WIRE AND WIRELESS ACCESS POINT FOR ANALYZING ABNORMAL ACTION BASED ON MACHINE LEARNING AND METHOD THEREOF

摘要

The present invention relates to a wired/wireless router for analyzing an abnormal action on the basis of machine learning and a method thereof, capable of effectively blocking an attack on a wired/wireless router by repeatedly learning a normal action type and an abnormal action type using a wireless signal and state information transmitted/received through a wired/wireless router, and by detecting and analyzing a state and a wireless signal of the wired/wireless router based on the learning result. According to the present invention, a wired/wireless router for analyzing abnormal action on the basis of machine learning is a wired/wireless router which repeatedly learns a normal action type and an abnormal action type of a wired/wireless router and determines whether or not the wired/wireless router is abnormal on the basis thereof. The wired/wireless router of the present invention comprises: an information collecting unit which periodically collects a wireless signal transmitted and received by the wired and wireless router and state information of the wired and wireless router; an information analyzing unit which analyzes the collected wireless signal and state information and determines whether an operation of the wired and wireless router is abnormal; and a security processing unit which performs a security policy when the operation of the wired and wireless router is abnormal. The information analyzing unit may analyze the collected wireless signal and status information, analyze whether or not there is an abnormal action, and operate in conjunction with a machine learning engine that repeatedly learns a normal action type and an abnormal action type.