Performance of TCP over satellite networks under severe cross-traffic using GA

Optimisation of control parameters for genetic algorithms to test computer networks under realistic traffic loads

“…On the other hand, Baldi et al [4] and Karthik et al [5] indicate that neither approach is sufficient to give enough information about the protocol performance. While the FDTs deliver a mathematically proven answer, it is forced to work on an over-simplified view of the protocol.…”

“…Baldi et al [4] and Karthik et al [5] show how to integrate a GA with a network simulator to drive the generation of critical background traffic in a computer network so that a worst case analysis can be done. Baldi et al use the static control parameter values that are shown in Table 1 and Karthik et al use those shown in Table 2.…”

“…As we have indicated, Baldi et al [4] and Karthik et al [5] show how to integrate a GA with a network simulator. The GA aims to generate the worst-case traffic for the protocol under analysis given some constraints on the traffic bandwidth.…”

“…Karagiannis et al [2] and Cao et al [3] argue that internet traffic can be well characterised by using Poisson models. On the other hand, Baldi et al [4] and Karthik et al [5] indicate that the use of Poisson models is a clever choice of background traffic pattern to obtain useful results, and that the worst case analysis cannot be done. They propose a mixed simulation technique in which a genetic algorithm (GA) that explores the solution space is integrated with a network simulator to drive the generation of critical background traffic.…”

A connection-level call admission control using genetic algorithm for multi-class multimedia services in wireless networks

“…Overflow replacement policies in location management are studied in Lou et al (2004). Genetic algorithm is also used to study the performance of TCP over satellite networks in Karthik et al (2004), and other publications Biaz and Ji, 2005).…”