MATHEMATICAL ANALYSIS OF ADAPTIVE CRYPTOGRAPHIC COMMUNICATION & CAT-BOOST BASED INTRUSION DETECTION IN IOV

Abstract

The advanced computing technologies such as cloud computing, IoT, IoV, and machine learning create vast amounts of information that require efficient, reliable, and prompt processing. The use of the IoV technology becomes essential since it allows establishing communications between the vehicle and roadside infrastructure.

Nonetheless, IoV communication networks are highly exposed to security vulnerabilities due to their decentralized architecture. Such threats as malicious vehicles, data injection attacks, and others can significantly influence the operation of the network and its users. To tackle the problem, the present research paper offers an architecture for secure communication based on trust management and hybrid encryption. Considering the idea of the hybrid encryption algorithm, both of the algorithms should be taken into account, as both of them are efficient in their specific areas. For instance, since the RSA algorithm may be employed for user authentication and key exchange, the efficiency of the AES algorithm will be helpful in encrypting the data quickly. Moreover, the Cat-Boost classifier should be utilized in our study, in order to create a novel IDS that will detect all malicious behaviour exhibited by vehicles and traffic problems according to the UNSW-NB15 dataset. We found out that the efficiency of our model equals 93.57

Some of the accomplishments which have been recorded using this suggested method have included the application of machine learning algorithms to analyse the data sets that include the UNSW-NB15 data set, which were analysed using the Cat-Boost technique, and which yielded an accuracy of 93.57

On the other hand, IoV -based networks suffer from many security issues due to their open and decentralized structure. There is no central entity managing the communications in this network. Hence, malicious vehicles can inject false data into the system, affect traffic behaviour, and even pose threats to the system as a whole by pretending to be some of the legitimate nodes. Thus, data protection and authentication have become crucial concerns that need to be addressed properly.