- Title
- Ant colony optimisation-based algorithms for optical burst switching networks
- Creator
- Gravett, Andrew Scott
- Creator
- Gibbon, Timothy B
- Subject
- Distributed algorithms
- Subject
- Ants -- Behavior -- Mathematical models
- Date Issued
- 2017
- Date
- 2017
- Type
- Thesis
- Type
- Masters
- Type
- MSc
- Identifier
- http://hdl.handle.net/10948/18939
- Identifier
- vital:28757
- Description
- This research developed two novel distributed algorithms inspired by Ant Colony Optimisation (ACO) for a solution to the problem of dynamic Routing and Wavelength Assignment (RWA) with wavelength continuity constraint in Optical Burst Switching (OBS) networks utilising both the traditional International Telecommunication Union (ITU) Fixed Grid Wavelength Division Multiplexing (WDM) and Flexible Spectrum scenarios. The growing demand for more bandwidth in optical networks require more efficient utilisation of available optical resources. OBS is a promising optical switching technique for the improved utilisation of optical network resources over the current optical circuit switching technique. The development of newer technologies has introduced higher rate transmissions and various modulation formats, however, introducing these technologies into the traditional ITU Fixed Grid does not efficiently utilise the available bandwidth. Flexible Spectrum is a promising approach offering a solution to the problem of improving bandwidth utilisation, which comes with a potential cost. Transmissions have the potential for impairment with respect to the increased traffic and lack of large channel spacing. Proposed routing algorithms should be aware of the linear and non-linear Physical Layer Impairments (PLIs) in order to operate closer to optimum performance. The OBS resource reservation protocol does not cater for the loss of transmissions, Burst Control Packets (BCPs) included, due to physical layer impairments. The protocol was adapted for use in Flexible Spectrum. Investigation of the use of a route and wavelength combination, from source to destination node pair, for the RWA process was proposed for ACO-based approaches to enforce the establishment and use of complete paths for greedy exploitation in Flexible Spectrum was conducted. The routing tuple for the RWA process is the tight coupling of a route and wavelength in combination intended to promote the greedy exploitation of successful paths for transmission requests. The application of the routing tuples differs from traditional ACO-based approaches and prompted the investigation of new pheromone calculation equations. The two novel proposed approaches were tested and experiments conducted comparing with and against existing algorithms (a simple greedy and an ACO-based algorithm) in a traditional ITU Fixed Grid and Flexible Spectrum scenario on three different network topologies. The proposed Flexible Spectrum Ant Colony (FSAC) approach had a markably improved performance over the existing algorithms in the ITU Fixed Grid WDM and Flexible Spectrum scenarios, while Upper Confidence Bound Routing and Wavelength Assignment (UCBRWA) algorithm was able to perform well in the traditional ITU Fixed Grid WDM scenario, but underperformed in the Flexible Spectrum scenario. The results show that the distributed ACO-based FSAC algorithm significantly improved the burst transmission success probability, providing a good solution in the Flexible Spectrum network environment undergoing transmission impairments.
- Format
- x, 186 leaves
- Format
- Publisher
- Nelson Mandela Metropolitan University
- Publisher
- Faculty of Science
- Language
- English
- Rights
- Nelson Mandela Metropolitan University
- Hits: 459
- Visitors: 492
- Downloads: 44
Collections
NMU School of Computer Science, Mathematics, Physics and Statistics
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | SOURCE1 | Adobe Acrobat PDF | 2 MB | Adobe Acrobat PDF | View Details Download |