Design new symmetric cryptographic algorithm for mobile communication systems

Abstract

The evolution of mobile communication systems began with the 1st generation (1G) in the 1980s, the second generation Global System for Mobile Communications (2G-GSM) in the early 1990s, the third generation Universal Mobile Telecommunications System (3G-UMTS) in 2002, and the fourth generation Long Term Evolution (4G-LTE) technology in 2010. There are more people going online via mobile communication systems, and huge amounts of data pertaining to bank transactions, emails, and voice calls are transmitted every hour. This makes online security paramount, and security systems such as symmetric cryptographic algorithms are used to ensure security of information confidentiality, authenticity and integrity. Symmetric ciphers are ubiquitous in information security of mobile systems, where the SNOW 3G cipher, ZUC cipher, and AES cipher have been used to secure 3G-UMTS and 4G-LTE mobile systems. Certain cryptanalysis techniques (e.g. Chosen-IV attack, Differential Cryptanalysis Technique, and Meet-in-the-Middle with Biclique Cryptanalysis Technique) can be used to attack these symmetric ciphers, therefore, there is a need for more analytical studies to address the security problems associated with these ciphers. This research presents a framework that can be used to investigate the security of symmetric ciphers based on analysing randomness properties, correlation properties, and applying cryptanalysis attacks to search for cryptanalysis weaknesses in the aforementioned ciphers. It focuses on analysing the specifications, components, operation, and the implementation of the three ciphers (SNOW 3G cipher, ZUC cipher, and AES cipher). First, analyse the mobile generations (from mobile 1G - 4G) and cryptographic algorithms used in them. The study is based on certain factors, such as features, services, standards, and techniques adopted in each generation. The randomness properties were analysed and evaluated for the basic components of the SNOW 3G cipher, ZUC cipher, and the AES cipher. The NIST test suit (SP 800-22) was used to evaluate and test their corresponding randomness properties. The correlation properties of different components of the three symmetric ciphers have been investigated, and different cryptanalysis attacks have been applied on the three symmetric ciphers to evaluate the security level of these ciphers. Based on the results in this research, weaknesses were found in the length of IV and secret keys (K) and in the initialisation procedure of the SNOW 3G and ZUC ciphers. The undesirable randomness properties were noted in the ZUC and AES ciphers. Also pinpointed some drawback pertaining to the cryptanalysis in the SNOW 3G and ZUC ciphers. Finally, the Secure Cipher Algorithm (SCAL) was designed and proposed to deal with problems in current symmetric ciphers and enhance the security in mobile communication systems. New techniques introduced in our design of SCAL cipher algorithm. The design includes three stages (Stage#1, Stage#2, and Stage#3) connected in a cascade manner. The proposed design relies on Multi Address Permutation, Multi Address Dynamic Substitution Box (DS-Box) and Multi Address Function. The results showed that the SCAL cipher passed the randomness tests, with random correlations, lack cryptanalysis weaknesses, and is more secure compared to the three symmetric ciphers based on the compared cryptanalysis attacks.