Energy efficiency in 5G systems: A systematic literature review

6.1Phase 1 – plan review

The appropriate searching strategy is used to specify the important research questions and the creation of review protocols in this first stage of the research technique.

(a) Searching keywords

An effort has been made to focus our search on the most pertinent search phrase to ensure that the evaluation closely covers Digital Twin and Energy Efficiency for 5G Systems.

As a result, we begin with the keywords and then completed the following steps:

• • Taking the key terms from our study questions and separating them.

• • Using alternative term spellings.

• • Replacing the search queries with relevant papers’ keywords.

To find the most immediately pertinent articles in the literature, we used the primary alternatives and included the “OR operator” and “AND operator,” as shown in Table 1.

Table 1

Inclusion and exclusion criteria description

Id	Keywords
1	(“Digital Twin” AND “5G network”)
2	(“Energy Efficiency” OR in AND “5G Systems” AND using OR “Digital Twin”)
3	(“Energy Efficiency” AND in OR “5G Systems”)
4	((“Energy Efficiency” AND in OR “5G Systems”) OR (“Digital Twin” AND “5G network”))
5	((“Energy Efficiency” OR in AND “5G Systems”) AND (“Digital Twin” OR “5G network”))
6	((“Energy Efficiency” OR in OR “5G Systems”) OR (“Digital Twin” OR “5G network”))
7	((“Energy Efficiency” AND in AND “5G Systems”) OR (“Digital Twin” AND “5G network”))
8	((“Energy Efficiency” OR in OR “5G Systems”) AND (“Digital Twin” OR “5G network”))
9	((“Energy Efficiency” OR in OR “5G Systems” OR using OR “Digital Twin”))
10	(“Energy Efficiency” AND in OR “5G Systems” OR using AND “Digital Twin”)
11	(“Energy Efficiency” AND “5G systems”)
12	(“Energy Efficiency” AND “Digital Twin”)
13	(“5G Systems” AND “DDoS”)
14	(“Energy Efficiency” AND “DDoS”)

Figure 3.

Methodology flow.

6.2Phase 2 – conduct review

We carried out the review at this step-in accordance with the research questions, protocols, and keywords. As per Table 2(a) and (b), this phase focuses mostly on the addition and elimination of publications.

6.3Report review

The five research questions were answered using data that was taken from the primary studies. In the reporting of the results, the recommendations of the references [15, 16] were strictly adhered to.

7.1Which of the domains focuses primarily on 5G systems research?

This section reviews 30 studies that covered 5G Systems. A list of accomplishments to 5G Systems is provided here. The Unmanned Aerial Vehicles (UAVs), a new area of aerial robotics, have drawn a lot of attention from researchers in the field of wireless networking. Swarms of UAV will begin to populate the skies of our smart cities as soon as national laws permit them to fly autonomously [17]. These UAVs will be used for a variety of tasks, including delivery services, infrastructure monitoring, event recording, surveillance, and tracking. 5G/B5G cellular networks are advantageous for the UAV ecosystem because they may be used in a variety of ways to improve UAV communications. These smart gadgets provide a huge variety of communication systems and use cases due to the intrinsic properties of UAVs pertaining to flexible movement in 3-dimensional model, autonomous operation, and intelligent placement. With the UAV being incorporated as a fresh aerial User Equipment (UE) to establish cellular networks, the UAV intends to present an in-depth investigation of incorporation efficiencies between 5G/B5G communication network and UAV technology [18]. The UAVs in this integration take on the function of flying users while in cellular coverage, hence the phrase “cellular-connected UAVs.” The UAV’s major goal is to give a thorough analysis of integration difficulties along with significant 5G/B5G technological advancements and current work on design development and field tests supporting cellular-connected UAVs. The UAV covers current third generation partnership project (3GPP) standardization progress updates and underlines socioeconomic considerations that must be taken into consideration before this potential technology is successfully used [19].

To satisfy the demands for large frequencies, low latency, high performance, and high dependability, certain nations have begun developing 5G networks. Applications based on the Internet of Things (IoT) are also growing daily [20]. 5G would make it possible for more Internet of Things applications, including connected autos that can manage traffic. It will be crucial to take security issues into account once 5G and IoT are combined. The current research has covered a range of 5G concerns as well as the security risks posed by IoT. The repercussions of the most serious cyberattack, known as a Distributed Denial of Service (DDoS), call for a lot of attention, making finding solutions essential. As a result, an overview of the development of 5G technology allows IoT applications, and the size of DDoS attacks in such applications [21].

