To mitigate primary user emulation attack trajectory using cognitive single carrier frequency division multiple access approaches: Towards next generation green IoT

The growing cognizance of spectrum scarcity has become a more significant concern in wireless radio communications. Due to the exponential growth of data transmission in intelligent wireless sensor networks, energy spectrum detection has become a promising solution for resolving spectrum shortages. Primary user emulation attack (PUEA) has been identified as a significant attack vector in the cognitive radio (CR) domain's physical layer. In comparison, the CR is a promising method to increase spectrum efficiency by allowing unlicensed secondary users (SUs) to access licensed frequency bands without interfering with approved primary users (PUs). The study's primary findings are the methodology for preventing PUEA using authentication tags, which are unique sequences. This research blends SC‐FDMA with CR to protect CR networks from PUEA attacks, a Latin square (LS) matrix tag generation system is proposed to mitigate the PUEA effect. The technology is meant to provide effective authentication and protection against malicious users. In a secured environment, the LS tag technique is utilized to track and estimate the PU. For example, the BER of both techniques is virtually identical between 0 and 4 dB, while the BER performance of the suggested LS tag generation improves with increasing signal‐to‐noise ratio (SNR). As a result, the suggested LS tag generation is less susceptible to PUEA. To diminish the influence of PUEA in CR networks, an efficient enlightening approach for making the future Green Cognitive Radio Wireless networks structure is proposed. The simulation results also demonstrate the resilience of the proposed CR spectrum sensing techniques for energy‐efficient knowledge at varying degrees to reduce the adverse effects of environmental technologies.

son, the CR is a promising method to increase spectrum efficiency by allowing unlicensed secondary users (SUs) to access licensed frequency bands without interfering with approved primary users (PUs).The study's primary findings are the methodology for preventing PUEA using authentication tags, which are unique sequences.This research blends SC-FDMA with CR to protect CR networks from PUEA attacks, a Latin square (LS) matrix tag generation system is proposed to mitigate the PUEA effect.The technology is meant to provide effective authentication and protection against malicious users.In a secured environment, the LS tag technique is utilized to track and estimate the PU.For example, the BER of both techniques is virtually identical between 0 and 4 dB, while the BER performance of the suggested LS tag generation improves with increasing signal-to-noise ratio (SNR).As a result, the suggested LS tag generation is less susceptible to PUEA.To diminish the influence of PUEA in CR networks, an efficient enlightening approach for making the future Green Cognitive Radio Wireless networks structure is proposed.The simulation results also demonstrate the resilience of the proposed CR spectrum sensing techniques for energy-efficient knowledge at varying degrees to reduce the adverse effects of environmental technologies.

