Abstract

The speed of the development of cloud computing and mobile networks has brought significant security risks that require innovative solutions to secure volatile and decentralized environments. The conventional models of perimeter-based security do not fit this scenario because they do not support the dynamics and multi-faceted nature of new threats to the network. This paper outlines a standalone security system coupled with the Zero Trust architecture and dynamic threat detection system to allow a greater level of security in cloud and mobile networks. The system suggested would guarantee constant verification of users, devices, and applications, with the use of AI-driven threat detection, whereby breaches to security are detected and addressed automatically. The proposed framework is experimentally evaluated and found to be very effective as compared to the traditional security models. The autonomous framework has a detection accuracy of 98.4% as opposed to the traditional models with a detection accuracy of 85.2%. In addition, the system was shown to have a much shorter response time (45 ms vs. 150 ms) and false positive rate (4.1% vs. 12.3%), as well as scalability because it could support up to 5000 users (compared to the 500 users that the traditional systems could support). This is indicative of the system in capturing large-scale dynamic networks. The inclusion of the Zero Trust principles further improves the security of the network since no user or device will be trusted by default, no matter where they are located. The autonomy of the system allows it to constantly adapt to new threats to enhance its capability to respond to changes in attack vectors. The paper identifies that the proposed approach is a scalable real-time solution to the current security dilemma facing cloud and mobile networks, which offers firm protection against internal and external threats. Further optimization of the system to deploy even bigger ranges will be considered in future work, and more rational methods of AI will be integrated.