Research And Practice Of Side Channel Attacks Based On Clock Synchronous Acquisition

Side-Channel Attacks challenge the security of cryptographic algorithms greatly,and have attracted extensive attention and research from researchers in the security field.The collection of Side-Channel information as the first step in a Side-Channel Attacks are the basis for subsequent analysis and attacks.The method of the asynchronous collection reduces the signal-to-noise ratio of the Side-Channel information,and also makes it impossible to align the Side-Channel information in the future,which causes great difficulty in attacking the Side-Channel information.Therefore,it is very important to greatly improve the collection efficiency of Side-Channel information and study the Side-Channel Attacks method based on the Side-Channel information.The main research contents and results of this article are as follows:1.For the traditional asynchronous Side-Channel information misalignment problem,this paper qualitatively analyzes the cause of the problem and proposes a leakage assessment method for the asynchronous Side-Channel information problem,which proves that the synchronous acquisition method improves the signal-to-noise ratio of the Side-Channel information.2.For the problem of traditional asynchronous Side-Channel information acquisition,this paper proposes a clock synchronization logic design based on Phase-Locked Loop with the configurability of FPGA clock resources,and builds a clock synchronization acquisition platform.For the problem of high coupling noise of the synchronous acquisition platform,this paper proposes a design of the coupling circuit in combination with the characteristics of the optocoupler chip to isolate the noise.Under the optimal digital oscilloscope sampling clock frequency,the single-point Signal-toNoise Ratio of the traces improves by 58.9%.3.For the inefficiency of the traditional non-modeling method of asynchronous Side-Channel information,this paper proposes a nonmodeling type of Side-Channel Attacks method for synchronous SideChannel information.Guessing the entropy index proves that the synchronous Side-Channel information is easier to implement Side-Channel Attacks.Under the optimal digital oscilloscope sampling clock frequency,the absolute maximum correlation value of the conventionally collected Side-Channel information is 0.039,and the maximum correlation absolute value of the synchronously collected Side-Channel information is increased to 0.12,which is at least three times the original increase rate.4.This paper explores the machine learning Side-Channel Attacks of synchronous Side-Channel information.With some different evaluation characteristics of algorithms such as classification accuracy,confusion matrix,and guessing entropy,it proves the limitation of machine learning to asynchronous Side-Channel information attack.The support vector machine evaluation results show that the synchronous acquisition method proposed in this paper improves the guess entropy of the Side-Channel Attacks by at least 15.