INTRODUCTION
Cognitive radio (CR) is a most promising, emerged, and key enabling technology of future networks technique that has been extensively investigated over the last two decades to solve the rising shortage of spectrum resources. 1,2SUs in a CR network locate the radio spectrum environment using spectrum sensing and utilize available spectrum opportunistically.4][5] The PUEA 6 is a type of malicious attacker 7,8 that modifies the radio spectrum environment by transmitting fake PU signals to interfere with the authentic SUs spectrum sensing process.It is important to note that the PUEA does not do any damage to the PU system, which ensures its secrecy and energy efficiency.In the case of a CR network suffering from PEA, if an SU detects the vacant channel being targeted during the spectrum access stage, it will falsely believe that a PU is utilizing the channel and will remain silent to avoid interfering with the PU.The attack is effective in this scenario be-because it stops the SU from exploiting vacant channels and provides the SUs with false information.In the alternative, if no SU attempts to perceive the attacked channel or if a PU occupies the channel, the attack has no effect on the SUs.Obviously, a necessary but not sufficient prerequisite for an effective attack is that it should concentrate on spectrum holes, that is, channels that are not utilized by the PUs.However, the reality is that the PEA has no prior knowledge of the radio spectrum environment and is hence likely to snip occupied channels rather than vacant ones.As a result, the PEA's assaulting strategy, which involves dynamically deciding which channel to attack during each time slot, has a substantial impact on the attacking performance and efficiency.In another light, a thorough understanding of the attacking method can also assist the PUs/SUs in quantifying the impact of PEA attacks on the SUs performance, which is critical for evaluating the accompanying detection and defense measures. 9,10It will help alleviate spectrum scarcity.As a result of this dynamic spectrum access, spectrum-sharing systems are vulnerable to malicious attacks.For starters, non-licensed users are especially vulnerable to harmful assaults due to a lack of spectrum ownership.So, protecting their spectrum resource access from competitors is difficult.Two factors complicate the implementation of suitable security countermeasures: fluctuating spectrum availability and scattered network structures.
Finally, attackers might use modern technologies like Machine Learning to launch increasingly complex and unpredictable attacks. 10Therefore, securing the SS network is clearly a major concern.Authentication, non-repudiation, compliance, access control, and privacy are all mentioned in Reference 11.SS requires confidentiality, especially when using databases.Transparency ensures that user-to-user data is not tampered with in any way.Availability ensures that users may access the spectrum/database when needed.User identification and verification are required for authentication.Denial of receipt/sending of a message or access to the spectrum at a specific location and time is called nonrepudiation.Detecting non-compliant conduct that causes harmful interference is part of compliance.User access to the spectrum/database is restricted via access control.][14][15][16][17][18][19][20] Cellular wireless networks must improve their capacities as demand for mobile broadband networks grows.One of the most difficult problems in cognitive radio networks (CRN) is the identification of PUEA.Through the vicinity of PUs, the attackers imitate PUs signal features to deceive legitimate SUs who are evacuating the network for themselves to exploit the network. 21The proliferation of IoT technology has aided in the rapid growth of Internet-connected gadgets.The exponential proliferation of these smart connected devices is revolutionizing the social lives of people. 22The fifth generation and beyond-5G connectivity is a step forward has addressed.The IoTs are powered by 5G which is reshaping the future of smart connectivity in society.The quality of service for various IoT services is assessed using criteria such as latency, dependability, power consumption, and throughput.The performance of connected devices is also determined by their computer capability, memory, and energy efficiency. 23The use of CRN techniques in future wireless networks could aid in the realization of the tactile Internet, a long-term information-centric network, and intelligent next-generation networks that have been envisioned. 24he CR on the other hand necessitates a large amount of energy to conduct its functions.However, due to the slower pace of advancement in battery technologies compared to semiconductor technologies, present battery technologies are unable to fulfill the higher power requirements associated with the flow of data created by Consumer Internet of Things (CIoT) devices.As a result, the effective use of CIoT networks to construct a flash flood early warning system is critical to saving lives and reducing the destructive effects of natural disasters.As a result, flash flood monitoring necessitates the deployment of a CIoT network across a vast geographical area without the need for regular battery replacement.Furthermore, in network architecture, optimal battery energy utilization for CIoT devices is critical.It has the potential to bring cost-effectiveness, increased network longevity, and fewer environmental problems. 25,26For 5G wireless systems and beyond 5G, the spectrum is an essential resource to meet the sustained rapid increase in bandwidth demand.To familiarize with the vital facts for green CRN research a Dynamic Spectrum Management (DSM) system with spectrum sharing among distinct types of users is required 27 for improving communication capacity in the design of next-generation green wireless networks.Such approaches will enhance the energy efficiency of the developed IoT devices with convincing glitches in everyday life which eventually make the environment safe and healthy.Such remarkable developments are the foundation of green CRNs.Detecting spectrum availability is one of the most important difficulties in CR communications. 28In cognitive communication networks, signal classification is critical for detecting and avoiding interference. 29,30he primary channel metrics have received a great deal of attention because of their importance in improving the performance of DSA/CR systems. 31Spectrum sensing is needed in CR systems to protect PUs from hazardous interference. 32,33n the literature, the SSDF attack is discussed.This attack is due to malicious users (MUs) tampering with locally collected sensory data which are sent to the FC to mislead it into making a bad judgment, causing the chirp spread spectrum (CSS) system to operate poorly.
Many approaches have been devised to detect MUs and reduce the harm caused by their attacks in CSS.However, existing anti-MU approaches are limited.Energy detection is among the most often utilized spectrum sensing techniques since it does not require any prior knowledge of the PU signal attributes.The adaptive learning-based mechanism in the CRN 34 analyses the transmitters' power features to detect and protect against PUE attacks.Attackers were distinguished from low-power PUs using cyclostationary feature detection (CFD) analysis.As a result, the proposed improved classification accuracy and communication rate.In Reference 35, the authors studied a channel-based technique based on multi-path channel behavior to detect PUE attacks using Machine Learning.It was given training using four extracted feature vectors.This approach could efficiently distinguish between legitimate and malicious users.PUE and jammer assaults were examined in the CRN by Roy et al. 36 A compressed received signal-dependent detection algorithm with sparse coding was proposed.It used channel-dependent dictionaries to distinguish between the spectrum hole, genuine PU, and emulation or jammers using convergence patterns.The decision-making process used Machine Learning based classification and the algorithm's efficiency was evaluated in Smart cities, remote monitoring, and eHealth applications require a new strategy to improve cloud-based dependability and availability.Since the scope and diversity of the cloud environment is overwhelming, most internet services, including software and hardware, have failed, a way of authenticating end-to-end data transmission among secondary users needs a large amount of power, which leads to harmful interference to primary users.The motivation of this research is to address the challenges and limitations in the current cloud environment and internet services related to authenticating end-to-end data transmission among secondary users.Here the main contribution highlights the harmful interference to primary users caused by a large amount of power needed for authentication, and the need for large bandwidth efficiency to overcome energy consumption.In such a way, the research presents advanced green IoTs as a potential solution.It emphasizes the importance of developing the next-generation internet of medical things (IoMT) to advance human health management by designing and developing an advanced bio-analytical system with network-linked devices.To overcome such massive energy consumption, large bandwidth efficiency is always a pivotal concern therefore, presented research is presenting advanced state-of-art green IoTs are focusing.In this direction, the outcomes of this research will support the development of the next-generation internet of medical things (IoMT) to design and develop an advanced bio-analytical system that combines network-linked devices, needed advancing human health management. 37