Numerous systems and gadgets are connected via 5G in almost real time, which has both advantages and disadvantages. To accelerate the safe delivery of 5G-enabled prospects to market, MITRE and its collaborators are validating the technology, policy, and standards. Many of us learn about 5G through advertisements. Therefore, it makes sense that people could assume that this 5th wireless technology will lead to quicker internet or the next step in gaming. However, the opportunities go well beyond consumer device improvements [22].

Globally, 5G network commercialization is accelerating. 5G communications are viewed from the viewpoint of industries drivers as being essential for improvements to consumer consumption experiences and digital industrial revolution. Globally significant economies need 5G to be a crucial component of long-term industrialization [23]. In terms of business, 5G will penetrate hundreds of industries, and technically, 5G must further combine DOICT and other technologies. Hence, the suggestions made are that ongoing study on the 5G advanced network’s needs further evolution [24].

The link between online consumer behavior and e-commerce Quality of Service to better understand consumer psychology activities and lower the risks associated with cost control for businesses needs to be examined. It also hopes to study e-commerce by segmenting online consumer groups [25]. The consumer consumption patterns and expenses using 5G technology needs tracking. First, it is determined that the core service performance, system performance, performance, guarantee and completion, as well as the performance of after-sales service, are the e-commerce Quality of Service assessment factors based on the body of existing literary works and interviews conducted with online consumers. The reliability of the collection variables was examined using SPSS software for statistical analysis based on the information gathered from the questionnaire in Y. Si research, thereby raising the caliber of the sample data. On these measurement items, they run a statistical analysis that is descriptive. Through the examination of experimental examples, it can be demonstrated that the deployment of 5G technology can very well fit consumers’ consumption habits and messaging demands and can assess the reliability of e-commerce operations more correctly and scientifically. Each characteristic calculates the relevance of each dimension [26].

5G is now in its early stages of commercialization. With its improved capabilities and cutting-edge features, the 5G System will completely transform the current wireless network. The Third-Generation Partnership Project (3GPP) is currently working on 5G New Radio (5G NR), also known as the global standardization of 5G, which may operate over a variety of bandwidths from less than 6 GHz to Millimeter wave (mmWave) (100 GHz). The three core 5G NR use cases – Ultra-Reliable and Low Latency Communication (uRLLC), Mega Machine Type Communication (mMTC), and Enhanced Mobile Broadband – are the main areas to look at for 3GPP Enhanced mobile broadband (eMBB). In comparison to LTE systems, several features such as scalable numerology, flexible spectrum, onward compatibility, and ultra-lean architecture are added to achieve the goals of 5G NR [27]. A quick rundown of 5G NR’s new features and important performance indicators. There is a good job of addressing the problems associated with inter-radio access technology (RAT) handover synchronization and the adaptation of higher modulation schemes. A proposed next-generation wireless communication approach takes these issues into account. The architecture serves as the foundation for the transition to networks that go beyond 5G and 6G. It also provides an overview of various 6G network technology and applications. Edge computing, quantum computing, wireless optical communications, hybrid access, and haptic services will all be included in the 6G network. A virtualized network slicing-based 6G architecture is suggested to provide these various services [28].

By advancing linked business realities, the fifth generation of smartphone bandwidth, cellular technology, and networks, or 5G, promises to significantly alter mobility. Effective Service-Level Agreements (SLA) and the foresight of Service Level Objective (SLO) breaches become essential in such a developing environment. It becomes difficult to ensure the necessary service quality while simultaneously assuring effective resource allocation [29]. Furthermore, a variety of Quality of Service (QoS) assurances for a broad range of services, applications, and consumers with a diverse set of needs are anticipated from 5G networks. The ability to translate user-friendly business terminology into resource-specific monitoring features that may be utilized to manage resources in the 5G core network, however, it is becoming more challenging. SLA management framework with QoS provider is introduced to close these gaps. An adaptive tracking algorithm will support the architecture to deliver amazingly precise information in real time without adding extra network traffic. The method eliminates the fixed time interval utilized in the monitoring system. To ensure that pertinent QoS measures are incorporated into the service-level agreement (SLAs), the proposed architecture additionally includes a recommendation system that assesses the importance of various QoS metrics using enriched metadata data from an Network functions virtualization (NFV) Catalogue [30].

With positive effects on energy use, available resources, resilience, and customer experience, the centralized cloud computing landscape is increasingly shifting to distributed and decentralized clouds [31]. Industrial Artificial Intelligence (AI), block chain technology, and 5G wireless technology are three examples of rising IT trends that have an impact on the creation and implementation of next-generation cloud computing, according to most studies. An integrated edge-fog-cloud architecture is suggested for effective application in manufacturing systems after the merging of these innovations in cyber-physical network and cloud-based paradigms in 5G as discussed [32].

