Development of an Integrate E-Medical System Using Software Defined Networking and Machine Learning

Abstract

Scholars and medical professionals have recognizes the importance of electronic medical monitoring services for tracking elderly people's health. These platforms generate a large amount of data, requiring privacy and data security. on the contrary, Using Software Defined Networking (SDN) to maintain network efficiency and flexibility, which is especially important in the case of healthcare observation, could be a viable solution. Moreover, machine learning can additionally utilized as a game changing tool which incorporated with SDN for optimal level of privacy and security. Even so, integrating SDN into machine learning, which heavily relies on health sensors of patients, is incredibly difficult. In this paper, an Integrate Medical Platform (IMP) with a focus on SDN and Machine learning integration is proposed. We produce a platform that reduces complexity by identifying high level SDN regulations based on the extracted flow classes and utilizing machine learning traffic flow classification techniques. F or various types of traffic, We employ supervised learning approaches based on models that have already been trained. We use four algorithms for supervised learning: Random forest, Logistic Regression classifiers, K-NN, and SVM, with different characteristics. Finally, we evaluated IMP by using accuracy, precision, TPR, TNR, FPR, MAE, and energy consumption.