LITERATURE SURVEY
Sharifi et al. 10 suggested a unique CSS with attack awareness in the PUEA where the possibilities of malevolent existence as well as the absence of the licensed PU signal were assessed.Then, the attained parameters were employed to discover the optimum threshold that diminished the total error probability.Simulation results were offered to indicate the enhanced performance of the recommended method against PUEA, and the results were contrasted to that of the conventional method.To prevent SUs from communicating in CR networks, a malicious attacker emits signals that mimic the characteristics of PUs.The PUE attack is one example of this type of attack.Yu et al. 11 provided a full introduction to PUEA, from the basis of the attack and how it affects CR networks to a strategy for attack detection and defense.
A two-level database-aided detection approach was suggested to protect CR networks from PUEA.For the purposes of quick and robust detection ED location verification were combined.To arrest the performance deterioration of a CR network due to PUEA an admission-control-based defense method was recommended.AlahMadi et al. 12 suggested using an efficient advanced encryption standard (AES)-centered digital TV scheme; The reference sequence to generate the P2 pilot symbols in the DVB-T2 frames was encrypted using the AES algorithm to allow PU and malicious user detection.Accurate PUEA detection was then possible over the sub-carriers or sub-bands where the P2 symbols were present.Through energy harvesting, an effective communication scheme was also implemented for SUs under PUEA and the sum rate of the SU network could be substantial.The worst case of PUEA interference was evaluated in terms of minimizing the sum rate for the SUs.Here the Simulation outcomes come with the worst jamming scenario if the malicious user performs equal power allocation overall white space subcarriers Haghigh et al. 13 introduced a scenario of a smart PUEA that directs fake signals resembling the primary signal.A smart attacker distinguished from an ever-present attacker was assumed to be aware of its CR environment which might switch transmission strategies.Cooperative spectrum (CSS) principles were then derived for a CR network in this case, and a new spectrum sensing scheme based on ED was proposed to mitigate the adverse effects of PUEA.Chen et al. 14 also derived CSS principles for the CR network in the presence of a PUEA and proposed a new CSS system based on ED.The data fusion center (FC), which makes a global decision about the presence or absence of the primary signal, employs the logical AND/OR rule.In contrast to conventional ED spectrum sensing, simulations have shown that the destructive effect of PUEA in spectrum sensing can be mitigated.Additionally, Li et al. 15 proposed a new penalty mechanism based on cognitive trust value.The interactive, authentication, trust value collection, configuration, storage and up-date, and punishment.A hierarchical structure had been introduced formed by the FC and cluster heads (CHs) to oversee the trust values of cognitive users.Malevolent users are penalized by the FC for diminishing their trust value.Ahmadfard and Jamshidi 16 studied a channel hopping-based defense strategy to avoid (NEP) of the interactions among the multiple cognitive users, and the attacker was obtained using an algorithm at which no throughput increase is possible by the cognitive user deviating from it.Most of the above techniques, however, do not consider the CR network under uncertain conditions. 17he spectral sensing focused on fractal dimension methods to protect against PUEA.It provides a set of assessment guidelines to classify various digital signal modulation types.Nevertheless, its rate of recognition is very less. 18In the study of Reference 19, a Kalman filter framework-based solution for identifying primary user emulation as-saults with a non-stationary primary user was developed.Even though the limitation of this work is locating the primary user's initial position and the difficulty of managing the measurement uncertainty.A hybrid location method for PU emulsion detection in CRNs was approached in Reference 20.An AoA, uses spatial dimensions to identify the PUEA, and the RSS utilizes merely range to determine the PUEA.It is reported that the range-based class is more accurate but more complex.A trust system for the identification of malicious nodes in CRNs has been proposed in Reference 21.From a security perspective, a legitimate handoff can be imitated by an attacker or malicious user with the purpose of degrading network performance.Such findings illustrate the need for a strategy to signal propagation environments to reduce the probability of false alarms and probability of miss detection.This led to the technical creation of CR that maximizes the use of spectrum.Moreover, spectrum sensing provides the basis for CR which is among the most critical techniques that allow the CR to maximize the use of spectrum.For instance, India has already begun its 4G bidding process, nevertheless, mobile operators are struggling to meet the humongous rise in demand for spectrum.An additional 4G spectrum would enable operators to expand network coverage and broaden the scope of their systems to underprivileged remote parts of the country.Increasing demand to migrate online is on the rise and the pandemic has indeed affected the number of mobile networks for people and companies.However, mostly the researchers examined the fact that far more than 90% of the bandwidth is underused.
Meanwhile, the spectrum shortage is known to be a bottleneck in wireless connectivity as techno communications.This CR is a promising approach for managing bandwidth efficiently.3][24][25][26] In this paper, one such attack is called the PUEA which deteriorates the efficiency of CRNs dramatically.The existing fixed bandwidth allocation policy only grants the contractual user, also known as the licensed, incumbent, or primary user, the frequency usage. 38Over the last couple of decades, spectrum recycling through CR technology has emerged as a method of resolving spectrum restrictions. 39,40DoS attacks can be used to exploit vulnerabilities in CRNs and cause significant performance loss.The CRs spend a significant amount of time during operation identifying idle (free) channels for transmission.2][43][44][45] As heterogeneous communication networks and multimedia services are rapidly expanding. 46,47In the virtual environment, a CDN is trained to learn the best spectrum access strategy.9][50] However, the data-based strategy is more significant for learning the process through historical data monitoring in semiconductor industries to meet the higher power requirements related to the flow of data generated using different IoT and sensor-related devices. 51Furthermore, the knowledge-based method frequently necessitates trained competence and domain knowledge to generate a suitable judgment based on the flawed data correlation structure.The semiconductor industry has yearned for the ability to identify the root causes of faults in real time to reduce time and increase yield rate.However, because of the lengthy and complex processes, data is generated in excess.The future of wireless communication networks will be defined by the capacity to effectively employ resources to fulfill the demand for expanding diversity in services and user behavior. 52,53In Reference 54, in DSS contexts, this study introduces a new energy-efficient spectrum sensing method for MIMO-OFDM systems.