Many nations and governments view smart cities as a solution to resource depletion, population increase, and global warming. The process of building a smart city is fraught with difficulties [33]. Digital Twins and 5G hold great promise for the conversion of the current urban governance paradigm toward smart cities, along with the Internet of Things. Fifth-generation wireless communication systems, block chain technology, collaborative computing, simulation, and Artificially Intelligent technologies all use both DT and 5G Systems [34]. The idea of a Digital Twin city (DTC) is put forth in most research, a DTC’s properties, core technology, and use cases are described. They also go over the framework, hypotheses, and future directions for DTC research [35].

By considering a mobile edge computer network with ultra-reliable, low-latency, and delay-tolerant services. By improving user identification, allocation of resources, and offloading probability pursuant to the Quality-of-Service requirements, there is a hope to reduce normalized energy consumption, which is described as the energy consumption per bit [36]. The mobility management entity (MME) controls user association, and each access point controls allocation of resources and offloading probability (AP). In some deep learning (DL) design, the DL algorithm is trained at a central server using Digital Twin and 5G of an actual network environment. The pre-trained deep neural networks (DNN) can supply the MME with a real-time user association strategy. Given that real networks are dynamic, the Digital Twin continuously tracks these changes and adjusts the DNN as necessary. There is a suggestion of an optimization approach to determine the best allocation of resources and offloading probability at each AP for a certain user association scheme. The simulation findings demonstrate that, in comparison to a current method approach strategy to the efficiency of the optimal global solution, the method achieves lower normalized energy usage with less computational complexity in 5G Systems [37].

7.2How can digital twin address the 5G system’s difficulties with energy efficiency?

Eleven (11) research papers covering the topics of Digital Twin (DT) networks in 5G Systems and improved Energy Efficiency were chosen for this part. Each study’s specifics are given in Table 7.

For creating and improving 5G networks and beyond, the DT technology was launched, and various DT functions and designs were examined. The DT may undoubtedly assist the research community in moving away from traditional network designs (which involved actual deployment) and toward digital/virtual ones, including measures for testing and validation. However, since this DT technology is not currently commercially available for 5G, there is a unique possibility for both DT and 5G study communities to develop ideas that will enable revolutionary 5G use cases for the society of the future. The cost and efficiency of the production line for hollow glass are significantly influenced by the customized design [50]. Most articles suggested quick, personalized production line design based on Digital Twins. Some studies suggested an idea of “iteration between static design and dynamic execution” by analyzing configuration factors in the development of production lines.

As such, it is designed for operating process requirements to meet process limitations and the associated optimization problems. The production line’s performance can now be more accurately digitally simulated to the suggested Digital Twin concept [33]. Some proposed decoupling methods can offer design with an optimum guide, as shown by a good design application in a thin glass matching warehouse system. A variety of controlled flow type manufacturing systems, including a production line for custom furniture and a line for 3C products, can use the suggested Digital Twin based paradigm. In the future, focus on integrating big data analytics will be taken into the Digital Twin approach for both production line design and operation. One of the fundamental pillars of the innovation process is the technology of the Digital Twin [35, 51]. It enables the creation of digital representations of physical systems, which has the potential to accomplish many advantages like real-time monitoring and greater productivity and efficiency. To provide users of Digital Twins with a high level of service and enable the implementation of Digital Twins in far more domains, most papers highlighted the significance of the network that links the physical and Digital Twins around each other, as well as the prerequisites that should be available in the connecting network. The DTC will change city systems of governance and laws as a new paradigm for creating smart cities and infuse ongoing energy for urban development and transformation [52].

