Research On Decoding WPA/WPA2-PSK With High Speed Based On GPU

WPA(Wi-Fi Protected Access) is a kind of system which is used to protect the safety of wireless network.WPA includes WPA standard and WPA2 standard and decoding WPA/WPA2-PSK(Pre-SharedKey) has high practical value .In order to meet the demand of speed on decoding WPA/WPA2-PSK, this paper propose a method that use CUDA to let WPA/WPA2-PSK decoding algorithm run on GPU(Graphic Processing Unit).This dissertation analyzes the Tesla GPU graphics and computing architecture, and the CUDA computing unified device architecture. It described how to decompose a compute work load to parallel form, and map it to Tesla GPU through the leveled programming model of CUDA.In the implement part of this dissertation, according to the software development process, first two hash algorithm SHA1 and MD5 which is used frequently in WPA/WPA2-PSK decoding algorithm is implemented on GPU using CUDA.Second,use CUDA to let WPA/WPA2-PSK decoding algorithm run on GPU and test the speed,then compare and analysis the speed which are respectively run on CPU and GPU.Finally,the deciphering algorithm which is implemented using CUDA is optimized to make the program more efficient.