An improved dictionary cracking scheme based on multiple GPUs for Wi-Fi Network

The Internet has penetrated all aspects of human society and has promoted social progress. Cyber-crimes in many forms are commonplace and are dangerous to society and national security. Cybersecurity has become a major concern for citizens and governments. The Internet functions and software applications play a vital role in cybersecurity research and practice. Most of the cyber-attacks are based on exploits in system or application software. It is of utmost urgency to investigate software security problems. The demand for Wi-Fi applications is proliferating but the security problem is growing, requiring an optimal solution from researchers. To overcome the shortcomings of the wired equivalent privacy (WEP) algorithm, the existing literature proposed security schemes for Wi-Fi protected access (WPA)/WPA2. However, in practical applications, the WPA/WPA2 scheme still has some weaknesses that attackers exploit. To destroy a WPA/WPA2 security, it is necessary to get a PSK pre-shared key in pre-shared key mode, or an MSK master session key in the authentication mode. Brute-force cracking attacks can get a phase-shift keying (PSK) or a minimum shift keying (MSK). In real-world applications, many wireless local area networks (LANs) use the pre-shared key mode. Therefore, brute-force cracking of WPA/WPA2-PSK is important in that context. This article proposes a new mechanism to crack the Wi-Fi password using a graphical processing unit (GPU) and enhances the efficiency through parallel computing of multiple GPU chips. Experimental results show that the proposed algorithm is effective and provides a procedure to enhance the security of Wi-Fi networks.