Digital Twin cities are being planned in many significant cities throughout the world. Building Digital Twin cities is now conceivable to the quick advancement of Digital Twin technology. Some studies suggested a pattern for Digital Twins by examining the current Digital Twin technologies and basing it on the growth of smart cities [53]. Through mapping and surveying technology, IoT perception, develop solutions, simulation, and deep learning (DL), a platform for urban management and maintenance is built that is identity, conscience, self-organizing, self-executing, and adaptive. Investigating how URLLC and delay-tolerant services could lower consumers’ average energy consumption in a Multi-access edge computing (MEC) system is done. To train the algorithm off-line, a presentation of a deep learning (DL) design for user association in which a Digital Twin of the network environment was set up at the central server [37, 54]. The MME that handles user association received the DNN after the training phase. Suggestion on a low-complexity optimization approach that improves allocation of resources and offloading probability at each AP with a specific user association scheme is done. Additionally, as compared to an existing technique and strategy for the global optimal solution, most modelling schemes can lower normalized energy consumption with the DL algorithm and with less computational complexity [55]. Technical literature actively examines concepts for developing Digital Twins to address a variety of issues that can occur for different types of businesses. Using Digital Twins, significant new results can now be obtained with advances in technology and the production of the requisite software. Some studies put out the idea of creating a system of Digital Twins to address a variety of real-world issues in intricate communication networks. These tasks are seen as examples of two applications [34]. For businesses involved in the life cycle of a communication system, the initial application is the management of information feedback: “elaboration of modernization concepts, creation of technology, design, construction, and operation.” A second application is keeping an eye on traffic, including any unusual behavior, to prevent mistakes from being made again in case of emergencies.

The Digital Twin connection will also be a useful tool for multidisciplinary study, which is another key element [36]. By creating a digital version of the physical entities, Digital Twin (DT) gives precise advice for inter scheduling algorithm in 5G edge computing-enabled distribution grids. In most studies, discussion on the important issues of low accuracy, long iteration delays, and security concerns in DT construction and DT-assisted resource scheduling was made. Description of a secure and delay Digital Twin assisted resource scheduling algorithm and a collaborative learning-based DT infrastructure Security Administrator’s Integrated Network Tool (SAINT) was done. By jointly optimizing its total iteration latency and loss function and utilizing abnormal model identification, SAINT provides low, accurate, and safe DT antimicrobial resistance (AMR). To enhance the effectiveness of deep Q-learning, SAINT uses DT to provide intelligent resource scheduling deep Q network (DQN) [56]. Because of the consideration of long-term limits, SAINT provides connect precedence and energy consumption awareness. SAINT performs better than cutting-edge algorithms in terms of cumulative iteration latency, DT loss function, energy usage, and access preference deficit. Addressing the 5G edge computing-enabled distribution grid’s resource scheduling issue with DT assistance [57]. To obtain a low-latency, accurate, and secure DT, SAINT was proposed. SAINT reduces the DT gradient descent by 58:06% and 70:39%, the DT error function increases by 46:92% and 58:53%, and the cumulative iteration latency by 12:23% and 31:89% compared to CS-unit control block (UCB) and RS-DNN [58]. Due to access priority and energy consumption awareness, SAINT also achieves the lowest access urgency deficit and cumulative energy consumption. The Digital Twin makes 5G testing easier and more efficient for university R&D initiatives and the kinds of private networks that are now cropping up on campuses, in factories, and in distribution and manufacturing facilities around the world [59]. Digital Twin makes a perfect testing-on-demand solution when combined with other cutting-edge Spirent products like Lab as a solution with intelligent automation. It is quickly emerging as the preferred option for realizing unified connectivity, optimized assistance, and secure communications within a particular location, such as a campus, factory, or geo-fenced area, to its significant advantages for modelling and supporting dedicated, next-generation operations [60].

7.3What methods can be employed to address the issues of energy efficiency in 5G systems?

Energy Efficiency has been employed in previous research to enhance network performance. Numerous methods and frameworks for Energy Efficiency were suggested in the literature evaluations on Digital Twin’s network and 5G Systems. To satisfy the needs of consumers and owners, DT analysts employ a variety of strategies in various sectors. Therefore, 28 studies that emphasize the fundamental methods and their contributions to enhancing the network’s Energy Efficiency in terms of speed, accuracy, and performance have been chosen for this part. Each study’s specifics were covered.

To increase operational effectiveness, productivity, and automatization, Cyber-Physical Systems (CPS), Big Data (BD), and the Internet of Things (IoT) are merged. Despite emphasizing these advantages, smart factories must also use energy sources that are sustainable and renewable [78]. According to estimates, industry operations consume around 35% of the world’s total electrical supply and contribute about 20% of all carbon emissions. Examination on the effect of energy upon that reach of Industry 4.0 in its environment. It was suggested to conduct a thorough literature review of current research on smart factories and energy use. To address the questions, a selection of papers was assessed. As a result, revisions have been made to more than 2,500 articles that were published in the last ten years [79, 80]. A trustworthy approach of article selection was used to identify the most important themes in energy conservation in smart factories.

