Automated Secure Updates for IoTs

Security is one of the biggest challenges in the Internet of Things(IoT). The heterogeneity and resource constraint nature of IoT devices make it more vulnerable. This is the reason, IoT devices are considered as an attractive target for attackers.  Most often, an IoT device get corrupted due t

Project Title

Automated Secure Updates for IoTs

Project Area of Specialization

Internet of Things

Project Summary

Security is one of the biggest challenges in the Internet of Things(IoT). The heterogeneity and resource constraint nature of IoT devices make it more vulnerable. This is the reason, IoT devices are considered as an attractive target for attackers.  Most often, an IoT device get corrupted due to the malicious software or firmware updates. These updates are written intentionally in a way to steal the private information of users. Henceforth, it is of utmost importance to check for the validity of these software/firmware updates before sending it to the device.

By exploiting the layered architecture of IoT,  we aim to provide a deep learning based automotive framework to secure the IoT device updates from malicious attacks. These updates will be sent by the original equipment manufacturer of the device to the cloud/fog nodes. The cloud/fog checks for the confidentiality, integrity, and authentication of the updates and sent to the gateways. For further investigation, the deep learning models will be deployed on the gateways to detect if the received update is malicious or benign. If the update is benign only then it will be sent to the device, else discarded. Moreover, to train the  deep learning models, the data set is collected from virus share repository, kaggle, and vision research lab. 

Project Objectives

• To create an experimental set up.
• To collect the data set of malware samples from the known resources.
• To prepare the data set of benign samples.
• To implement an efficient deep learning model for detecting the malicious updates

Project Implementation Method

• Uploading of the update in zipped file from Original Equipment Manufacturer Server and transfer it through HTTPS protocol to Dart Dav application at the gateway.
•  After receiving at the gateway,  it will be transferred to a "Zipped Docs" folder and it will extract there.
•  The update file will be put into the deep learning model.
•  File will be placed in a benign folder if it went through all the checks correctly else will be placed in a malicious folder.
• Gateway will broadcast the updated file to the IoT devices & wait for acknowledgment.
• Gateway will store the update until all the IoT devices receive the update.

Benefits of the Project

It is an advanced Deep learning model framework which makes the devices capable of protecting itself by distinguishing among the files, which will be sent from the Original Equipment Manufacturer, and categorize them under a benign or a malicious update file tag.

This framework will minimize the need of deploying some heavy duty security algorithms within IoT devices.

Technical Details of Final Deliverable

1. Successful update transfer

2. Metadata checking

3. Log maintenance

4. Implementation of MQTT framework

5. Classification of malicious & benign

6. Successful authentication

7. Building ML/DL models for security

8. Testing framework

9. Start the research paper draft

10. SSL security

We have successfully implemented MQTT protocol in our framework. Our framework automatically  receives the updates from registered companies servers, which then authenticate the updates via Metadata checking mechanism. This mechanism mitigates the risk of replay and freeze attacks. Upon the results updates are either labeled as benign or malicious.In either case  broker keeps the Log of all the updates for later use. The client connects to the broker through providing  username and password to authenticate themselves. Only the authentic nodes connects to the broker. Update comes in .exe file and it will be converted into image file and will pass through the Deep Learning model and will be classified as malicious or benign. On benign it will be proceed further else the update will be discarded.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Security

Other Industries

IT

Core Technology

Internet of Things (IoT)

Other Technologies

Artificial Intelligence(AI)

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com