In Reference 55, the goal of this research is to reduce energy consumption and evaluate network lifetime in a wireless cognitive sensor network made up of randomly dispersed sensors that perform MCCSS.D2D interactions have long been regarded as a critical enabler for reducing traffic congestion on 5G mobile networks.Cognitive spectrum sensing can be used to detect temporarily available spectrum chunks for direct linkages among user devices to optimize radio spectrum usage in such a communications scheme.In Reference 56, Spectrum sensing in CR protects licensed users from hazardous interference and maximizes spectrum utilization.As a result, this research provides a technique for optimal channel estimation and spectrum sensing for the MAC layer protocol in CR networks, thereby addressing the scheduling concerns.In Reference 57, the authors presented a swarm-based performance-tuning co-operative spectrum access method for a wireless cognitive sensor network is detailed.In Reference 58, this research presents a new SCSTN, which combines the cognitive satellite domestic system and a decentralized CSS infrastructure.In Reference 59, the authors investigate a scheme to reduce SU energy consumption while optimizing overall spectrum utilization and maintaining accurate sensing by maximizing sensing time detection while keeping SU power and data rate low.In Reference 60, the member nodes perform roles of the CSS system and transmit the localized testing statistics to their CH, which is located nearby.The CH then aggregates the sensing data and uses the likelihood ratio test to arrive at a cluster choice by utilizing the spatial correlation of the members.The fusion center uses a hard fusion approach to make the final choice about spectrum occupancy after receiving decisions from all clusters.Our technique not only improves sensing performance but also improves energy efficiency, according to simulation results.But the CSS system information is needed about prior PU.In Reference 61, the research provides a learning-based identification framework for heterogeneous signals modulated with OFDM in a virtual environment at an undetermined frequency.
The energy detector and the sphericity test are two frequently used spectrum sensing techniques 62 that take advantage of the varied features of the signal received at the SU terminal.In Reference 63, due to the general intrinsic openness of satellite communication and the increasing intelligence of devices, it is subject to spectrum misuse, posing a severe danger to the satellite communication system's dependability and efficiency.The identification of spectrum misuse in the satellite-terrestrial spec-rum sharing system is the subject of this study.The STFT is used in this research to introduce a TFSS approach for vehicle detection using a single acoustic sensor.Vehicle identification, tracking, and classification are among the applications of spectrum sensing in the time-frequency plane.31,64 The challenge of spectrum sensing for NC signals is managed in CR networks with uncelebrated multiple antennas.The NCC approach is a novel robust spectrum sensing technique that fully uses the statistical quality of the NC signals by including both the standard covariance and complementary covariance information included in the NC signal. 65The MUs can submit fraudulent detecting data to a centralized CSS system, which can drastically decrease the CSS system's performance. 32,33,66This article discusses a signal processing approach for receiving non-contiguous RF channels by mixing different RF bands into neighboring IF channels simultaneously. 67The CFO is caused by the offset of fundamental clocks of transceivers or non-ideal components in practical CR systems, which reduces spectrum sensing accuracy and degrades reliability. 30,68Cooperative sensing is being investigated as a means of enhancing system performance.The optimal threshold value for the generalized FFT-based 1-bit quantization system is determined in such a way that the aggregate error rate is minimized while still allowing for decision fusion collaboration. 69 novel opportunistic spectrum-sharing system for CR networks is proposed in this research, which is based on OFDM-IM.The PT transmits OFDM-IM signals to the primary receiver via an AF relay in the proposed OFDM-IM-based CR architecture 70 considers a CR network in which the feedback information from the primary network is used to build an access method for the secondary network to make use of the underutilized primary spectrum resources. 71Due to their low use radar frequency bands have recently been suggested as feasible options for CR. 72Identification of a PU using spectrum sensing is critical for a SU in CRNs so that interference-free transmission can be achieved. 73An effective adaptive detection approach for CRNs is given in this research, in which several SUs collaborate to locate idle spectrum bands. 74 revolutionary technology-independent machine learning framework capable of independently locating and retrieving statistics from unlicensed frequency transmissions.The machine learning application for reducing false alarm and misdetection rates in conventional energy detection. 9,75After extracting features in CSS mode, the classifier can use soft or hard combining schemes to make decisions.The kind is decided by the feature vectors.Hard combining SUs digitize their local observations, while soft combining SUs explicitly exchange their local decisions.The most frequent hard combining CSS approaches use and/or.To be occupied in And-based CSS, the PU must be active for all SUs.A PU is active if at least one of its SUs determines that the channel is busy. 76here are three types of ML-based CSS schemes: unsupervised, supervised, and reinforcement-based. 77 Unsupervised-based CSS feeds the classifier features without labeling them 78,79 utilizing the k-means technique and its revised version, fuzzy c-mean, as ML classifiers with energy detection (ED) feature extraction.The Normalized energy determines the channel state in the ED-based method 80 presented a technique based on geodesic distance calculation.In supervised-based, the classifier fills in the features and their labels to make the final decision.Numerous studies have been done on SVM. 39,81While SS networks and ML have many similarities, combining the two is a natural fit.For all the above frameworks, users/coordinators in the SS network must monitor spectrum resource utilization and make decisions based on observation, learning, and reasoning. 82,83Various PUE attack detection and defense tactics have been discussed above.But no one has discussed the optimal offensive techniques.5][86] Assailants can use ML algorithms to increase their performance, which can help defense plan design.They showed two GAN-based models that successfully replicated PUs in References 34-36.They offered dumb and smart generator models based on prior knowledge of the PU's feature space.Two DNN-based discriminator models were created to identify the PU and EPU from their generator counterparts.Each GAN model's generator and discriminator got smarter with repeated and sequential training.During the deployment phase, discriminators could detect approximately 50% of PUE attackers without GAN training, while both GAN models could attain 100% accuracy during the learning phase.After GAN training, the dumb generator's discriminators achieved 98 percent accuracy, whereas the smart generator's discriminators achieved 99.5% precision.Their studies show that if there is no observation in the attacking position, the attacker loses.However, seeing at least one channel can dramatically boost attack performance.En outre, observing many channels does not increase the attacker's advantage, however, it does reveal the smallest constant factor. 87here is a game model 88 created, which combined anti-jamming and jamming subgames.Q-learning found the best channel access approach where Minimax-Q learning outperformed Nash-Q learning in an aggressive setting.Friend-or-foe Q-learning was the best explanation for scattered portable ad hoc networking where centralized control was scarce.An eavesdropper can target the principal system, and genuine SUs can operate as friendly jamming to prevent data spillage. 89In Reference 90, SU aided anti-eavesdropping communication for primary system defense.The PUs were the game leaders, and the SUs were the followers in a Stackelberg game-based cooperative jamming strategy.The SUs transmits the jamming impulses to increase the vested power's data rate.The system required to make more spectrum available to SUs to encourage them to provide friendly jamming.To access spectrum holes, the PU traffic load had to be modified.The system's evolution was modeled using a continuous-time Markov chain.The suggested system investigated optimal PU and SU techniques as Stackelberg equilibrium.CRN has a data-driven self-awareness module to guard against harmful assaults and establish safe networks. 91Smart attackers might use radio frequency to educate the CR on unpleasant habits and make bad decisions.Simultaneously, a basic SA device was discovered to help include innovative generative models and detect abnormal radio spectrum activity where two actual real-world methods based on data dimensionality and sample rates were shown.