In addition to offering a new taxonomy to structure projects that correlate the consumption of energy as well as the new demands of Industry 4.0, the contributions include highlighting unresolved problems and difficulties in the field [81]. The survey that is being presented attempts to assess the state-of-the-art interaction between smart factories and energy. To extract and arrange the articles to respond to the stated survey questionnaire, a systematic literature search methodology was used to see how effective Energy Efficiency was. A focus on ways to reduce energy use and challenges encountered while tackling energy in smart manufacturing [10, 82]. Along with the survey, tools, their connection, and generated data were all investigated. Finland has some of the highest mobile data consumption rates in the world. The use of mobile data has rapidly increased, and future expansion is anticipated [83]. The absolute yearly energy consumption associated with the operations of mobile providers appears to be rising even though the energy used per transferred gigabyte has significantly dropped. Consumer behavior is altering concurrently, even though modern consumer electronics use less energy, we use mobile devices more and more frequently [84].

Does rising usage negate any energy savings made? What options are there for addressing the rising energy demand? Recently, there has been an increase in interest in finding ways to reduce energy usage in wireless communications [85]. New network architectures like wireless links, dispersed antennas, cross cellular; advanced physical layer techniques like multiple-input multiple-output (MIMO) and frequency division multiplexing (OFDM), cognitive radio, network coding, collaborative communication; radio and network resource management schemes like various cross layer optimization algorithms, switching devices saving, multiple radio access technologies [86]. In most posts, provision of an overview of various technologies and the most recent information on each topic was made. The difficulties that must be overcome in the location are also described, because it will have significant economic and ecological impacts soon, cellular networks’ Energy Efficiency has recently attracted scholarly attention from academia and industry. A strategy for collaboration between Long Term Evolution (LTE) and the next wireless communications is put out in some papers [87]. The fifth generation (5G) wireless technology seeks to compromise between wireless network performance (maintaining the focus on high-speed packet rates throughout congested traffic durations) and Energy Efficiency (EE) by alternately switching 5G base stations (BSs) off/on based on the traffic immediate load condition while, at the same time, assuring broadband service for mobile subscribers by the current active LTE BSs [88]. The optimal LTE BS parameters (technical support, total high antenna, spectral range frequency, and transceiver ratio) that ensure maximum reception for the entire region as during switch-off session for 5G BSs were determined using the optimization of particle swarms (PSO) technique. With a maximal data throughput of close to 22.4 Gbps during peak traffic times and 80.64 megabits per second during such 5G BS switched-off period and assured full coverage and over entire area by the remaining operational LTE BSs, simulation findings show that the energy reduction can exceed 3.52 kW per day [89].

The suggested BSs switching-on/off algorithm has merit since the existing dense BSs placement, which is the driving force behind the narrow coverage area concept, has led to unexpected traffic patterns for each BS. The suggested solutions reduce energy use by keeping an eye on the network’s traffic load and making decisions about whether to turn on or off specific BSs. Most studies investigated the possibilities of balancing network efficiency and energy savings in LTE and 5G dual wireless access cellular networks by turning on and off the 5G-BSs in response to the current state of traffic load while maintaining service and coverage [90]. However, the number of BSs which will be switched off will determine how much energy is saved by this method. Energy Efficiency (EE) is the cell inside the User Equipment (UE), and it has since evolved into a crucial component of UE designs. Factors including UE power usage, UE battery capacity, and UE transceiver are now being studied. Recent years have seen an increase in the importance of Energy Efficiency as a determining factor [91]. A key design element that determines the potential for Energy Efficient transmission is the type of reference signals that must be transmitted irrespective of the network demand, including cell-specific reference signals (CRS) for 3GPP LTE systems. The coverage regions for green CoMP with collaborative dimensions of 2 and 3 are inadequate to link all the UEs originally connected by turned-off evoled node B (eNBs), there is a large energy consumption. Although it is projected that increased transmission power will lead to decreased Energy Efficiency (EE), its gap gets smaller as cell size increases [92]. For whatever transmission power investigated, the EE is essentially same at 1200 meters. On the other hand, when Bandwidth (BW) increases, EE increases significantly.

