The Role of 6G Technologies in Advancing Smart City Applications: Opportunities and Challenges

Abstract

The deployment of fifth-generation (5G) wireless networks has already laid the ground-work for futuristic smart cities but along with this, it has also triggered the rapid growth of a wide range of applications, for example, the Internet of Everything (IoE), online gaming, extended/virtual reality (XR/VR), telemedicine, cloud computing, and others, which require ultra-low latency, ubiquitous coverage, higher data rates, extreme device density, ultra-high capacity, energy efficiency, and better reliability. Moreover, the predicted explosive surge in mobile traffic until 2030 along with envisioned potential use-cases/scenarios in a smart city context will far exceed the capabilities for which 5G was designed. Therefore, there is a need to harness the 6th Generation (6G) capabilities, which will not only meet the stringent requirements of smart megacities but can also open up a new range of potential applications. Other crucial concerns that need to be addressed are related to network security, data privacy, interoperability, the digital divide, and other integration issues. In this article, we examine current and emerging trends for the implementation of 6G in the smart city arena. Firstly, we give an inclusive and comprehensive review of potential 6th Generation (6G) mobile communication technologies that can find potential use in smart cities. The discussion of each technology also covers its potential benefits, challenges and future research direction. Secondly, we also explore promising smart city applications that will use these 6G technologies, such as, smart grids, smart healthcare, smart waste management, etc. In the conclusion part, we have also highlighted challenges and suggestions for possible future research directions. So, in a single paper, we have attempted to provide a wider perspective on 6G-enabled smart cities by including both the potential 6G technologies and their smart city applications. This paper will help readers gain a holistic view to ascertain the benefits, opportunities and applications that 6G technology can bring to meet the diverse, massive and futuristic requirements of smart cities.

1. Introduction

The recent trend in massive urbanization has resulted in several infrastructural issues such as traffic congestion, clean water availability, waste management, power supply, health care, sanitation, etc. The percentage of the population living in urban areas was just 17.95% in 1978, but had risen to 54.7% by 2016, and is expected to reach around 70% by 2030 [1]. This phenomenal rise in urbanization can create severe challenges in areas related to sustainability [2,3], resilience [4], and de-carbonization [5]. These issues pose an urgent need for the efficient use of limited resources, a challenge that can be addressed through the implementation of advanced digital technologies and the intelligent use of data collected from ICT-based infrastructure deployed in cities [6]. Digital technologies have the transformative power to deliver progressive, sustainable and positive impacts on urban life [7,8]. The focus of both practitioners and academicians is to understand and develop cities that are both sustainable and smart, aligning with the United Nations’ Sustainable Development Goals (SDGs) and 2030 Agenda [3]. So, it would be more suitable to refer to smart cities as “Smart Sustainable Cities” (SCCs). Since the last decade, several countries have been focusing on developing smart cities. The world-wide anticipated market size of smart cities in 2023was estimated at USD 1.48 trillion and is expected to rise to about USD 12.02 trillion by 2033 [9].

Mobile communication networks, in the last few decades, have witnessed phenomenal development. The enormous capabilities of 5G technology have brought rapid growth in numerous emerging applications such as virtual reality (VR), the Internet of Everything (IoE), immersive communication, artificial intelligence (AI), machine-to-machine (M2M) communication, etc. The large-scale commercial deployment and continuous progress of 5G mobile communications have made IoT application scenarios in smart cities more diversified, but it is now constantly revealing its limitations. For example, the upcoming level of virtual augmented reality (VAR), extended reality (XR), and holographic communication (HC) will demand super-high data rates (in the Tbps range) with extremely low latency that cannot be fulfilled by 5G communication even with the use of new frequency bands. Apart from limitations in data rate and latency, the demand for massive connection density to meet the demands of users in densely populated urban areas and industrial equipment will soon surpass the capacity for which 5G was originally designed. Among the devices that will contribute to the phenomenal growth in numbers will be the Internet of Everything (IoE) devices that will be deployed in smart city environments (as shown in Figure 1) for services like traffic management, telemedicine, remote patient monitoring, the Internet of Medical Things (IoMT), the Internet of Vehicles (IoV), environment monitoring, drone communications, Industrial automation, the Industrial Internet of Things (IIoT), logistics, transportation, etc.