The C-GAN was used to contract the high dimensions generalization state vectors extracted from the spectrum illustration data.Two methods learned low-dimensional generalized state vectors with sub-band information.The proposed methodologies for aberrant signal detection were evaluated on both real mm-Wave and simulated OFDM data sets.SS approaches will help alleviate spectrum scarcity.However, malicious attackers can exploit SS systems due to the dynamic connectivity of spectrum resources.First, non-licensed users are especially vulnerable to harmful assaults since they do not control the spectrum.So, opportunistic spectrum access is difficult to defend against.Second, the dynamic spectrum access and fully distributed frameworks make security countermeasures difficult to implement.Security concerns are major concerns in the SS system.Diverse network threats can hinder the SS system from satisfying the above requirements.This effort will focus on dangers and mitigation measures in the SS network's physical layer. 92We investigate two types of spectrum sensing attacks in the SS network, PUE and SSDF, that try to disrupt spectrum monitoring and user access.ML has become a vital aspect of security and privacy measures in numerous applications to further improve SS system security.ML can discern between normal and abnormal behaviors depending on how components in the SS system interact during spectrum access.The system may gather and analyze each component's behavior to discover regular patterns of interaction, allowing it to detect malicious behavior early on.
By studying existing records, ML can also intelligently forecast new attacks, which are typically mutations of earlier attacks.For effective and secure systems, SS networks must evolve from mere secure communication to security-based intelligence provided by ML.Before accessing the licensed channel, an SU must first perform spectrum sensing to determine spectral occupancy (idle/busy).During this step, the SU must distinguish between PU signals and background noise.Spectrum sensing is thus a categorization challenge.Automatic modulation recognition is essential for CR adaptive modulation technique to sense and adapt to changing circumstances.Automatic modulation recognition is a classification problem, and deep learning excels at it.Several studies combine DL with automatic pattern recognition in CRNs.In Reference 85, the authors suggested a framework based on deep learning for obtaining satisfactory automatic pattern recognition.By successfully identifying and detecting inherent spectrum dynamics, ML may automate CR functionalities.But there are two major obstacles.First, ML requires a lot of training data to capture complex channels and emitter characteristics.In addition, when the channel and emitter circumstances change, the training data cannot be used. 93hese issues have led to the development of robust spectrum sensing techniques which proposed a new technique to augment training data and domain adaptability.To increase classifier accuracy and adjust training data to spectrum dynamics, a GAN containing DL structures were used.This method can be utilized for spectrum sensing with minimum training data and no prior knowledge of spectrum statistics, 94 proposing a robust DL-based spectrum sensing paradigm.
The SU receiver filtered, sampled, and sent the received signals straight into a CN.TL was added to the framework to increase the classifier's adaptive capabilities.Using ML to assist the SU to make more efficient decisions as to which channel to sense as well as how or when often to sense improves the SU's sensing performance.Unlicensed SUs must identify the licensed frequency band (spectral hole) owned by licensed PUs in CRN. 95SUs can access the spectrum hole without interfering with any PU.If a PU reactivates, the SUs must leave the spectrum.The capacity to forecast can help SUs execute spectrum sensing more efficiently.Sensing time can be lowered by allowing SUs to pick data transmission channels and estimate channel idle time.The CRN has unique and severe problems in this uncertain context.To overcome these obstacles, ML algorithms can assist in increasing system performance.The most critical step in fighting against such attacks is identifying malicious attackers.This is done by extracting specific features from received signals.Distinct traits may reflect the transmitters' personalities, making them distinct.The user location-based approach is widely used to distinguish attackers from PUs.Since Received Signal Strength (RSS) changes by location, it can be used to identify both location and user type.Finally, physical layer detection uses transmitter or channel characteristics to detect intruders and typical transmitter fingerprints include phase and frequency changes.
According to the literature, there is a high demand for spectrum and offend to enable secure communication.However, the CR's physical layer security is a critical concern.Malicious users interrupt the spectrum sensing process by posing as the PU in the absence of the PU, causing the CR to make an incorrect judgment about the PU's existence.PUEA refers to this type of attack.As a result, a reliable system for identifying and mitigating rogue users must be devised.Because of this, the research focuses on creating such a tag and embedding it in an appropriate manner to mitigate the PUEA and SSDF attacks on the network.This paper focuses on a scenario in which a helper node based on the ED method is utilized to sense a PU's availability utilizing the tamper-proofed LS Tag method over an SCFDMA system.Several experiments have been conducted, with simulation results demonstrating the benefits of the proposed approach.The absence of mature green network modules in prevailing simulators offers new research direction to enterprises an experimental testbed to provision green networks.In such a way, the promising innovations in biosensor technology predict the detection of the life-threatening COVID-19 pandemic in remote regions, which will benefit all human beings who have difficulty commuting for rapid, selective, and sensitive detection. 96,97reen IoT refers to the use of environmentally sustainable and energy-efficient technologies in the development and deployment of IoT devices and systems.This approach aims to reduce the environmental impact of IoT by optimizing energy consumption, minimizing waste, and reducing carbon emissions.To achieve this, advanced technologies such as AI, Unmanned Aerial Vehicles (UAVs), and cognitive IoT are utilized to create sustainable and eco-friendly smart cities.One example is Smart Packet Transmission Scheduling in Cognitive IoT Systems, which can reduce transmission delay and packet error rate while maintaining energy efficiency. 98Additionally, Tethered Balloon Technology for Green Communication in Smart Cities and Healthy Environments can help reduce CO2 emissions and minimize radiation and ecological hazards. 99Green IoT is a crucial aspect of Industry 4.0, as it enables pollution monitoring and promotes energy efficiency through the adoption of green ICT technologies.1][102] The green IoT framework suitable for future wireless communications has a holistic view that considers all the risk assessments, and technological facts, and tries to maintain the ecosystem for developing next-generation green communications.The research progress in this field is illustrated in Figure 1, where the expressive concern and awareness of green CRN research is the subject of attention.

