Soft Computing Techniques for Physical Layer Security of IoT Devices

• Published in: Advanced Soft Computing Techniques in Data Science, IoT and Cloud Computing Vol. 89; pp. 27 - 51
• Main Authors: Ismayil Siyad, C., Tamilselvan, S.
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Studies in Big Data
• Subjects: Deep learning; Fuzzy logic; Genetic algorithm; IoT; Machine learning; PHY; Physical layer security; Soft computing
• ISBN: 9783030756567; 3030756564
• ISSN: 2197-6503; 2197-6511
• DOI: 10.1007/978-3-030-75657-4_2

Internet of Things (IoT) is an archetype of the Internet where the devices are connecting each other utilizing the Internet. Due to the enormous attention from industries and academia IoT is considering one of the technologies that revolutionize the future. IoT is considering as one of the potential enabler technologies for beyond fifth-generation wireless technologies. The physical layer aspects of IoT is one of the rapidly developing areas in IoT research. The developments in the arena of soft computing techniques such as fuzzy logic, genetic algorithms, machine learning, and deep learning algorithms have an unequivocal role in this buildout. Data-driven techniques revolutionized various aspects of physical layer techniques such as the generation of adaptive waveforms, energy harvesting, the energy efficiency of the network, spectrum sensing, multiple access techniques, cooperative communication, power allocation, and carrier allocation, etc. Physical Layer Security (PLS), where communication security is achieved by the techniques used in the physical layer is a major application area of soft computing techniques. Several PLS techniques for IoT based on soft computing techniques are proposed by various researchers. Soft computing techniques transformed many of this hardware define techniques to software-defined. In the literature, many researchers are recently reporting many advancements in this domain. One of the major advantages of this type of data-driven ad knowledge-driven technique is its inherent ability to adapt and cognitive capacity to behave differently with the time-varying characteristics of the physical medium. Another advantage of soft computing technique in the physical layer is its propensity to solve nonlinear problems which are difficult to solve with mathematical algorithmic models and its ability to approximate complex dynamic systems according to the renowned universal approximation theorem. Also, for some solvable multivariate optimization problems soft computing techniques give low complexity solutions by training the same algorithm generated data. Thus, creating a low complexity representation of nonlinear models. In this regard, this chapter presents a comprehensive overview of the state-of-the-art approaches towards the application of soft computing algorithms to physical layer security techniques for IoT network. Qualitative and quantitative insight into soft computing techniques for the IoT physical layer security is included in this chapter.