Differential Power Analysis Of The Stream Ciphers With Filter Structure

Recent years, in the wake of fast developments in kinds of embeddedsystems (e.g. smart cards), their applications are increasingly wide.Followed by this fact, the security issues of those embedded systems arepaid more and more attentions. Side-channel attack (SCA) which is one ofthe most representative approaches in the terms of attacking on theembedded systems is any kind attack based on the information leakagesfrom the software or hardware implementation of the cryptosystem, ratherthan brute force or theoretical weaknesses in the algorithms. Among kindsof SCA, power analysis techniques are well-known and powerful.The main goal of the present master thesis is to develop generaldifferential power analysis (DPA) techniques which can be employed toattack the stream ciphers based on the filter structure, with linear feedbackshift registers (LFSRs) and filter functions. The results of the analysisillustrate the vulnerabilities of LFSRs in the analysis of DPA. As anexample, the technique is also applied to attack a stream cipher Crypto-1which is widely used in real life. After developing a full hardwareemulation platform, the present thesis first implemented Crypto-1byhardware emulation. Only through the analysis of7*128power traces, thecomplexity of exhaustive search to Crpyto-1is reduced to220from248.In final part of the present thesis, several potential countermeasuresare discussed and their effectiveness and cost are considered.