Along with the surge in connection density in smart cities, there will be pressing requirements for more energy-efficient networks. As compared to 5G, the next-gen 6G networks are expected to be 100 times more energy-efficient. Additionally, there are a number of services and applications that demand global coverage, extending over ground, space, and underwater along with high-precision positioning and pervasive intelligence [10]. The worldwide data traffic generated from mobile devices is projected to surge by 55% each year through the period from 2020 to 2030 [11], and it is expected that in 2030, the monthly data traffic will surge to 5016 exabytes (EB)/month [12]. Also, subscriptions for mobile and M2M devices will increase to 17.1 billion and 97 billion, respectively, by 2030. The massive data handling capability, along with the abovementioned requirements particularly in the smart city paradigm, cannot be completely fulfilled by 5G. In order to address such limitations, researchers are now aiming to develop the sixth generation of mobile cellular systems. With the use of 6G technologies, smart cities can leverage a wider range of advanced applications, for example, in autonomous vehicles, e-healthcare, smart grids, and water and waste management. In addition, 6G can open the door to a host of new applications, such as immersive gaming experiences, augmented reality, holographic communications, and virtual reality.

The international mobile telecommunication draft IMT-2030 (6G) has defined six usage scenarios. Out of these six scenarios, three are extensions from IMT-2020 (5G), namely eMBB, mMTC and URLLC. The three additionally added usage scenarios are ubiquitous connectivity, integrated sensing and communication (ISAC), and AI and communication [13]. To support these 6G scenarios, non-terrestrial networks (NTNs), along with conventional terrestrial networks, will be required to enable extended coverage and mobility. UAVs can be equipped with onboard base stations (BSs) to provide mobile connectivity. They are easily deployable and have a strong line-of-sight links, which makes them highly suitable for SC applications such as in emergency deployments.

Additionally, AI, cloud computing, and mobile edge computing (MEC) will seamlessly provide 6G-enabled SSC. To address the challenges of centralized cloud-based scenarios, mobility-enhanced edge computing (MEEC) will become a vital element of 6G technology, allowing data to be processed locally and thus avoiding long-distance data transmissions, which improves latency, saves bandwidth and reduces security risks. Sixth-generation (6G) networks will be more human-centric [14], so a greater focus will be on privacy, security and confidentiality.

Along with this, additional requirements for ultra-low latency, super-high speed, and very-high reliability will be hallmarks of futuristic 6G networks. The sixth-generation (6G) networks will aim to offer performance that is 10–100 times better than that of 5G networks, e.g., user-experienced data rates in the range of 1–10 Gb/s, peak data rates of more than 1 Tb/s, connectivity to up to 107 devices/km², and over-the-air latency of 10–100 μs [15].

Table 1. Comparison of 5G and 6G (summary).

Parameter	5G	6G
Data Rate, Band	~20 Gbps, sub-6 GHz, Crowded	~1 TBPS, ultra-fast (THz)
Services	Limited capability to support new communication	Holographic communication, augmented reality, immersive gaming, etc.
Latency	Low latency	Ultra-low latency and high reliability
Architecture	Massive MIMO	Cell-free massive MIMO, intelligent surfaces
Coverage	Infrastructure-based	Ubiquitous connectivity (space–air–ground–sea)
Security	Security issues	Blockchain and quantum communication.
AI Integration	Partial	Full
Satellite Integration	No	Full

provides a concise comparison between 5G and 6G.

• Literature Survey and Review Process

In this work, we use the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method to retrieve relevant research records from databases. Later, a systematic literature review is performed to analyze these retrieved articles. The literature search method allows an unbiased review, which substantially impacts the results of a systematic review [16]. This is specifically important for upcoming technologies and research areas such as 6G and smart cities, where the digital transformation is happening at a rapid pace. Figure 2 displays a step-wise procedure for performing literature extraction from well-known scientific research databases, including IEEE Xplore, Web of Science (WoS), Taylor and Francis, Scopus, ASCE Library, and Springer. Precise search strings have been used to query the mentioned databases in order to extract applicable research records related to smart cities and 6G technologies.

• Step-1: Literature Retrieval

Carefully crafted search queries were framed as follows:

(“Artificial Intelligence” OR “THz” OR “ISAC” OR “Block Chain” OR “UAV” OR “RIS” OR “IoT”) AND (“6G”) AND (“Smart Cities”). These queries were aimed at retrieving papers published on the subject of 6G-enabled smart cities between 2019 and April 2024. We also included articles gathered from other sources. By combining the articles generated from the search string and other sources, we captured 1830 articles covering the research area concerned. After carrying out screening for titles and abstracts, removing duplicate and extraneous documents, and applying inclusion and exclusion criteria, 287 documents were finally considered for inclusion in the systematic review process.

