problems, optical network routing and wavelength assignment problem (RWA), network fairness problems, network topology design. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Indicative Reading List Books:
Articles: None | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Other Resources None | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||