Ambrose, Barry E. (1995) Dynamic routing of telephone traffic using network management tools. Dissertation (Ph.D.), California Institute of Technology. http://resolver.caltech.edu/CaltechETD:etd-09192006-170430
In this thesis, the consequences of automating network management of telephone networks are examined. The role of network managers is to monitor the network for exceptional conditions and place controls into the network if necessary to deal with these network exceptions. One potential consequence of automating network management is a network which is capable of adjusting itself quickly to changing traffic conditions, also known as a network with dynamic routing. Simulations are used to show that there are benefits to be gained from implementing dynamic routing by automating the actions of the network managers.
In this thesis, the application of learning techniques such as neural networks and linear predictors to the tasks of network traffic management is also examined. Three network management tasks considered are: (i) recognition of traffic patterns in the network (ii) learning suitable thresholds for network congestion control and (iii) time series prediction of trunk group occupancy. It is found that non-linear learning techniques such as neural networks can give small gains over the more standard technique of linear predictors.
|Item Type:||Thesis (Dissertation (Ph.D.))|
|Degree Grantor:||California Institute of Technology|
|Division:||Engineering and Applied Science|
|Major Option:||Electrical Engineering|
|Thesis Availability:||Restricted to Caltech community only|
|Defense Date:||31 May 1995|
|Default Usage Policy:||No commercial reproduction, distribution, display or performance rights in this work are provided.|
|Deposited By:||Imported from ETD-db|
|Deposited On:||19 Sep 2006|
|Last Modified:||26 Dec 2012 03:01|
- Final Version
Restricted to Caltech community only
See Usage Policy.
Repository Staff Only: item control page