ANALYSIS AND DEVELOPMENT OF ENERGY EFFICIENT TECHNIQUES FOR 5G GREEN COGNITIVE RADIO NETWORK

Abstract

WITH the growing cognizance of environmental concerns and global warm ing related to communication technologies, researchers have been seeking some solutions to minimize the consumption of energy in the telecommunication indus try. There is a remarkable advancement in mobile communication from simple voice based devices to ubiquitous data-hungry smart phones. The telecommunication indus try expressions a serious energy consumption challenge. The existing static spectrum allocation-based technologies are not in a position to fulfill this extra spectrum require ment and handle this future traffic load. This volatile evolution of global traffic data urges research attention globally and can be handled by future cognitive radio networks (CRNs). This directed to the development of the idea of inclusion of CR technology with green networking. The Green Cognitive Radio Networks (GCRNs) are able to re move this limitation related to spectrum scarcity. The application of CR technology will be utilized to make the green (radiation free) environment in order to increase the spec trum resource opportunities available for next generation (6G) networking. We analyse that the energy- efficient green communication and seamless networking are very im portant pillars of smart city construction, and connect the different essential elements of smart cities. Emerging technologies such as green communication, artificial intelli gence, cognitive technology, Internet of Things, machine learning and cloud computing are now being used in a significant manner to convert cities into ”smart cities”. As a consequence, the main objective of the thesis is to investigate the schemes to allocate the resources/power efficiently in cognitive radio technology-enabled green networks to support intelligent telecommunication systems. To start with, this thesis provides an introduction, subsequently by a overview of CRNs, spectrum management, energy efficiency measurement and power allocation. A detailed review of the current literature on the concerned has been presented. iii The first objective of the thesis is to investigate various next generation green wire less communication networking techniques, with consideration of energy-efficient trans mission. The futuristic technologies like cognitive radio, carrier aggregation, Terahertz communication, Internet of Things (IoT), massive MIMO (multiple-input multiple-output) and mm wavelength are briefly reviewed to prepare for advancing recent research con tributions. It is followed by a discussion on the green CRN architecture and cognitive cycle. Further, the challenges related to green CRN and spectrum management are also reviewed. The second objective examines two proposed channel selection strategies: probability based and sensing-based channel selection strategies. The proposed channel selection methods evenly allocate the CU’s traffic load among various applicant channels. Re sults of the work present that in the circumstances of huge traffic, SCSS reduces the total network time, while in the situation of low traffic, PCSS gives better results. These observations offer a vital perception in designing of traffic-adaptive channel selection strategy in the existence of PU’s interruptions and sensing errors. The proposed strate gies can minimize the total network time by 60% as compared to non-load balancing strategy for λcu = 0.05. Next, we calculate the total energy consumption at various op erational modes in GCRN. The results indicate that the arrival rate of the CUs and the time spent on channel scanning affect the energy consumption of the network. The pro posed channel selection strategies reduce energy consumption by 75% as compared to non-load balancing strategy. The third objective analyzes the benefits of cooperation between SUs for detecting the PU’s spectrum, through which the rapidity of the network can be improved. Two cases (having a distinct level of cooperation) have been exploited to reduce the sensing time. The first one is non-cooperative, in which all SUs independently sense the PU, and the first user who senses first, informs the presence of the PU to the other SUs via a central controller. The second is cooperative, in which SUs follow the protocols of Amplify-and-Forward cooperation to minimize the sensing time. The results show that the proposed joint cooperation spectrum sensing (JCSS) scheme increases the sensing probability for a vacant spectrum by as much as 34%. After this, we propose two distinct spectrum sensing schemes preset spectrum sensing (PSS) and viscous spectrum sensing (VSS) that presents the energy savings percentage in GCRNs under specific conditions. iv These results conclude that the energy consumed by the user’s contention increases due to the increase in sensing time. The proposed schemes are better in terms of scalability because it is not essential to sense all spectrums in these schemes. The fourth objective has analysed a cooperation-based energy-efficient scheme for cognitive users in GCRN to improve the energy efficiency of CU. The proposed cooperation based energy-aware reward (CEAR) scheme supports CUs to actively cooperate by uti lizing temporal and antenna diversity to improve energy efficiency. The proposed CEAR scheme is compared with other existing schemes, and it is presented that the CEAR scheme provides up to 28% improvement in energy efficiency. In this work, the optimal stopping protocol is used for problem formulation, and the backward induction method is employed for solving the decision problem. This chapter has contributed significant insight in terms of energy efficiency, spectral efficiency, throughput, and consumed en ergy, which motivates the design of future green communications systems. In the final objective,a real-time learning-based scheme has been proposed to con trol transmission power and decrease the overall network power consumption while supporting QoS for multilayers. The reinforcement learning method takes into account the influence of cognitive transmitters’ actions on the transmission power policy that has been chosen. In addition to this, the proposed ROPC scheme is based on the upgra dation method for the Q-value. This feature of scheme helps to decrease the state/action pair and improves convergence speed. The suggested scheme’s performance is proved by simulation, which shows that it achieves faster convergence and higher EE, SNIR, and SE than existing schemes. In the end, the thesis briefs the research objective findings and come up with the proposal for the future aspect of the research work.