The Research Of Encryption And Decryption Algorithms In The Embedded Trusted Computing Platform

With the embedded devices widely used in our daily life as well as the develop-ment of network technology, Embedded mobile devices such as cell phones、handheld tablet and so on can be easily connected to the existing network, and along with the rise of the Internet of Things, embedded devices between the Internet can facilitate the realization. As a result, users can directly access and manipulate the embedded devices through the network, and the virus also can attack and destroy the embedded devices from the network by damaging the operating systems and applications for embedded devices, taking up system resources, and even undermine the embedded terminal hardware resulting in irreparable failure. Therefore, the traditional security issues are no longer limited to home PC and other computing ability of the equipment. Security issues also need to be resolved in the computing limited embedded platform. But the computing limited embedded platform is not good at computing some complex algorithms, so we could not directly transplant the security mechanisms of the ability of PC and other computing devices to the embedded platform.With the introduction of the concept of trusted computing, in the international the Trusted Computing Group (TCG) proposed a set of programs to solve a variety of computing devices, including security problems of the traditional PC equipment and a variety of embedded devices. The core idea of their program is the credibility mechanism through the RSA asymmetric cryptographic algorithm to guarantee the identity of the different platforms and data security in the communication process. But the RSA asymmetric cryptographic algorithms have the low efficiency of the implementation in the embedded trusted computing platform. Adopting the AISC method to implement the cryptographic algorithms will need higher hardware costs, but the embedded computing devices often require the lower the better hope that the cost of the computing platforms. So, this topic uses a relatively new instructions optimized method that extend the existing instruction set to optimize the encryption and decryption algorithm in the Trusted Computing Platform. This method can make the cryptographic algorithm efficient execute by using small-scale hardware resources comparing to the ASIC method. At the same time, after research and analysis the existing trusted computing platform specifications described by TCQ we find that the definition of Trusted Computing Module (TPM) is in the lack of the support for symmetric ciphers. Symmetric ciphers and asymmetric cryptographic algorithm have their own advantages in encryption and decryption of data, and particularly the lower the efficiency of the implementation of the RSA asymmetric cryptographic algorithm, when the encryption and decryption of data blocks is very large, even using the pure hardware implementation cannot meet the embedded trusted platform for real-time terms of the requirements. Therefore, this subject introduces the American Nation Standards of Technology Advanced Encryption Standard (AES) to the trusted computing platform, and proposes new embedded computing platform architecture based on symmetric ciphers and asymmetric cryptographic algorithms.Because this study is based on32bit trusted computing platform, so when analysis the encryption and decryption in the trusted computing platform, we first use the Very Long Instruction Word (VLIW) to optimize the software implementation of the encryption and decryption algorithms. After that, implement the software optimization algorithm by high-level language such as C language followed by dividing the high-level description of the algorithm into Basic Instruction Block (BIB), and each of the BIB can be converted to a Data Flow Graph (DFG). We find the optimization of the instruction from the DFG Our predecessors have been concluded many ways for finding the optimization of the instruction, so after research and analysis the automatic custom instruction generation methods, we summarize our own method for finding the optimization of the instruction. This subject uses this method to find the optimization of the instruction in the DFGAfter finish finding the optimizing of the instruction based on the cryptographic algorithms in the embedded trusted computing platform, this subject uses a newer Electronic System Level (ESL) method to design the Application Specific Instruction Processor (ASIP) using LISA language. Combined with the previous instruction optimization method, this project propose a set of overall model design validation ideas. Then transform the LISA language designed ASIP model into the Verilog RTL-level description of the processor model, and simulate the processor model by ModelSim tool. Finally synthesize、place and route the RTL model into Gate-level using the Synplify tool, and program the Gate-level model into the Xilnx Virtex5LX110T verifying the validity of the model.