Multi objective optimization of VPN design by linear programming with risks models

Abstract

Optimal design of IPsec-based virtual private networks (VPN), in general, depends on multiple factors and parame- ters, such as the VPN's architectural model, hardware and software setups and the technical platform solutions, network topo- logy models, modes of the tunnel's operation, levels of the Open Systems Interconnection (OSI) model, encryption/decryption algorithms, modes of cipher operation, security protocols, security associations and key management techniques, connectivity modes, parameters of security algorithms, computer architectures, the number of tunnels in the VPN, and other factors. This paper presents an innovative approach to using methods of linear programming with risks to solve a multi objective optimiza- tion problem of VPN design. In particular, it describes the proposed conceptual VPN design model, VPN information security space, index of effectiveness for multi objective optimization, a new classification of scales, a system-based approach to risks and mathematical modeling of a risk, a hierarchy of scripts and a theorem of description of passwords, VPN design optimization process and particular procedures, sets of legal and illegal VPN design methods, and a VPN design optimization algorithm based on multi objective optimization by linear programming with risks models. The proposed and developed VPN design optimization algorithm was tested by developing specific VPN design methods for various types of VPN users.