The impact on EE is comparable whenever Mobility Controllers (MCs) transition from a middle level to a higher order scheme. In fact, EE becomes increasingly susceptible to MCs changes as bandwidth grows. It is believed that enhancing Energy Efficiency is essential to reducing operating expenses and environmental effects [93]. By automatically updating internet bandwidth to the real traffic requirement at a given time, the network’s radio access can achieve most of its energy savings. Energy Efficient radio resource planning is one of the best ways to reduce the energy use of wireless devices. Developing heterogeneous networks (a combination of femtocells, small cells, macrocells, and femtocells) and various relay and cooperative connections need also be considered in addition to Energy Efficient cell-size design [94]. A great way to improve Energy Efficiency is by installing Energy Efficient infrastructure and switching off as numerous base stations as you can. It is essential to increase the Energy Efficiency of wireless cellular networks. To make wireless cellular networks more Energy Efficient, extensive research must be done [95]. It simply relates to the UE’s charger and has since developed into a crucial part of UE designs. Metrics including UE power usage, UE battery capacity, and UE development of renewable energy are part of current research efforts. For building Energy Efficient wireless networks, the Multiple Input Multiple Output (MIMO) and Analogous frequency division transmission method (OFDM) techniques, connectivity techniques, Energy Efficient network strategic planning, various relay, and cooperative communications, and cross-layer optimizations were all introduced [96].

The idea is to first develop concepts for Energy Efficiency before improving them to improve QoS metrics. Then, by perfectly alright the energy-saving technique considering the technical expertise, QoS may be quickly improved. To prove the value of QoS technology, energy conservation methods should be taken into consideration during QoS upgrading [97]. The amount of energy used depends on both the volume of data and the separation between sensor nodes. The spacing distance may have an impact on how much power the transmitter and receiving circuits use. If the amount of data transferred between numerous sensor nodes is the same, a smaller gap distance between them would result in less energy being used [98].

7.4How can qos be used to increase the energy efficient of a network?

Six (6) studies that show the effectiveness of Quality of Service (QoS) and Energy Efficiency employed in 5G Systems were chosen for this part. Each study’s specifics were given.

It is crucial to create a configurable data forwarding that is conscious of QoS and is adjustable to enforce the QoS obligations to fully satisfy the various Quality of Service (QoS) needs imposed by various network slices for various vertical applications. For the edge and core subnetworks of a 5G network, a concentration on creating prototyping, and evaluating a unique privacy data-plane network slicing architecture. By device congestion control, priority configuration, and scheduling, the framework may manage differentiated services. A prototype that suggested framework using the most successful and able field-programmable circuit array platform, and empirically assess the system’s performance. Results from experiments show that the suggested framework can perform proxy broadband segmentation at the data layer. To partially meet the enormous demand for high throughput in 5G networks, the network infrastructure is vertically dense with so-called limited base stations. The efficiency of 5G networks inherently suffers from: 1) significant inter-cell interference brought on by network density and 2) significant energy waste because of the high redundancy of weakly occupied, often, and small-cell base units [111]. The use of continuous learning method to intelligently direct the search toward useful answers to get around the repetitive quality of exploring in the enormous strategy space. More specifically, to balance high bandwidth usage and Energy Efficiency, some approaches use orthogonal frequency division multiple access (OFDMA) sequencing and dynamically learn effective user-equipment association strategies. Inter-cell disturbance and the variety of user QoS requirements are both considered by their model. Information and computing technology (ICT) research have recently concentrated on Next-Generation Networks, Software-Defined Networking, 5G, and 6G. One of the research topics in 5G is the optimal operational state for device-to-device (D2D) communication, which aims to enhance service quality using the frequency spectrum structure. D2D connection working modes are chosen to satisfy the predetermined system requirements and offer the network’s highest throughput.

Due to the intricacy of the direct answers, a structured issue as an optimal solution and, using extensive simulations, identified the best operating modes for various system characteristics [112]. The use of computed network throughput after identifying the ideal operating modes for the links, and as a result, was able to achieve the highest throughput. To satisfy system requirements, D2D communication couples are more likely to employ full-duplex (FD) mode at small distances, and as a result, most communications occur in FD style at these distances. These findings show that FD communication over short distances provides higher QoS and conditions than QoS-D2D technique. The way we live will change drastically because of 5G technology. The 5G network can adapt to emerging simulation results followed by a significant growth in traffic and the number of connected devices to a flexible combination of improving current infrastructure solutions and innovative wireless ideas and meeting the needs for efficiency and scalability [113]. The dynamic resource exchange and context sharing requirements of the upcoming 5G network age call for ubiquitous, high-speed connectivity. Operators must be able to comprehend and control the services’ performance and quality to meet professional Quality of Service (QoS) and Quality of Experience (QoE) requirements. The subjective acceptability of a telecommunication service’s perceived quality by the user is known as quality of experience. In contrast, Quality of Service refers to technical measures that are objective at the network level. Most papers discuss the selection of service relevant QoE measures and modelling of how these are impacted by various 5G QoS indicators [114]. The effect of QoS requirements on QoE as well as a potential technique for QoS/QoE mapping are provided. One of the key issues with cellular networks is mobile communication. A mobile community sets up the right public model to provide a data service packet at a high rate.

