Unimodal biometric system competes with several problems like noisy data, spoof attacks, changes in intra-class, restricted degrees of freedom, non-universality, and unacceptable error rates. Several of these limitations can be handled by the application of multimodal biometric systems that integrate multiple sources of information. Multimodal biometric systems involve the fusion of information from many biometrics. These biometric technologies are gaining great attention nowadays due to its possibility to overcome limitations in unimodal biometric systems. This paper presents an overview of multimodal biometrics used in multiple levels of security, it also includes modules of multimodal biometric system, different levels of fusion in multimodal biometrics and related work. The area to improve upon the fusion approach of multimodal technology is suggested as future work for researchers.