On the efficient design of network resilient to electro-magnetic pulse attack-elastic optical network case study

The rising popularity of the telecommunication networks and their growing support in our every-day life make them a potential target of various security threats and malicious attacks. The international security organizations warn against increasing likelihood of nuclear weapon or electro-magnetic pulse (EMP) attacks, which can be extremely harmful also for the networks. On that background, we study efficient design of a network resilient to the EMP attack wherein the required protection level is provided by the application of multipath routing and military grade bunkers (advanced electro-magnetic radiation resilient approaches protecting whole network node) implementation. Formally, we define and study the problem of bunkers location, routing and spectrum allocation (BLRSA) in an elastic optical network (EON). In the problem objective we address two criteria - network resilience (measured by the average lost flow per potential attack) and spectrum usage. For that problem we propose integer linear programming (ILP) model and two dedicated heuristics - 1S-RSA and 2S-RSA. Then, we perform extensive numerical experiments divided into three parts: (i) tuning of the proposed approaches, (ii) comparison with reference methods, (iii) realistic case study - efficient EMP-resilient network design. In the case study we analyze benefits and costs (especially in terms of the spectrum usage) of the proposed protection scheme. Moreover, we also analyze vulnerabilities of three realistic network topologies to the EMP attacks and identify their critical nodes. The investigation proves high efficiency of the proposed approaches and shows that they allow to save up to 90% of the traffic lost in the case of no protection against these types of attacks.

Automated summary

The article investigates an EMP-resistant network design scheme that employs multipath routing and military-grade bunkers for protection. It proposes an ILP model for the BLRSA problem and two heuristics, and conducts extensive numerical experiments demonstrating the efficiency of the proposed methods.