Research On Memory Integrity Protection Technology

With the popularity of the computer application,the information stored in the computer is more and more.As a result,the security of the data stored in the computer must be protected.This method mainly protect the integrity and confidentiality of data stored in the memory.And at the same time,it can detect whether the data is tampered with by the attacker by the way of integrity.The attacker can not understand the meaning of the data even he obtains the data by the way of confidentiality.Firstly,this paper introduces the achievements of the previous researchers in the terms of integrity and confidentiality.It protects the confidentiality of the data by encrypting the data and often uses OTP(One-time pads)to encrypt the data as the delay is very small.MAC(Message Authentication Code)and hash function are proposed by the researchers to protect the integrity of the data.When the computer writes data into the memory,it writes the data as well as the MAC of the data into the space.While the computer reads the data,it firstly reads the data and the MAC.Later the computer re-computes the MAC of the data and compares the two MACs.The data is correct if they are the same,otherwise the data is tampered with.The advantage of the two methods are simple,the drawback is they can't resist against the replay attacks.So the tree mechanism is proposed.As the overhead of the traditional integrity method is very larger,the Unbalanced Hash Tree Integrity Protection is put forward in this paper.Firstly,it introduces the basic principle and the advantages of this method.Secondly,the paper proposes the questions which must be solved in order to realize this method.Then the rules and the algorithm of the way is introduced.Finally,using the simulator Simplescalar tool set 3.0 simulates our method,and compare it with other integrity protection methods(base,2-ary Tree and Merkle Tree).Simulation results show that performance overhead of this method is lower than traditional Merkle Tree and the CHTree.The advantage is not obvious when the second level cache is not large.By making some improvements this paper puts up with Dynamic Cached Hash Tree Integrity Protection which takes advantage of the principle of locality of the program.So the length of the data which accessed more frequently is shorter than that accessed less frequently.The basic principle and the advantages of our way are firstly introduced.Secondly,the paper proposes the questions we have to solve in order to realize our method.Then the rules and the algorithm of the way is introduced.Finally,this paper proves that it can resist against physical attacks such as spoofing attacks,relocation attacks as well as reply attacks.So our method is security.