Machine Learning for Resource Management in Next-Generation Optical Networks

Project Description

Recent developments in optical networking technology offer the prospect of greater flexibility and configurability over increasingly short timescales in order to address the demands of large capacity, highly bursty, intermittent data transfers, typically in accordance with performance constraints. The aim of this project is to investigate the application of confidence machines to the prediction of highly dynamic traffic behaviour in next generation optical networks and, consequentially, enable these networks to be operated more efficiently. This project is a new collaboration bringing together three strands of recent research: pre-booking resource management, multi-fractal traffic modelling, and confidence machines. As the traffic pattern varies across time granularities, the proposed pre-booking resource management mechanism is hierarchical, whereby the traffic prediction is decoupled into multiple levels. A key point of novelty in the proposal lies in its approach to prediction; namely, the use of confidence information when evaluating plausible alternative resource allocations over a continuum of timescales. Unlike conventional machine learning techniques, the predictions these confidence machines make are hedged: they incorporate an indicator of their own accuracy and reliability. These accuracy reliability measures allow service provider and network carrier to choose appropriate allocation strategies by eliminating unlikely resource demands. Therefore, resource management process can effectively perform a cost-benefit evaluation of alternative actions. The project will employ a "technology-agnostic" approach allowing a number of possible evolution scenarios for next generation optical networking to be considered. The outcome of this research will have important industrial repercussions for optical network efficiency and revenue generation capability, as well as theoretical advances to the evaluation of performance risk in the context of dynamic network behaviour. This latter aspect is likely to have further application, for example, with regard to the performance and resilience of utility computing, and not just the underlying transport.

People

• Dr. Chris Phillips, Principal Investigator, Queen Mary, University of London
• Dr. Zhiyuan Luo, Principal Investigator, Royal Holloway, University of London
• Prof. Jonathan Pitts, Co-Investigator, Queen Mary, University of London
• Ali Hassan, PhD student, Queen Mary, University of London
• Mikhail Dashevskiy, PhD student, Royal Holloway, University of London

Simulator

This file is an optical network simulator developed in OPNET modeler 14.5 in order to simulate dynamic routing and wavelength assignment in WDM/DWDM optical networks like ASON. This read me file explains the steps and pre-requisites for using this simulator.

Publications

• Mikhail Dashevskiy and Zhiyuan Luo, Time Series Prediction with Performance Guarantee, IET Communications, Vol. 8, No. 5, pp.1044-1051, 2011.
• Ali Hassan, Particle Swarm Optimisation for Routing and Wavelength Assignment in Next Generation WDM Networks, PhD Thesis, Queen Mary, University of London, 2010.
• Mikhail Dashevskiy, Prediction with Performance Guarantees, PhD Thesis, Dept of Computer Science, Royal Holloway, University of London, 2009.
• Ali Hassan, Chris Phillips, Zhiyuan Luo, Swarm Intelligence Based Dynamic Routing and Wavelength Assignment for Wavelength Constrained All-Optical Networks, 9th IEEE International Symposium on Communication and Information Technology (ISCIT 2009), Korea, September 2009.
• Mikhail Dashevskiy and Zhiyuan Luo, Prediction of Long-Range Dependent Time Series Data with Performance Guarantee, Fifth Symposium on Stochastic Algorithms, Foundations and Applications (SAGA 2009), October 26-28, 2009, Sapporo, Japan.
• Ali Hassan and Chris Phillips, Chaotic Particle Swarm Optimization for Dynamic Routing and Wavelength Assignment in All-Optical WDM Networks, 3rd IEEE International Conference on Signal Processing and Communication Systems (ICSPCS 2009), Nebraska, September 2009.
• Mikhail Dashevskiy and Zhiyuan Luo, Reliable Probabilistic Classification of Internet Traffic, International Journal of Information Acquisition, Vol. 6, No. 2, pp. 133-146, 2009.
• Mikhail Dashevskiy and Zhiyuan Luo, Predictions with Confidence in Applications, LNCS 5632, International Conference on Machine Learning and Data Mining (MLDM 2009), July 23-25, 2009, Leipzig, Germany.
• Ali Hassan and Chris Phillips, Particle swarm optimization-based DRWA for wavelength continuous WDM optical networks using a novel fitness function, Artificial Intelligence Review Journal, Volume 29, Pages: 305-319, DOI: 10.1007/s10462-009-9142-5, 2009.
• Ali Hassan and Chris Phillips, Improved PSO-based Static RWA Solver Avoiding Premature Convergence, Proceedings of London Communication Symposium 2009, London, UK.
• Mikhail Dashevskiy and Zhiyuan Luo, Network Traffic Demand Prediction with Confidence, IEEE GLOBECOM 2008, 30 Nov-4 Dec. 2008, New Orleans, USA.
• Mikhail Dashevskiy, Aggregating Algorithm for a Space of Analytic Functions, ALT2008, Hungary. (Winner of the Best Paper Award)
• Mikhail Dashevskiy, Zhiyuan Luo, Reliable Probabilistic Classification and Its Application to Internet Traffic, International Conference on Intelligent Computing (ICIC2008), Shanghai, China.
• Mikhail Dashevskiy and Zhiyuan Luo, Guaranteed Network Traffic Demand Prediction Using FARIMA Models, Lecture Notes in Computer Science, Vol 5236, pp 274-281. The 9th International Conference on Intelligent Data Engineering and Automated Learning, 2-5 November 2008. (Winner of the Best Paper Award)
• A. Hassan and C. Phillips, Swarm Intelligence Inspired Routing and Wavelength Assignment for All-Optical WDM Networks, IEEE Proceedings of ICTTA08, pages: 495-496, April 2008, Damascus, Syria.
• Zhiyuan Luo, On-line Network Resource Consumption Prediction with Confidence, ChinaCom2007.
• Mikhail Dashevskiy, Network Traffic Classification Using Venn Machines, EPSRC PGNET 2007, June 2007.
• A. Hassan, C. Phillips, Dynamic Routing and Wavelength Assignment using Hybrid Particle Swarm Optimization for WDM Networks, EPSRC PGNET 2007, June 2007.