Planning Beneficial and Profitable Network Upgrade Paths

  • I B Crabtree
  • D Munaf
Part of the BT Telecommunications Series book series (BTTS, volume 8)

Abstract

The basic task is to produce a plan of maximum cost-effectiveness that upgrades switches in a communications network. While upgrading all switches immediately achieves maximum customer satisfaction, it is prohibitively expensive. The task is clearly a constrained resource allocation problem. This chapter describes briefly a number of novel approaches to searching that are suited to this class of optimization problem, and shows the results obtained using simulated annealing on a large multi-user network. The integration of this technique into a tool to support this type of network planning is also described. In terms of problem modelling, as outlined in Chapter 1, the tool performs the search for the optimal solution and also allows sensitivity analysis and what-if analysis to be performed on resulting solutions.

Keywords

Genetic Algorithm Tabu Search Customer Satisfaction Vehicle Route Problem Minimum Coverage 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Metropolis N et al: ‘Equation of state calculations by fast computing machines’, Journal of Chemical Physics, 21, pp 1087–1092 (1953).CrossRefGoogle Scholar
  2. 2.
    Buchanan J T (Ed): ‘Some novel approaches to resource allocation problems’, University of Strathclyde, Centre for Network and Resource Management (October 1993).Google Scholar
  3. 3.
    Prosser P, Muller C and Brind C: ‘A preliminary study of stochastic search techniques applied to vehicle routeing problems’, University of Strathclyde (February 1994).Google Scholar

Copyright information

© British Telecommunications plc 1996

Authors and Affiliations

  • I B Crabtree
  • D Munaf

There are no affiliations available

Personalised recommendations