Smart Home Network Traffic Abnormally Detecting Research And Implementation

With the rapid development of information technology, smart home network technology is gradually entering into the ordinary household. However, the related security problem is becoming the main bottleneck of the popularity of the smart home network. Currently, the studies of smart home network security mainly come from traditional network including static protection technology such as authentication and access control, and lack of real-time detection system in comparison. Anomaly Traffic Detection Technology can provide real-time response decision-making to the virus invasion and Illegal equipment join attack by specific decoding agreement and application reduction of smart home network. Therefore, using Anomaly Traffic Detection Technology to have a real-time analysis and rapid response decision-making to the security event of smart home network has a great practical significance in the security of smart home network.This essay has a thorough analysis of related theories, technologies and methods of smart home network and flow testing, and proposed an anomaly detection method in smart home network which is based on the thought of classification. This essay designed and realized based on classification distributed abnormal traffic detection prototype system for specific periodicity and aperiodicity network flows in the environment of smart home network. Specific research and development work mainly includes:(1) Has an analysis and research of the rules of traffic in smart home network, and the traffic in smart home network is classified from the following aspects:the application types, the flow features, the degrees of importance, and influence degrees of equipment quantity changes, etc. It helps to distinguish and extract control command stream, condition monitoring stream and media data stream from complicated smart home network flow and finally analyze the features of different flows.(2) Using differentiation anomaly detection method to different traffics at the combination of the specific agreement and application. The methods includes:using real-time detection by adaptive threshold detection to the state monitor flow which has a much more stable and no obvious periodic; using sequence similarity searching technology to have periodic anomaly detection for media data stream and control command stream which can happen suddenly in a short time and have a change in cycle in day or week; and using the method of threshold adjustment to judge the abnormal behavior in smart home networks according to the degree of the importance to of different traffics. The experiment has proved that the comprehensive use of the above two detections has improved the effect of anomaly detection in smart home network significantly(3) Designed and realized a distributed anomaly detection system which is based on the thought of flow classification detection and very suitable to smart home network environment. What's more, writer has validated the real-time and effectiveness of the classification of traffic and anomaly detection methods by experiment.