关键词:
IoT
Lightweight cryptography
TEA algorithm
XTEA algorithm
XXTEA algorithm
Fiestel structure
FPGA
ASIC throughput
摘要:
Y The current era of ubiquitous computing has led to the emergence of a sub-domain in cryptography called Lightweight Cryptography, which deals with imparting adequate security to resource constraint devices like IoT devices, RFID tags with a suitable choice of design metrics. The advancement in network connectivity and data handling capabilities shows the tremendous growth of IoT in physical life. The number of connected devices have been increasing throughout IoT applications, leaving severe security concerns behind. There is a need of secure communication which can be fulfilled with lightweight algorithms. The motive of this work is to implement the optimized lightweight ciphers and model its design metrics. Design has to be simulated to implement the cipher in hardware from which different metrics can be measured. TEA, XTEA and XXTEA ciphers have been used to fulfil the objective stated and were modelled, implemented and optimized on specific hardware technology like Field Programmable Gate Array (FPGA) and Application Specific Integrated Circuit (ASIC) platform. Designs have been implemented to examine numerous properties like block sizes, rounds of implementations and key scheduling part. This paper presents four hardware architectures namely TEA (T1), XTEA (T2), XXTEA (T3) and hybrid model (T4). T1, T2 and T3 have been implemented using pipelined method with percentage improvement in frequency is 75.9%, 162% and 89%. Similarly, the enhancement in terms of area has been 85.43%, 57.08% and 90.79%. T4 presents a hybrid model which is also a pipelined architecture that combines TEA family (TEA, XTEA and XXTEA) in a single design. The percentage improvement of gate equivalent (GE) for T2 is 47.50%. The hybrid model (T4) has the same throughput as that of T1, T2 & T3 and has less GE when compared to combined GE of all three. Efficiency improvement of all the novel architectures are more than eighteen times as compared to the existing literature.