The present paper suggests a hybrid solution to address two key authentication challenges: data privacy and the limitations of mobile devices resources. The former is addressed using Partially Homomorphic Encryption based on Paillier Algorithm for the encryption. While the latter is handled using a combination of Deep Convolutional Neural Network and Local Ternary Pattern for face recognition. We compare the accuracy and performance of our proposed solution to others proposed by the literature on ORL dataset and Extended Yale data set. Our findings suggest our proposed methods return higher recognition rates including 98.75% on encrypted ORL data set and 98.78% on encrypted Extended Yale data. In contrast, the existing methods achieved lower recognition rates, which have achieved only 92.50% and 95.44% of recognition rates on encrypted ORL and Extended Yale data set, respectively.

Remarkable progress in the areas of deep learning and the mobile cloud environment. The security has become most challenging part on mobile cloud., this is due to the growth of persons who use their mobiles for access to the services provides by cloud such as health care., storage., games and so on. This paper aims to suggest using deep learning framework for mobile “Tensorflow lite” to make the recognition on mobile without the need of the cloud or computational resources and to avoid that users sends their picture for each time. Improving the accuracy of the work that have been done with 99.50 % of accuracy to 100 % by applying the fine-tuning method using Keras library in Python.

Security in Mobile Cloud Computing (MCC) has become mandatory nowadays. This is due to the increasing use of mobile devices for accessing various accounts (e.g. Health record, Gaming, Facebook, Gmail and so on). Many researchers proposed biometric authentication in MCC, with a classical model for training and classification like using Local Binary Pattern (LBP) for the extraction of features and Support Vector Machine (SVM) for classification and so on, Deep Convolutional Neural Network (DeepCNN) outperform classical models in a number of cases. This paper aims to propose a Deep authentication model using Biometric Face Recognition based on DeepCNN in MCC. The proposed model uses the front camera of a mobile device to take a picture of the user and upload it to the cloud for computation (features extraction using DeepCNN). Due to the huge data and complex computation in deep authentication, we propose to allocate the training process to the cloud. The proposed authentication process achieves 99.50 % of accuracy. This method is implemented in Python using Keras library for deep learning.