Intrusion Detection and Prevention

The succeeding paragraphs discuss some of the scientific works that are related to the heterogeneous networks that we have explored in our ongoing research process. Thus, the problems of machine-to-machine communication (M2M) are analyzed. ETSI, the European Telecommunications Standards Institute, defines the protocols to be considered when using IoT devices that lack computing power. The daily lives of people can potentially be influenced by machine-type communications, which allows us to consider heterogeneous networks as a reliable, technical option for the 5G IoT network implementation. Machine-type communication is associated with an ever-increasing volume of transmitted data. In addition, paper "Reinforcement Learning for Real-Time Optimization in NB-IoT Networks" described the corresponding quality of service (QoS) policies. The paper "Wireless Beam Modulation: An Energy- and Spectrum-Efficient Communication Technology for Future Massive IoT Systems" suggests that a considerable number of 5G IoT devices can be deployed, which will provide certain services through their interactions and also ensure a proper balancing of data traffic. The contribution that is reported in "Investigation of Future 5G-IoT Millimeter-Wave Network Performance at 38 GHz for Urban Microcell Outdoor Environment" describes a system that actively handles data traffic generated by registered NB-IoT devices using the same ML-based intrusion detection engine. It is important to develop and deploy large 5G network topologies, such as narrowband IoT (NB-IoT) and millimeter wave (mm wave). Scientists are working on the design of intrusion detection and prevention systems for 5G and beyond networks. The authors of "Space-Reserved Cooperative Caching in 5G Heterogeneous Networks for Industrial IoT" offer to design the federated IDS architecture by means of federated learning for 5G networks. The authors of "Enabling the IoT Machine Age With 5G: Machine-Type Multicast Services for Innovative Real-Time Applications" offer a novel IDS for 5G networks to efficiently identify the attacks. The main problem with the offered works is that their designed IDS cannot work efficiently in real-time.

The heterogeneous networks (HNet) concept implies an additional paradigm that supports the design and realization of 5G logical networks in which services are hosted. We used this system to design a runtime that allows the IDS to correctly handle all data flaws in a 5G data network. It must be mentioned that this additional mechanism provides an opportunity to configure the virtualized network settings properly. Some data types are given priority during processing over others by defining appropriate QoS policies. This approach is unique and differs from the approaches offered in the related works.

The main advantage of our system is that it works in real-time, which is very important for the security of 5G and beyond networks. The integrated intrusion detection system, which is presented in this paper, relates favorably to similar existing contributions. Thus, the authors describe intrusion detection systems that ensure a fast processing of the data traffic that flows through the 5G network core; however, the detection accuracy is not satisfactory. Moreover, papers present intrusion detection systems that generate a satisfactory level of detection accuracy, but they are not able to scale well for large real-world 5G data infrastructures. Furthermore, other papers propose relatively comprehensive surveys concerning significant intrusion detection approaches. Nevertheless, none of the presented models fulfill all technical performance criteria, at least when considering large, real-world deployments. It is relevant to note that the paper proposes a data processing model that is based on ensemble learning, while the paper discusses the security of optical data transmission mediums relative to 5G infrastructures. Essentially, the fundamental requirement that was envisaged is related to the mandatory automatic real-time processing of large amounts of data traffic that flow through the telecommunications operator's 5G network core. It is relevant to note that the proposed algorithmic and architectural structures fully comply with these constraints.

The performance assessment process, which is described, and the reviewed similar contributions suggest that the integrated intrusion detection system, which is presented in this paper, is one of the very few relevant systems that are proven to detect known and unknown threat patterns in a large 5G network core, with high accuracy and without interfering with the low-latency levels of the implied data network, as they are perceived by the end users.