Early warning model of cyber threats in 5G networks using Markov processes

Security of telecommunication networks, in which the transmission channel can be used by many users simultaneously, is a particularly important problem. In wireless metropolitan networks, this problem is compounded by the fact that the communication channel is publicly available. In other words, information transmitted in such networks can be easily intercepted by intruders. This can lead to theft of personal data, financial losses, or even to a breach of security of critical infrastructure. Information security can be compromised by failures that affect the availability, integrity, or confidentiality of information. These failures can be caused by vulnerabilities, namely, defects in software or hardware that can be exploited by attackers to gain unauthorized access to information. Information security is one of the components of 5G networks reliability. The main security threat to such systems is vulnerabilities, primarily of software components. Despite the fact that information about vulnerabilities of software products is publicly available, there is not enough data to quantify the security of these products using a single general criterion. It is also impossible to predict how well they will be protected from attacks in the future. One of the main problems of choosing the most secure 5G configuration is the difficulty in quantifying the level of information security. In addition, it is difficult to choose adequate evaluation indicators that take into account all the factors affecting successful network penetration and the amount of potential damage. The search for vulnerabilities in software components is an urgent and resource-intensive task that has recently been taken up by large companies and research centers. Analysis of vulnerability detection and remediation processes shows that they can be described by a mass service system with an infinite queue length. A model for early warning of cyber threats in 5G networks using Markov processes has been developed. Using simulation modeling in the Matlab environment, a time diagram of the arrival of requests for vulnerability detection was obtained. The change in the probabilities of states was also obtained. Thus, knowing the intensity of the flows, it is possible to model and predict the processes of the arrival of requests for vulnerability detection in real time.