• Step-2: Literature Filtering

Thorough reading of each paper was carried out in this step, particularly for the content given in the abstracts, introductions and final conclusions. The process worked towards selecting articles that matched with 6G potential technologies and smart city applications.

• Step-3: Classification

This step involved segregating the shortlisted papers into respective 6G technologies and smart city applications.

Table 2. Brief summary of parameters used in literature search.

Source Databases	IEEE Xplore, Web of Science (WoS), Taylor and Francis, ASCE Library, Scopus, and Springer
Search String	(“Artificial Intelligence” OR “THz” OR “ISAC” OR “Block Chain” OR “UAV”) AND (“6G”) AND (“Smart Cities”)
Time period	2019–2024
Article Type	Journal, Review, Letter, Book Chapter, Short Survey, Article
Language Restriction	English
Included Subject Area	Computer Science, Engineering, Energy, Business, Management and Accounting, Mathematics, Environmental Science, Decision Sciences
Excluded Subject Area	Chemical Engineering, Arts and Humanities, Health Professions, Agricultural and Biological Sciences, Neuroscience, Multidisciplinary, Psychology, Pharmacology, Toxicology and Pharmaceutics, Immunology and Microbiology, Nursing, Social Sciences, Economics Econometrics and Finance, Physics and Astronomy, Materials Science, Medicine, Biochemistry, Genetics and Molecular Biology, Chemistry, Earth and Planetary Sciences

briefly summarizes the criteria applied in shortlisting the available articles, including time period, search query strings, and subject areas included and excluded. Further,

Table 3. Summary of recent significant articles on 6G and the smart city context.

Ref.	Authors	Year of Public.	Research Area	Major Contribution
[17]	Fong, B et al.	2023	Vehicular	Investigates technical issues regarding the design and implementation of vehicle-to-infrastructure (V2I) systems to enhance reliability in a smart city with 6G as backbone.
[18]	P Mishra et al.	2023	IoT, Vision	Proposes framework, architecture and requirements for 6G IoT network. Discusses emerging technologies for 6G concerning artificial intelligence/machine learning, sensing networks, spectrum bands, and security.
[19]	Nahid Parvaresh, Burak Kantarci,	2023	UAV base station	Network performance of UAV-BS is improved by use of proposed continuous actor-critic deep reinforcement learning method to address the 3D location optimization issue of UAV-BSs in smart cities.
[20]	Z. Yang et al.	2023	Edge cloud, Energy efficiency	Paper analyzes challenges in developing a low-carbon smart city in 6G-enabled smart cities. Also proposes a visual end-edge-cloud architecture (E 2 C) that is AI-driven for attaining low carbon emission in smart cities.
[21]	N. Sehito et al.	2024	IRS, UAV, NOMA, Spectral efficiency	Paper introduces a new optimization scheme by utilizing IRSs in NOMA multi-UAV networks in 6G-enabled smart cities, resulting in significant performance enhancement in terms of spectral efficiency.
[22]	Prabhat Ranjan Singh et al.	2023	AI, Technology evolution, Smart city applications	Paper covers evolution of network technology, AI approaches for 6G systems, importance of AI in advanced network model development in 6G-enabled smart city applications.
[23]	Murroni, M et al.	2023	Vision, Enabling technologies	Paper furnishes an update on the smart city arena with the use of 6G. Paper describes the role of enabling technologies and their specific employment plans.
[24]	Kamruzzaman	2022	IoT, Energy efficiency, Use cases	Presents key technologies, their applications, and IoT technologies trends for energy-efficient 6G-enabled smart city. Also, identifies and discusses key enabling technologies.
[25]	Kim, N et al.	2024	Standardization and key enabling technologies	Paper provides key features and recent trends in standardization of smart city concept. Paper highlights potential key technologies of 6G that can be used in various urban use cases in 6G-enabled smart cities.
[26]	Ismail, L.; Buyya, R	2022	AI-enabled 6G smart cities	Discusses evolution of wireless-technology generations, AI implementation in 6G and its self-learning models in smart city applications.
[27]	Zakria Qadir et al.	2023	Survey, IoT	Emerging 6G connectivity solutions and their applications in IoT to serve smart cities are surveyed in this paper.
[28]	Misbah Shafi et al.	2024	6G technologies	The framework of 6G network is presented with its key technologies that have substantial effect on the key performance indicators of a wireless communication network.

summarizes recently published papers that have significantly contributed to the subject of our discussion.

On the basis of various aspects of urban life, the applications in a smart city can be classified into six domains, which are mobility, governance, economy, people, environment and living [29]. Applications under these domains can be further sub-categorized as shown in

Table 4. Possible classification of smart city applications (domains and sub-domains).

