Distributed aggregation/disaggregation algorithms for optimal routing in data networks.

A gradient projection algorithm using iterative aggregation and disaggregation is proposd and analyzed for box-constrained minimization problems. In a variation of the distributed computation model, the algorithm is shown to converge. The algorithm is applied to optimal routing in a large interconnected data communication network. The aggregation/disaggregation method proposed results in a multilevel hierarchical clustering of a large network, which fits naturally the hierarchical topological structure of large networks. A numerical simulation of a 52-node network shows that the serial version of the algorithm achieves 35% saving of computation time as compared to a path-formulated gradient projection code developed by D. P. Bertsekas, B. Gendron and W. K. Tsai (1984), which is among the fastest existing programs for path-formulated optimal routing.

Distributed aggregation/disaggregation algorithms for optimal routing in data networks. Conference Paper