Vedic Multiplier-based International Data Encryption Algorithm Crypto-Core for Efficient Hardware Multiphase Encryption Design

Abstract

At present, there are several pieces of research on designing and implementing new cryptographic algorithms that are lightweight and resistant to several, if not major forms of security attacks. However, some algorithms such as the International Data Encryption Algorithm (IDEA), which has been around for some time is yet to record any real threat against its functionality. To ensure its continued usage, current implementations rely on multiphase encryption where it is combined with other algorithms such as ROTation (ROT) and Data Encryption Standard (DES) for maximum security strength. Multiphase encryption implies that there is a tendency for an increase in hardware area and a reduction in overall speed. In such cases, having fast and reduced area algorithms are much desired. This paper, therefore, proposes an efficient hardware implementation of the IDEA cipher that is based on arithmetic modulo multiplication—one of the main computations of the IDEA—on a novel Vedic multiplier architecture. The increase in efficiency of the IDEA crypto architecture and the reduction in resources utilization is achieved through an enhancement of its structural architecture to utilize a fixed set of resources for all eight identical rounds of computation and the use of a proposed fast and lightweight Vedic hardware multiplier. The proposed hardware modification and resulting architecture are designed using the Xilinx ISE and Vivado tools. The architecture is synthesized using Precision Synthesis Tool (PS) and simulated using Modelsim SE 10.6d and ISIM simulation tools. The proposed IDEA cipher is 100% more efficient when designed based on the Vedic multiplier compared to existing designs. The hardware architecture is implemented on Spartan-6-FGG484 Field Programmable Gate Array (FPGA) using Verilog HDL. Verified results show that the proposed Vedic-based IDEA occupied 212 Slices with the Vedic multiplier only occupying 28 Slices out of the total 212. The proposed architecture operates at a maximum frequency of 253.3 MHz.