Data envelopment analysis and interdiction median problem with fortification for enabling IoT technologies to relieve potential attacks

作者：Khanduzi Raheleh; Peyghami M Reza; Sangaiah Arun Kumar^*

来源：Future Generation Computer Systems-The International Journal of eScience, 2018, 79: 928-940.

DOI：10.1016/j.future.2017.08.056

• Data envelopment analysis
• Facility fortification
• Disaster relief
• Bi-objective model
• NSGA II

Fortifying the facilities to relieve the impact of disaster can be a significant operation for service industries. This study introduces an integrated assignment-allocation model that combines three issues: customer facility assignments, allocation of fortification resources and data envelopment analysis (DEA). To justify the approach used to deal with a more reasonable solution for real world applications, we consider the outputs of the model as a function of the assignment variables. In new proposed model, the DEA efficiency measure is used to allocate fortification resources and select the best facilities for serving the customers after interdiction operations. For this purpose, we present a new bi-objective optimization model that optimally serves customers' demands in a remarkably efficient manner and allocates fortification resources to relieve potential disaster, including interdiction operations. Also, Internet of Things (IoT) optimizes the performance of the customers and enhances the safety of facilities. Indeed, a new parking management system based on loT is proposed to assist parking drivers in managing and handling parking in the presence of protection and interdiction operations considerably. The proposed model is multi-objective and Non-deterministic Polynomial-time (NP)-hard; so, a weighted metric approach by using the CPLEX solver through MATLAB and Non-dominated Sorting Genetic Algorithm II (NSGA II) are utilized in problem solving. Several numerical instances are generated and solved by CPLEX and NSGA II. Experimental results show that NSGA II solves instances in considerably less time than CPLEX. Besides, it is seen that none of the solutions obtained by these two approaches are not dominated.

• 出版日期2018-2