An integer linear programming formulation for power equalization placement to limit jamming attack propagation in transparent optical networks

Authors

  • Nina Skorin-Kapov,

    Corresponding author
    1. Centro Universitario de la Defensa de San Javier (University Centre of Defence at the Spanish Air Force Academy), Santiago de la Ribera, Murcia, Spain
    • Correspondence: Nina Skorin-Kapov, Centro Universitario de la Defensa de San Javier (University Centre of Defence at the Spanish Air Force Academy), MDE-UPCT, C/Coronel López Peña, s/n, 30720, Santiago de la Ribera, Murcia, Spain.

      E-mail: nina.skorinkapov@cud.upct.es

    Search for more papers by this author
  • Amornrat Jirattigalachote,

    1. Royal Institute of Technology KTH/ICT, Kista, Sweden
    Search for more papers by this author
  • Lena Wosinska

    1. Royal Institute of Technology KTH/ICT, Kista, Sweden
    Search for more papers by this author

ABSTRACT

Transparent optical networks can be highly vulnerable to various physical-layer attacks, such as high-power jamming, which can cause widespread service disruption and even service denial. The propagation of such attacks can be thwarted with wavelength-selective attenuators, referred to as power equalizers, installed at the network nodes. However, employing all nodes with power equalization functionality can lead to substantial costs. In previous work, we proposed a heuristic approach for sparse power equalization placement to limit jamming attack propagation cost-effectively. The approach provides suboptimal solutions quickly; however, it does not guarantee optimality. Because placement of such power equalization is a long-term planning problem affecting the capital expenditures of the network operator, solution quality is more critical than execution time. Thus, in this paper, we propose an integer linear programming formulation for the problem to guarantee optimality in terms of the number of power equalizers placed. Evaluation results show that our proposed integer linear programming formulation is able to solve moderately sized problems in reasonable time. These results also support the efficiency of our previously proposed heuristic by confirming its ability to find optimal solutions for the cases tested. Copyright © 2014 John Wiley & Sons, Ltd.

Ancillary