F I G U R E 1
Organization of the research.

PROPOSED METHODOLOGY
A growing number of apps and services that store and share data must be effectively protected from threats such as hacking, tampering, and unauthorized access. 103However, PUEA still stands as the foremost issue in CRNs.ED, commonly known as radiometry or periodogram, is one of the most popular and easiest methods of spectrum sensing. 104To diminish the influence of PUEA in CRNs, an efficient tag generation structure is proposed utilizing LS tag generations.As of now, energy-efficient approaches are focusing on the security of either network or mobile nodes.A secured green communication for both networks and mobile nodes, along with an energy-efficient encrypted technique is required for future wireless communication systems.

System model
Spectrum sensing is essential in a CRN.For communication to take place, this is the first step.Spectrum sensing is a hypothesis test that can be summarized as a problem of recognition.One of two hypotheses must be selected by the sensing device.
In this case, the signal received by the secondary users is S(t).d(t) is the primary user-transmitted signal.The variance-added white Gaussian noise n(t).Hypothesis H0 states that there is no PU and simply noise in the frequency band of interest, whereas H1 states that there is PU.Consider the ED system depicted in Figure 2, In the context of cognitive radio networks, primary users are the licensed users of a radio frequency band or spectrum who have exclusive rights to use the spectrum.Cognitive radio users, on the other hand, are secondary users who can dynamically adapt to changes in the radio environment and make decisions about how to use the available spectrum.A helper node is a device or system that provides support or assistance to another device or system in the network, such as by providing information about the radio environment or coordinating the use of the spectrum among multiple users.Cognitive radio with SC-FDMA (single carrier frequency division multiple access) is a multiple access technique used in cognitive radio networks that allows multiple users to share the same frequency band by assigning each user a unique subcarrier.However, there may be attackers who seek to cause harm to the network, such as disrupting its normal operation or stealing sensitive information.To prevent this, the spectrum sensing, and energy detection method is used to detect the presence of primary users and determine when it is safe for cognitive radio users to use the spectrum.Hence to ensure assume that n, SUs is located inside the coverage area of a primary base station (PBS).A PUEA near the SUs transmits PU-emulated signals.On the PU spectrum, the PBS transmits downlink signals to the PUs.The SUs looks at the presence of the PU signal in the PU spectrum.The ED method is used to detect the spectrum holes without disturbing the PU.To mitigate the PUEA tag generation method has been proposed over SC-FD the MA system.With the binary hypothesis, conventional CR methods describe the presence of the PU signal on the PU spectrum.With hypothesis H0, they denote the lack of the PU signal.Hypothesis H1 denotes the presence of the PU signal.Primarily, the cognitive users are authenticated, grounded on the distribution of "helper" nodes, fixed in a region of the CR Network.Hence, the helper node does the sensing action to discover the availability of the PU.The CR admits the arriving signals with the authentication tags only.The remaining signals get discarded.Latin squares are utilized to make the authentication tags.To ameliorate the working performance, FFT grounded energy detection is utilized for spectrum sensing.Figure 3. displays the diagram of the helper node (SC-FDMA process) where spectrum sensing is a critical component of CR's cycle.Its major objective is to distinguish between two states: the absence of the PU signal, represented by H(0), and the presence of the PU signal, designated by H(1).The following diagram illustrates these two states.
where Z'(a) denotes the received signal, B(a) is the transmitted signal, and Y(a) denotes the transmitted signal's noise.ED calculates the energy of received N samples as the squared magnitude of their fast Fourier transform (FFT) averaged across N samples. 4-8

SC-FDMA
SC-FDMA is employed to perform signal transmission via network nodes in the cognitive environment.The LS-generated authentication tag is XOR with the convolutional encoder output.Then the output of the encoder is fed to serial to parallel converter module.The subcarrier mapping is conducted on the generated SC-FDMA signal.SC-FDMA stands as an FDMA scheme that executes channel equalization and estimation in the frequency domain.The signal for every user is an ("Serial to Parallel") S/P converter and the parallel data is specified as input to the FFT module where the SC-FDMA signal is produced.Subsequently, subcarrier mapping is mapped and then Inverse FFT (IFFT), and parallel to serial conversion are applied.A sufficiently long cyclic prefix is added before the signal is conveyed over the channel to lessen Inter-carrier interference and Inter-Symbol Interference (ISI).At the receiver's side, the first inserted CP of the transferred signal is removed.After the conversion of S/P and successive FFT, subcarrier mapping along with IFFT, the user's signal in frequency province is de-mapped from the distributed subcarriers.

Helper node
The helper node generates the signal with the help of the SC-FDMA technique.Figure 4 demonstrates the flow diagram intended for the proposed tag generation scheme using a helper node.A helper node is presumed to be in a much nearer range to the PU.It acts as a link between the PUs and the SUs.The CR requests the helper node to do the sensing process.This node senses the availability or presence of the PU.In the case of PU absence, it directs the above facts to the cognitive user along with the authentication tag.

Tag generation
The helper node creates the signal after adding the authentication tag for the signal and then transmits it to the cognitive receivers.The nature and other information of this authentication tag are already acknowledged by the cognitive users.To produce the authentication tag, the Latin square matrix is utilized.These tags are attached to the output convolutional encoder.In general, the Latin square is considered as an x × x array packed with x diverse symbols {1, 2, … , n}, each exists specifically once in every row furthermore specifically once in every column.Mathematically, a Latin square with order x via a tri-tuple function of (r, c, k) as follows: where r signifies the row index of an element in L. C denotes the column index for an element in L and k denotes the symbol index for an element in L, are utilized for 1D signals.But in general Latin square matrices are 2D.Figure 4 designates the algorithm flow for the proposed LS tag generation (Algorithms 1 and 2).
Algorithm 1. Latin square matrix tag generation based on N sequence 1.Begin. 2. Input: Q1 and Q2 are two length N sequences.3. Output: L is a Latin square of order x. 4. Key generation of order (1 × x). 5. Generate random no.Sequence using LCG.6. for L = LSG (Q1, Q2). 7. Split the random sequences and store it in Q1 and Q2 a.

Algorithm 2. Steps on LS tag generation
Step 1 • Generate a key of order 1 × x.Key generation is utilized for generating 1D the matrix.The input is two sequences, Q1 and Q2, of length N.
Step 2 • A random number sequence is generated using the produce random numbers or sequences using PRNG for example, linear congruential generators (LCG), a simple method for generating a sequence of pseudo-random numbers.
Step 3 • The random number sequence generated in step 2 is split and stored in the Q1 and Q2 sequences.Then split the random sequences equally and store them in Q1 and Q2, both Q1 and Q2 are length-N (same size as key).
Step 4 • Q1 and Q2 are sorted using the SortMap function.This function rearranges the elements of the sequences in a particular order.Perform index mapping using sort map (Q) function.It first calculates the sorted version of the sequential inputs, executes index mapping betwixt sorted version and the input sequence in the ascending order and then obtains the seed and shift sequences.
Step 5 • Performs row shifting operation using Row shift (Q, v).Finally, the Latin square of the order x is obtained.For each row (r) in the range of 0 to x-1, a row shift operation is applied to the sorted sequences (Q_(seed)) .This operation shifts the elements of the row by a specified number of positions.The result of the row shift is stored in L'(r,:).

