A detailed analysis of Deep Learning for Internet of Things (IoT) Security
Security attacks are a showstopper for the Internet of Things (IoT), which has recently gained popularity. The heterogeneous structure of the Internet of Things (IoT) confronts it to a permanent uncertainty. The active nodes in IoT networks are resource-constrained, which makes them luring targets for security-attacks. One single attack can easily jeopardize its global performance. This paper covers different security threats faced in an IoT network and how Deep Learning techniques can be used to detect and address them.
IoT devices are typically connected over wireless networks where an intruder may access private information from a communication channel by eavesdropping. IoT devices cannot support complex security structures given their limited computation and power resources and the physical environment.
The existing security mechanism has to be suitably tweaked for seamless adaptation in IoT networks. For instance, applying existing defense mechanisms such as encryption, authentication, access control, network security and application security will be challenging. It would need suitable changes to address the scaling requirements of IoT networks.
This paper covers the following –
1. What are the potential threats is an IoT System?
2. What are the security requirements in an IoT System?
3. What are the various Deep Learning (DL) methods used in IoT Security?
4. What are the necessary steps in Deep Learning (DL) based Security Risk Detection?
5. How are security attacks detected using Deep Learning (DL) methods?