Natural Resources and Energy	Mobility and Transport	Living and Environment	People and Economy	Government
Smart Grid.	People Mobility.	Pollution Control.	Education and School.	e-Governance.
Public Lighting.	City Logistics.	Public Safety.	Entertainment and Culture.	Transparency.
Waste Management.		Health Care.	Entrepreneurship and Innovation.
Water Management		Public Spaces
		Welfare Services.
		Smart Homes.

.

The general layout of a possible smart city architecture [30,31], as shown in Figure 3, can be depicted in four layers: the sensing layer at the bottom, the transmission layer on the second level, the data management layer on the third level, and the application layer at the top.

Sensing Layer—This layer consists of physical components such as sensors and devices. These are used to collect large volumes of data from a diverse range of appliances/gadgets, such as smart watches, surveillance cameras, home appliances, vehicles, medical devices, etc.

Transmission Layer—This layer connects the physical devices with the management layer. The connectivity can be provided by wireless technologies (6G, 5G, or LTE) or using short-range links such as Bluetooth, Zigbee, M2M, etc.

Data Management Layer—This layer is responsible for data cleaning, data manipulation, data organization, its analysis, decision-making and storage.

Application Layer—This layer acts as an interface to provide services to citizens. The applications make decisions based on the data provided by the lower layers.

This paper attempts to review and discuss a comprehensive list of various potential 6G technologies in terms of requirements, architecture, visions, and usage, which are anticipated to be required in the implementation of future smart city setups.

Table 5. Summary of relevant works in the literature for 6G enabling technologies in smart cities.

Ref.	THz	AI	BC	QC	NTN (UAV)	MEC	RIS	ISAC	HC	VLC
[32]	√	√
[26]	√	√	√	√		√
[33]	√
[34]		√
[35]
[36]		√	√
[37]		√	√
[38]		√	√
[39]			√
[40]		√		√
[41]				√
[42]				√
[43]	√	√		√
[44]		√	√
[45]		√	√
[19]					√
[46]					√
[47]					√
[48]		√			√
[49]					√
[47]					√
[50]						√
[51]						√
[52]						√
[17]						√
[25]		√			√	√		√		√
[22]		√	√			√
[24]		√	√	√			√		√	√
This Paper	√	√	√	√	√	√	√	√	√	√

shows a comparison of 6G technologies covered in the available literature of the last 5 years. As can be seen from Table 5, this paper attempts a comprehensive discussion of potential 6G technologies that no other paper has covered so far. In addition to this and to the best of the knowledge of the authors, there has been no attempt to consolidate both the potential 6G technologies and their possible smart city applications in a single paper. Accordingly, there is a pressing need to furnish an exhaustive work that covers potential 6G technologies as well as their possible applications so as to give a complete and holistic perspective of the 6G-enabled smart city paradigm. Based on that, given below are the contributions of this research paper.

• Through this article, we have extensively covered in detail the potential 6G technologies in terms of requirements, architecture, visions, and usage, which are anticipated to be integrated in futuristic 6G-enabled smart cities.
• Secondly, we have discussed prominent smart city applications with underlying 6G technologies. This includes smart waste management, smart healthcare, smart grids, and more.
• Thirdly, potential challenges are also highlighted along with discussion on each technology and application. Also, at the end of this survey paper, challenges and suggestions for possible future research directions are also highlighted for the 6G-enabled smart city paradigm.

• Research Objectives

RO1—Identify the prime potential technologies for 6G wireless communication that are applicable in sustainable smart cities (SSCs).

RO2—Shed light on major urban use-cases with the implementation of potential 6G wireless technologies in the context of smart urban scenarios.

RQ3—List future scope of research.

• Structure of Paper

The survey structure is organized as shown in Figure 4. Section 1 provides the introduction, related literature discussion and recent advances in the field of 6G implementation and smart cities. Next in Section 2, fundamental 6G technologies used in the 6G-enabled smart city arena are discussed along with challenges and the future scope. In Section 3, smart city applications like smart grids, smart transport systems, smart manufacturing, etc., with 6G implementation are discussed. Finally, Section 4 concludes the paper with highlights for the future scope of research.

2. Potential 6G Enabling Technologies

The integration of advanced 6G communication technologies into smart city applications can bring about a revolutionary change in the development of intelligent smart city infrastructure. Furthermore, 6G-enabled technologies in the smart city can promote the inclusivity of citizens and elevate their quality of life by ensuring easy access to services, for example, transportation, healthcare, traffic management, clean water, energy management, and digital governance platforms.

