Parallel Processing of Big Heterogeneous Data for Security Monitoring of IoT Networks

Abstract

Networks of the Internet of Things (IoT) nowadays find greater widespread in many domains. Particularities of creation of IoT make the problem of their security monitoring rather actual, it is caused by necessity of processing of big amounts of heterogeneous data in real time. The problem may be solved by means of implementation of the parallel system for security data processing within IoT on the fly basing on complex event processing (CEP) technology. The paper considers basic solutions for creating such a system. The proposed system is oriented for usage of software environment Hadoop and includes data collection, data storage, data normalization and analysis, and data visualization components. The paper discusses the issues of architecture of this system, its implementation and experimental estimation. The experiments showed that the proposed approach to creation of the system on the basis of CEP technology provides suitable scalability and is capable to meet the requirements of processing data on security events in real time in IoT.