日本語 サイトマップ

Introduction

The IoT security with AI's protection
Today there are various personal devices have become more intelligent, and that is IoT society which gathers information and sends it to the cloud is spreading. Surveillance cameras, automotive electronics, household appliances, medical equipment, and factories and office robots will have Internet connectivity. However, many security risks are arising when devices originally assumed to be used individually are connected to the Internet. Unlike PCs and smartphones, these IoT devices operate with M2M (inter-device communication), and do not necessarily interact with humans, so monitoring by users is difficult task. Also, users do not only have to worry about Internet connection, but also PCs are exposed to physical attacks as they may be infiltrated from USB memory. For example, tampering embedded software from the debug port and giving erroneous information to the sensor. In addition, PC is kept in house, but car is left in parking lot. How will we deal with these new threats? The threat is arising from being connected to the network is actually experienced by the PC, the smartphone, and the control system. Besides that, there are still another potential security theme, such as detection of unauthorized access, malware detection, log analysis, virtual machine monitoring, but various methods are being considered. We believe that the threat of IoT may be in difficulties such as physical attacks and password-based authentication. Both, since IoT is M2M, the essence is that people are not involved. No matter how much the user is watching, incidents will occur where they can not be seen. What happen if human is not involved in some process? People should consider to use the AI technology to protect the IoT's security on behalf of people.

IoT

Research theme

  • Discovery of fraud by analyzing Web access log by RNN (Recurrent Neural Network)
  • Vulnerability assessment of IoT devices
  • Privacy protection on Wireless Sensor Network
  • Security of automotive embedded devices
  • Safety evaluation of counterfeit prevention PUF for semiconductors

Toshihiro Matsui

Notice

2018/03/08 Research contents page

2017/06/06 Members page

2017/05/16 The website has been released.