International Journal of Innovative Computing, Information and Control ICIC International c ⃝2019 ISSN 1349-4198 Volume 15, Number 4, August 2019 pp. 1441–1453 IRIS FEATURES VIA FRACTAL FUNCTIONS FOR AUTHENTICATION PROTOCOLS Nadia M. G. Al-Saidi 1,∗ , Arkan J. Mohammed 2 , Razi J. Al-Azawi 3 and Aqeel H. Ali 1 1 Department of Applied Sciences 3 Department of Laser Engineering University of Technology Baghdad 10066, Iraq * Corresponding author: nadiamg08@gmail.com 2 College of Science Mustansiriyah University Baghdad 10052, Iraq Received October 2018; revised February 2019 Abstract. Information technology is rapidly developing, thereby highlighting the need for new protection mechanisms to ensure the authenticity and integrity of communica- tion information over an open network. Identification and authentication protocols have emerged as significant tools in this field. Biometric systems have played an important role in protocol authentication and identification because of their individualized proper- ties used as appropriate measures. The iris has many advantages over other biometrics because of its features, including unique complex pattern, stability, and others. These properties make the iris an excellent choice for confidence authentication. The ability of fractal functions to model and describe complex natural patterns makes them highly interesting for use as a mathematical technique inefficient recognition and fast matching. In this study, the fractal encoding approach is used to transform a scanned iris into a compressed representative form. The authentication is satisfied by encrypting the repre- sentative parameters using one of the proven secure encryption methods to ensure secure transmission and storage. The resulting authentication system is analyzed to show the efficiency and accuracy of the proposed recognition method. Keywords: Fractal, Iterated function system, Iris identification, Authentication 1. Introduction. Authentication is considered as a significant tool for security perfor- mance in computer-based communication. Traditional authentication and identification techniques are based on physical token or knowledge, which have been proven to have numerous drawbacks. A good alternative to these traditional methods is the biometric technique using physiological or behavioral traits. Various methods and techniques to strengthen the capacity of the human identification system have emerged over the years. The individual identification based on fingerprints is the most popular method because of the simple mechanism of the fingerprint. Galton and Henry [1, 2] were the first to work on fingerprint identification. Face recognition is one of the most commonly adopted biometric identification methods. Thorat et al. [3] introduced an initial idea of using facial characteristics in 1964. In 1992, Antonini et al. [4] proposed a new image-coding scheme using the fractal coding method for face recognition, considering features both in space and frequency domains. In 1998, Okada et al. [5] developed a system (ZB-Face) DOI: 10.24507/ijicic.15.04.1441 1441