Internet routing control by link weights optimization with traffic engineering objectives is studied.It is modeled as a mathematical programming problem with the objectives of network congestion minimization and delay minimization.It is a NP-hard problem

and a chaotic group search algorithm based on local search is proposed

which makes use of chaos variable to find initial solutions with favorable distribution

and applies extended greedy principle to local search process to increase global search ability.Simulation results manifest that it improves network performance with decreased congestion probability caused by unbalanced traffic distribution

and restricted extra path hops.