Tabu search optimization in translucent network regenerator allocation

This paper introduces the Tabu Search optimization algorithm to solve the regenerator allocation problem in translucent networks. The problem consists of finding the minimum number of regenerator nodes which primarily affects the cost of the translucent network. The problem is first solved with an ILP formulation to find the optimal solution without taking into consideration its time performance. The optical reach limit due to the dispersion compensation module and full (static) traffic demand with a 1+1 protection scheme are considered in the network model. The proposed algorithm is then compared with two other heuristics: the maximum infeasibility reduction (MIR) algorithm and the maximum regeneration demand (MRD) algorithm. Numerical results show that the Tabu Search procedure either outperforms or equals the performance of the reference algorithms, while having a lower implementation complexity and comparable convergence speed.