To increase service quality, the bits of mobile data model are required (QoS). The enhancement of QoS and communication systems to 5G cellular operators are provided by some search efforts. A helpful resource allocation mechanism is used at the “Radio Access Network (RAN)” layer to enhance built-in services and the Quality of Service (QoS) in Energy Efficiency of 5G. A local hub or departure station can indicate to or interact with its neighbors to require special modifying of beneficial traffic with QoS labelling, a type of organizational communication. To give website users options for customization, QoS highlighting is useful. By applying the simulation model in 5G’s ultra-low latency, real-time connection, greatly enhanced capacity, and rapid speed, ISP function effectively and demonstrate the full potential of the ubiquitous network [115]. Device connections will become quicker, more effective, and less vulnerable to delays with 5G categories that can use lower, medium, and high broad bandwidth bands. Faster connectivity than 4G is available at close range with mid-bandwidth, which also offers better mobile broadband connections or higher machine-to-machine communication.

For distributing multimedia information like still photographs and movies, wireless multimedia sensor networks (WMSNs) are becoming a popular technology. Multimedia delivery over resource constrained WMSNs faces significant hurdles despite being highly scrutinized by research communities, particularly in terms of efficient energy and Quality of Service (QoS) assurances. Most papers surveys recent advances in methods for creating Energy Efficient and QoS-capable WMSNs. First, examine the distinctive qualities and pertinently enforced requirements of WMSNs. Also provides a summary of each requirement’s current answers. Then, with an emphasis on their techniques for prioritization and service differentiation as well as disjoint multipath routing protocols, recent research cover initiatives on Energy Efficient and computation communication channels, including media access control (MAC) protocols [116].

7.5Which algorithms are more effective in terms of energy efficiency for DDoS attacks in 5G systems?

Nineteen (19) research’s that demonstrate the threats posed by DDoS attacks on 5G Systems to lower the network’s Energy Efficiency are included in this section. Each study’s specifics are presented.

DDoS assaults’ effects on 5G-enabled IoT applications has offered a summary of remedies for shielding the Internet of Things from DDoS attacks. Among the recommendations, which are based on machine learning (ML), strategy might offer a productive setting, where DDoS attacks will be severely limited. Moreover, the research could be expanded to include ML approaches in creating a 5G network optimized IoT application [117]. Attacks such as Distributed Denial of Service (DDoS) can cause communication networks to become unstable by sending a large volume of malicious requests and traffic over the network. Computer networks develop an intricate web of nodes that results in a robust structure. Consequently, in this situation, it becomes difficult to give the user an effective and secure environment [118]. Many strategies have been used before to identify and stop DDoS attacks.

However, they fall short in terms of offering effective and trustworthy attack detection. Because of this, there is still a lot that can be done to strengthen the protection against DDoS attacks. The Gated Recurrent Unit (GRU), a form of Recurrent Neural Network (RNN), is used to protect against new DDoS attacks that are used in the real world. DDoS attacks are detected and identified using a variety of classification methods, including Gated Recurrent Units (GRU), Recurrent Neural Networks (RNN), Naïve Bayes (NB), and Sequential Minimal Optimization (SMO) [119]. The effectiveness of the artificial and deep learning classifiers is assessed using performance review metrics like accuracy, precision, recall and f1-score. According to experimental findings, reflection attacks and exploitation assaults using GRU both produce DDoS classification accuracy readings of 99.69% and 99.94%, respectively. Since Satoshi Nakamoto’s white paper regarding Bitcoin was released in 2008, blockchain has (slowly) grown in popularity as one of the most mentioned strategies for protecting data storage and transmission through decentralized, trustless, and peer-to-peer systems [120]. Most studies identify academic research that aims to use blockchain for cyber security reasons and provides a comprehensive overview of the most widely used blockchain security applications. The number of Internet of Things (IoT) devices and sensors has rapidly expanded with the development of IoT technologies [21, 121]. Large-scale sensor-based systems are predicted to dominate in our societies, necessitating the development of novel design and operation approaches. The cloud is moving to the edge of the network, where resources like routers, switches, and gateway are being virtualized, to serve the computing requirement of real-time delay-sensitive programs of widely dispersed IoT devices/sensors. The open structural layout of architecture and the widespread use of the concept itself give rise to typical security concerns for the currently used networking technologies [122].

