Biography
Enrollment Date: 2009
Graduation Date：2012
Degree：M.S.
Defense Date：2012.05.25
Advisors：Liji Wu
Department：Institute of Microelectronics，Tsinghua University
Title of Dissertation/Thesis：Design and Analysis of A Java Coprocessor for Smart Card
Abstract:
Speeding up financial IC card (bank card) in place of the traditional magnetism card is proceeding in our country. In the beginning, the bank card security is ensured by adding encryption algorithm such as 3DES, RSA and SHA-1, however, hidden trouble is still existing in the system of Melt payment. Secondly, the customer needs for more financial business can be met by one financial IC card.
Benefiting from characteristics of simple, cross-platform and safety, Java technology is suitable for intelligent IC card, so, the technology has become a trend and marketing is broad. A platform is needed to support the operation of the Java language, the platform is Java card Virtual Machine.
The author used a special designed CPU for financial IC card to implement a Java Card Virtual Machine. Due to the operation of Java Card Virtual Machine is on the stack-based rather than on the traditional register-based. Operations of pushing and popping of the stack are more frequent and operation times of some instructions of the stack is not only once. In summary, it is not conductive to reducing the running time and improving efficiency, so its power consumption is difficult to cut down.
The idea of author's design is as follows: the Java card virtual machine instructions are classified into eight types according to the operation of the function and the way of addressing. the 16 bits width of ROM table is used to encode instructions, the encoding instructions are pushed into the optimizer and judged whether it can be optimized or not, the optimization of the instructions are explained directly into executable instruction for local CPU.
The author has completed the 185 instructions of the Java card virtual machine and simulated. Design the ALU module and optimizer module. The number of bytes streaming can be optimized is 32. The typical application is run based on the Design. So that it can reduce running times and instructions, and its power consumption reduced to 2.5 mW.