Malware Detector using Decentralized System
A decentralized firewall system on a blockchain to determine whether or not a particular Portable executable file is malicious or not.
Created on 12th August 2018
•
Malware Detector using Decentralized System
A decentralized firewall system on a blockchain to determine whether or not a particular Portable executable file is malicious or not.
The problem Malware Detector using Decentralized System solves
The problem Malware Detector using Decentralized System solvesTrying to include neural networks to train our system on train data and test it on our server nodes. It prevents the malicious content from crossing the firewall and here, we plan to build a firewall which runs on a blockchain, which also help us gather insights and links between the machine learning algorithms and blockchain. We tried to think differently and convert a serious problem into a good machine learning challenge, by converting our PE files(Portable execuatble) into a image array and reduce its size by Python imaging library. We trained the data based on train data which the ML algorithm will draw insights and decide weights for our neural network which in turns helps in classification of the file as malicious or not. But we will not block the content but will bring out a query for the user to decide, and also it works on proof of work based consensus.
Challenges we ran into
Being very new to the area of blockchain, constructing a firewall using blockchain on ethereum was kind of out of reach, understanding the different aspects of theoretical concepts on the backend was equally important which we weren't exactly able to portray. Deploying and testing on truffle was an important part but we weren't able to do so due to the lack of skills but we did come up with a solidity code to build a structure for our blockchain.