Experimental setup
The proposed work's primary objective is to mitigate the PUEA effect with the aid of an authentication tag for the efficient utilization of spectrum resources.Using MATLAB 2020, the proposed secure communication was simulated.The performance of the proposed robust CR-SS approaches is demonstrated in this section using computer simulation.In this research, the Amplitude Modulation (AM) signal is used as the PU signal.The effectiveness of the designed ED-based spectrum sensing approach is first discussed.The influence of LS Tag methods on classification is next investigated.Finally, the proposed resilient CR-SS methods performance is discussed.

Simulation parameters
The primary aim of the proposed work is to reduce the PUEA effect by using an authentication tag to ensure efficient spectrum resource utilization.For signal transmission, SC-FDMA is used.The unused spectrum must be detected using the ED method, whereas the CR may efficiently sense the spectrum using the ED method without disturbing the PU.In a CR situation, the PUEA is critical.Introduce the tag generation utilizing the LS matrix approach to prevent the PUEA and successfully communicate the SC-FDMA system.The suggested SC-FDMA system is coupled with CR to mitigate PUEA and can enable secure communication.This task was done using MATLAB 2020 software.Table 1 summarizes the numerical values used in the simulation.The goal is to compute BER performances at the destination, which is depicted in Figure 5 as being located horizontally and vertically in a CR coupled with an SCFDMA environment.After, by computing the probability of detection and of false alarm (PD and PFA) when the PU is available to simulate the results.When the PU threshold value Q p is presented at level 104, PD and PFA are considered in the simulation.If it is greater than 104, it indicates that PU is present; if it is less than 104, it indicates that PU is absent.Table 2 summarizes the corresponding values used in the simulation.

Performance analysis
Most prevailing works on cognitive sensing concentrated on conducting hypothesis testing to choose the PU availability.The detecting performance of the proposed tag generation scheme is determined by the following metrics (a) Probability of False Detection and (b) Probability of False Alarm

Probability of detection with different SNR
The efficiency of spectrum sensing algorithms depends on characteristics like SNR, sample count, and noise uncertainty.Spectrum sensing is used to pick between two hypotheses (H0 or H1) based on received signals. 97The detection probability is the time proportion in which the detecting algorithm accurately detects the PU (licensed).The PU affects the system's performance.The PU will make greater use of its spectrum if the sensing period is enhanced, and the SU cannot interfere most of the time.The PUs will maximize their priority since greater spectrum sensing means more PUs recognized and less interference.The detection probability is the time proportion in which the detecting algorithm accurately detects the PU (licensed).The PU affects the system's performance.The PU will make greater use of its spectrum if the sensing period is enhanced, and the SU cannot interfere most of the time.The PUs will maximize their priority since greater spectrum sensing means more PUs recognized and less interference.The probability of the detection indicates the likelihood of a CR user proclaiming that a PU is existent when the licensed user indeed holds the spectrum.The mathematical manifestation is:

TA B L E 2
The signal is from the PU ∶ user presence H 1 = The signal is from the attacker ∶ user abscence .
The proposed LS Tag generation scheme is contrasted and Scrambled PN sequence tag generation. 22,23The proposed tag generation scheme shows a higher likelihood of detection when contrasted with other such schemes is perceived.4 The information needed to exclude challenged SUs from the sensing interaction process is extensive, including assault strength and SU location.For this reason, a secure and trustworthy sensing algorithm was developed.CSS then excluded the corrupted sensing reports.A reliability value was also assigned to each SU to allow for identification errors.The identification value has also been used to protect trusted users from misidentification.The likelihood that the detecting algorithm detects PUs when none exist.Low probability of false alarms should be desired for SUs to employ detected spectrum when accessible.So, the secondary network's throughput is higher.The FA probability indicates the likelihood of a CR user proclaiming that a PU is existent when the spectrum is free.The FA probability is mathematically signified as: The signal is from the PU ∶ user presence H 0 = The signal is from the attacker ∶ user abscence .
This energy TED is then compared to a pre-define threshold λa and λb; to obtain the sensing decision as follows: The algorithm's detection performance can be quantified using the probability of detection of PD and the probability of false alarm PFA.The probability of detection is calculated as the number of true detections (PU is present) divided by the total number of sensing trials, whereas the probability of false alarm is calculated as the number of times the PU is wrongly detected divided by the total number of sensing trials.These probabilities are denoted as follows:

RESULTS
This section covers the numerical results of spectrum sensing using ED method and to evaluate the performance analysis.
In Figure 5 depicts the link of each node in another network in SC-FDMA based CR net-work scenario and its number of PUs' slots long with spectrum sensing slot, network node positioning to evaluate the efficiency of the proposed model in MATLAB 2020 scenario.The cognitive network with 10 network nodes is considered.The helper node is deployed near-by the PU in the cognitive area.Thus, it is concluded that there is extremely less probability for malicious users attack in the proposed tag generation scheme.The SUs accesses the PUs spectrum band in their absence.In Figure 5A represents the network node positioning.Figure 5B illustrates how the nodes must be linked with another network in a CR environment.Figure 5C shows the spectrum slot between the number of PU spectrum and PU status.It is represented by 1 to 4 slots that are allocated in PU and 5-10 slots showing the PU is absent.It conveys that the CR user can be able to allocate the spectrum without disturbing the PU. Figure 5D depicts the number of users which is allocated for transmitting over the CR-based SC-FDMA system.
Figure 6A plot clearly shows that the proposed LS tag generation method outperforms the existing PN scrambled sequence method in terms of performance.Figure 6B shows the comparison graph of the probability of FA rate with and without LS tag generation schemes aimed at diverse SNR values.From Figure 6C it can be conjectured that the result acquired with the proposed work has a low error rate when contrasted with other plans.From Figure 6C, for 0 to 4 dB, BER for both schemes are equal, whereas with the increase of SNR, BER performances of proposed LS Tag generation are getting better.Thus, the proposed LS tag generation scheme is less prone to PUEA attack.Previously one proposed research suggests a performance analysis of Energy Detection (ED) based spectrum sensing over fading channels.The authors used MATLAB simulations to evaluate the performance of the ED method and compare it with other methods.The simulation results show that the ED method performs well in detecting the presence of primary users in the spectrum even in the presence of fading channels but there is a lack of sensitivity to noise, sensitivity to interference also the performance of ED-based spectrum sensing is highly dependent on the SNR of the received signal, and a low SNR also reduce the detection performance.However, limited detection range for ED-based spectrum sensing has also been addressed which can be affected by various factors such as the transmit power of the primary users and the frequency of the signal. 105n either case, that is, without LS Tag generation and with LS Tag generation, BER decreases with an increase in SNR.To track and estimate the PU in a secure manner, an LS tag method is used.for 0 to 4 dB, BER for both schemes equal, whereas with the increase of SNR, BER performances of proposed LS Tag generation is getting better.As a result, the suggested LS tag generation approach is less likely to be attacked by PUEA.With an increase in SNR, BER reduces in both cases, without and with LS Tag creation.Table 2 Comparison of Probability of detection rate for various generation schemes Therefore it demonstrates that proposed LS tag generation schemes give better utilization of the resource spectrum.The results of the simulation reveal that the suggested CR spectrum sensing approaches are resilient to MU attacks on the CR system.

