Multimodal Biometrics as Attacks Measure in Biometrics Systems Ifeoma U.Ohaeri, Michael Esiefarienrhe, Naison Gasela oh.ifeoma@yahoo.com, Department of Computer Science North-West University, Mafikeng Campus, Private Bag X2046, Mmabatho 2735, South Africa. Michael.Esiefarienrhe@nwu.ac.za, Department of Computer Science North-West University, Mafikeng Campus, Private Bag X2046, Mmabatho 2735, South Africa. Naison.Gasela@nwu.ac.za, Department of Computer Science North-West University, Mafikeng Campus, Private Bag X2046, Mmabatho 2735, South Africa. *Corresponding author: Ifeoma.U.Ohaeri Abstract: Multimodal systems are considered to be the most dependable among the other types of biometric systems because it is more difficult to forge multiple traits than a single one. Hence, having a fingerprint multi-algorithmic system would not make authentication more secure considering an attacker trying to gain unauthorized access provided with a fingerprint replica. Therefore, multimodal systems are widely deployed and the most generally accepted. Therefore, in this review paper we proposes an integration of fingerprint and face recognition systems to provide anti-spoofing and replay attacks measures making it difficult for an impersonator/imposter to steal multiple biometric traits of a genuine user. Keywords: Biometric, Multimodal Systems, Verification, and Authentication. 1. Introduction Biometric system is automated recognition of persons based on their biological and behavioral characteristics. This technique of physical and logical access controls to protect information systems from security threats such as spoofing and replay attacks is becoming increasingly popular compared to traditional token-based or knowledge- based techniques. One of the main reasons for this popularity is the ability of the biometrics technology to differentiate between an authorized person and an intruder who fraudulently acquires the access privilege of an authorized person. Two primary uses of biometrics would be identity verification and user identification. Identity verification is a one-to-one comparison of a person’s ’biometric template’ with his or her original previously stored in the system. The verification result is a “match”/“no-match”, or a similarity measure or class membership degree [1]. Impersonation is a very big security threat to biometric systems. This is performed by the use of artifacts or by finding an existing person with a similar biometric data and then fraudulently assuming that identity to spoof a verification check. points out that a biometric-based verification system works properly only if the verifier system can guarantee that the biometrics data came from the legitimate person at the time of enrollment so that during verification when a user claims an identity it is validated by comparing the stored biometric data against their presented biometric features [2]. For example, wolf attacks use a biometric sample such that the similarities between this sample and a number of templates are resulting in high false matches with these templates; wolf attacks took advantage of vulnerabilities on the specific matching algorithms to achieve their purposes [1]. However, the assassination of Al- Mabhouh, a co-founder of the military wing of Hamas in 2010 highlights the risk that sophisticated attackers can undermine existing identification systems by targeting individuals for impersonation. Interpol and the Dubai police believe the suspects which are up to 29 stole the identities of real people [3]. It is therefore important to examine the accuracy of biometric tools when subjected to such attacks. If biometric systems are to prevent these attacks, the systems need to be made complex for impersonations or impostors by combining more than one biometric data for verification and authentication purposes. Int'l Conf. Wireless Networks | ICWN'15 | 189