Achieving flatness: Honeywords generation method for passwords based on user behaviours

Abstract

Honeywords (decoy passwords) have been proposed to detect attacks against hashed password databases. For each user account, the original password is stored with many honeywords in order to thwart any adversary. The honeywords are selected deliberately such that a cyber-attacker who steals a file of hashed passwords cannot be sure, if it is the real password or a honeyword for any account. Moreover, entering with a honeyword to login will trigger an alarm notifying the administrator about a password file breach. At the expense of increasing the storage requirement by 24 times, the authors introduce a simple and effective solution to the detection of password file disclosure events. In this study, we scrutinise the honeyword system and highlight possible weak points. Also, we suggest an alternative approach that selects the honeywords from existing user information, a generic password list, dictionary attack, and by shuffling the characters. Four sets of honeywords are added to the system that resembles the real passwords, thereby achieving an extremely flat honeywords generation method. To measure the human behaviours in relation to trying to crack the password, a testbed engaged with by 820 people was created to determine the appropriate words for the traditional and proposed methods. The results show that under the new method it is harder to obtain any indication of the real password (high flatness) when compared with traditional approaches and the probability of choosing the real password is 1/k, where k = number of honeywords plus the real password.