FUTURE SCOPE
The emerging generation of networks is likely to provision incorporation of various networks having diverse services and protocols.The increasing demand due to expanding population has led to several technological advancements, namely urbanization, and industrialization.The advent of new technological advances has led to a proliferation of smart and intelligent devices, vehicles, homes, and cities.The increasing utilization of IoT, 5G communication networks, machine learning, and artificial intelligence has led to the idea of an "intelligent environment" where all aspects of people, information, things, and practices are interconnected. 106Therefore, vertical handover, management of interference, enhancement in network capacity with the attention of energy efficient communication system is an important research topic.There are several potential extensions to the work presented in this article for future investigation.While this research focused on SC-FDMA for the uplink, it would be fascinating to investigate a scenario in which NOMA, UAV, QR codes, SWIPT-based IoT systems, and CIoT devices are all examined.In future study directions, we will utilize raspberry pi and successfully analyze the system over CIoT-based design.It will be a study of how technology may support both individual-centered and institutional-centered care.To achieve an integrated deployment of CIoT, three primary trades must be prioritized: availability, scalability, and security.However, AI can overcome network defects and support a sustainable system, leading to the specific application of AI in the Cognitive Internet of Things (CIoT).The related article then highlights the challenges in improving smart packet transmission scheduling (TS) in CIoT and proposes a solution using Generative Adversarial Network and Deep Distribution Q Network (GAN-DDQN). 98The article effectively explains how GAN-DDQN is used to enhance smart packet TS by reducing the distance between estimated and target action-value particles.It also outlines how GAN-DDQN training based on reward clipping is used to evaluate the value of each action for certain states to avoid large variations in the target action value.The growth of internet-connected gadgets has resulted in the emergence of security flaws and concerns about security accountability.Green communication networks have been broadly subjugated in bestowing a wide range of real-time multimedia applications.Figure 7 demonstrates the extensive summary of upcoming future networking technologies.To realize a green communication network environment, an intelligent advanced energy efficient technique must be developed.Different energy-efficient techniques have been immensely deliberated in this survey.However, there are still new research areas for more effort to be achieved in this field such as a diagnostics approach based on a POC sensing technology that can be interfaced with the Internet of things and artificial intelligence (AI) techniques for investigating practical informatics via data storage, sharing, and analytics. 107This part of the section delivers future research directions that entail the consideration of the research community.
F I G U R E 7 Futuristic application and related research directions for green wireless communication.

CONCLUSIONS
This article explored the trends of the telecommunication system in the last decade which visualizes a move towards following energy-efficient green communication to design next-generation networks.For each energy efficient technique, an outline of the existing state-of-the-art research, merits, demerits, open issues, challenges, and possible forward way research direction has been investigated.The CR is a game-changing technology that aims to improve spectrum utilization efficiency and reduce costs in next-generation wireless networks.It will alter the way the radio spectrum is regulated, which will require the development of new enabling technologies.The intelligent agent used in CR perceives the environment, turns data into knowledge, and can solve complex allocation and classification problems.CR is an adaptive and intelligent radio and network technique that can automatically identify available channels in the spectrum and improve communication without interfering with licensed users.Hence, a new approach is suggested for detecting PUEA in CRNs with moving primary using ED-based SCFDMA.LS tag method is employed to secure PU estimation.LS tag generation is more resistant to PUEA attacks.It performs better than the current scrambled PN sequence approach.Simulation results show that ED-based CR approach is secure against PUEA attacks.The proposed method successfully detects PU in non-stationary environment.LS tag system is superior in terms of detection and miss detection probabilities.
The results indicate a satisfactory and secure way of PU identification.Overall, ED-based CR approach is robust against PUEA attacks.Furthermore, AI and ML technologies can be utilized to predict the performance, shortcomings theoretically, and practical solutions related to working the specific sensor before undergoing experimental evaluations resulting in saving time, cost, resources, contamination, and workforce. 108Applying ML to various SS frameworks has become a popular research topic and a promising future communication towards futuristic wireless networking.These technologies are blockchain, machine learning, artificial intelligence, big data, advanced beamforming.

F I G U R E 4
Algorithm flow for tag generation.
U R E 5 (A) Network node positioning (B), link of each node in another network (C), spectrum sensing slot (D) number of users.

F
I G U R E 6 (A) Detection rate probability of PU under various SNR, (B) false alarm probability under various SNRs, (C) bit error comparison for different SNR values.
Simulation parameters of the CR emerged with SC-FDMA System.
TA B L E 1 Comparison of Probability of detection rate for various generation schemes.
.2.3 Probability of false alarm at various SNR Identify attackers from legitimate SUs to avoid SSDF.There are two types of defense strategies: outlier detection and reputation management.Outlier identification removes the offending user from the network.The convergence of SSDF and PUE attacks puts the network at risk.The performance of existing SSDF protection mechanisms is compromised if SUs is assaulted or misled by PUE.A PUE attacker's signal can pollute surrounding SUs' sensing reports even if the attacker is identified.Examining previous works' attacks and ML-based responses can provide a firm knowledge of steps to protect in SS systems.In Reference 87, investigated CRN secure sensing under PUE and SSDF attacks.