Additionally, cooperation creates difficulties because new security issues may impair the systems’ normal functionality and functioning. Moreover, the IoT’s commercialization has given rise to various public security worries, such as the risk of cyberattacks, privacy problems, and organized crime. More precisely, a thorough analysis of security risks and difficulties across the various IoT system design layers were presented [123]. The literature suggested methods and counter measures to deal with these security vulnerabilities was also clarified. Due to its dispersed nodes, server, and software for efficient communication, the Internet of Things (IoT) is essential in many industries, including transportation and the medical industry. Existing Intrusion Detection approaches are unable to reliably thwart attacks even though this IoT paradigm has been subject to infiltration assaults and threats that have raised security and privacy concerns. As a result, the sparse convoluted network has been used to study the IoT intrusion danger to counter threats and attacks. A set of intrusion data, traits, and suspicious activity sets are used to train the web, assisting in the identification, and tracking of attacks, primarily Distributed Denial of Service (DDoS) threats [124].

The network is also optimized using an application that recognizes and detects common errors and attacks attempts under various circumstances. Neurons are used to evaluate complex hypotheses in the sparse network, and the event channel outputs are then propagated to additional hidden layer processes. This method reduces the interference with IoT data transfer. The network correctly distinguishes the normal and threat patterns by using training patterns effectively. When it comes to ensuring network security, Intrusion Detection systems were better than other forms of conventional network defense [125]. Some researchers used an improved genetic algorithm back propagation (IGA-BP) network, an auto encoder network model, and an improved evolutionary algorithm to detect attacks and tackle the increasing issue of internet safety in the big data era. One of the most important recent advancements in computer science is the Internet of Things. Modern computing infrastructures frequently include many low power devices with a wide range of hardware and software capabilities. Such systems cannot be sufficiently protected by existing security models, techniques, and solutions [126]. The use of lightweight agents installed at various Internet of Things (IoT) installations, such as smart homes, is suggested in most studies to jointly identify Distributed Denial of Service (DDoS) attacks carried out by IoT device botnets. A dramatic decrease in Distributed Denial of Service (DDoS) attacks because of the harm it does to organizations’ assets, computer security is a topic that is often discussed. However, it has difficulties due to the rapid increase in computer access speed and internet user traffic. There is need for a thorough study on the DDoS impact’s measurement data and a systematic assessment of the literature with regards to 5G.

Definition, forms of DDoS attacks, and different DDoS attacks currently in use for 5G Systems that need detection and various methods for predicting DDoS attacks [127]. An overview and analysis of the different DDoS attacks detection and prevention techniques has been seen. They have developed a conclusion that there is a significant threat from DDoS attacks to Internet-based shared and distributed systems are crucial, therefore shield of such attacks from occurring against our system is required. Despite the numerous methods for identifying and reducing the DDoS assault there is a flaw in the entire system since the system utilizes high bandwidth to identify genuine users’ data as a perpetrator [128]. Jamming methods range from straight forward attacks that just transmit interference signals continuously to more complex ones that target specific protocol weaknesses. Other preventative strategies call for elements that might not hold true in practical situations. For instance, evasion approaches need for the nodes’ mobility, which may not be viable, while other alternatives call for changes to the current protocols. Solutions that need significant changes (and cannot be implemented, for example, with a software fix) are unrealistic given the already extensive deployment of wireless systems. Attacks via Distributed Denial of Service (DDoS) damage the Internet’s digital accessibility.

The user’s expectation of receiving prompt and efficient services could be severely harmed by DDoS attackers [129]. There are numerous stories of DDoS attack incidents that have had catastrophic repercussions on internet users and web services. Hence, most users are inexperienced, DDoS assaults frequently target their devices, or they unintentionally join the advanced threat army. The 5G growth of mobile will soon take place, but there are rumors that 5G cellular are also vulnerable to DDoS attacks. As a result, it is important to recognize DDoS attacks as a threat. Inter-slice isolation is known to give powerful resource isolation; however, it might lower resource efficiency [130]. Utilization of better control is provided by intra-slice isolation in terms of security, resource availability, and cost trade-offs utilization. Currently, conducting extensive 5G testing is exceedingly difficult, although being gained through a modest testbed, some findings are thought to be highly valuable in this new field of study because they were obtained through actual tests. To determine if the testing experiments can be scaled up, the outcomes for larger testbeds will be the same. They tested and trained their suggested model using the Knowledge discovery (KDD99) dataset. To find the DDoS attack, they applied SVM and logistic regression algorithms. The software define network (SDN) environment has deployed the classification module to differentiate between legitimate and malicious traffic data, the decision tree and support vector machine (SVM) algorithms are used. Most research demonstrated that, in simulated environment, SVM performs better than decision tree approach and Intrusion Detection was the best [131].