In

Table 6. Potential 6G technologies and their brief descriptions.

Potential 6G Technology	Brief Description
Artificial Intelligence (AI)	AI can be used to analyze, manage and optimize resources and to efficiently support 6G networks. AI can be used for tasks like efficient channel estimation, energy efficiency, modulation recognition, data caching, traffic prediction, radio resource management, mobility management, etc.
Terahertz Communication (THz)	Uses frequency band 0.1 to 10 THz. Ability to attain ultra-high (up to 1 Tbps) data rates and wide bandwidth.
Blockchain (BC)	A type of distributed ledger technology to ensure safety, privacy, scalability and reliability in this complex heterogeneous architecture.
Quantum Computing (QC)	Based on quantum no-cloning theorem and the principle of uncertainty, absolute randomness is introduced by the use of the quantum nature of information, which provides security and enhanced channel capacity.
Non Terrestrial networks (NTN)	Includes drones and satellites and is used to extend coverage footprint of terrestrial base stations, provide additional capacity in dense urban hotspots. Used in disaster recovery and remote or rural areas.
Mobile Edge Communication (MEC)	By placing computing resources closer to end user, it reduces delays and latency and enhances processing speed and on-premise security
Integrated Sensing and Communication (ISAC)	Optimizes the allocation of scarce resources and contributes to better decision-making processes by combining both sensing and communication tasks, which enhances efficiency.
Reconfigurable Intelligent Surfaces (RISs)	A planar surface with array of passive elements whose characteristics can be altered dynamically. Used in 6G-THz to improve coverage, NLOS scenarios.
Holographic Communication (HC)	HC is an application used in transmitting human-sized immersive and interactive holograms consisting of 3D videos and images that require extremely high data rates with ultra-low latency.
Visible Light Communication (VLC)	VLC offers numerous advantages, such as, energy efficiency, cost-effectiveness, un-licensed spectrum, no electromagnetic interference, secure access technology, and large bandwidth.

below is a list, along with brief descriptions, of potential 6G technologies that can play a key role in delivering smart city applications.

In this section, we summarize fundamental enabling 6G technologies for implementation in smart city applications that are proposed in the recent literature. Table 5 lists relevant works in the literature and corresponding 6G technologies described in them. As can be clearly seen, our paper covers all potential 6G technologies.

2.1. Role of AI in 6G and Smart City Arena

Sixth-generation (6G) networks, in contrast to previous generations, will need to support a massive and diverse range of connected intelligent devices with very stringent requirements for ultra-low latency, broad frequency bands, very high data rates and high energy efficiency. These will also be the foremost requirements of any smart city implementation. All this will make the 6G networks more complex, dynamic, and diverse. The use of AI will enable 6G networks to become more agile, intelligent, flexible, and resilient in solving network challenges [53]. AI can be used to analyze, manage and optimize resources and to efficiently support 6G networks.

In last couple of decades, artificial intelligence (AI) has gained significant interest due to its potential to improve smart governance in cities [54]. For urban governance, AI-enabled policymaking and decision-making have the potential to promote efficient governance and services by using massive data collected from IoT systems comprising a diverse range of smart devices, such as sensors, security cameras, health sensors, vehicle sensors, GPS-enabled devices, home devices, etc. [55,56]. The data collected by IoT systems may be transmitted to the storage cloud, where big data platforms can be utilized in compiling the collected data. Further, these volumes of compiled data can be utilized to provide smart city services and make smart decisions [57]. Figure 5 illustrates a possible framework for AI implementation with IoT systems in a smart city scenario.

AI simulates the human brain’s capability to process the information used to develop theories and techniques, allowing machines to imitate the brain. ML is a sub-division of AI that performs predictions and helps in decision-making on the basis of data-learning [58]. Communication data can be analyzed by ML algorithms that can estimate signal loss in a given wireless environment. ML algorithms can also be implemented to design pilot signals and support efficient resource allocation [59].

To achieve the full potential of 6G, there is a need for the enhancement of prominent factors using intelligent and innovative techniques [60]. Machine learning techniques can be used for estimating and predicting the channel or detecting and decoding the symbols. Machine learning algorithms can be implemented in a wide variety of applications, such as beam selection, precoding, symbol identification, antenna selection, reduction of interference/noise, etc. The choice of using the best-suited machine learning algorithm depends on the given application, volume of training data, transferability of parameters between environments, etc. [61]. For an autonomous 6G network, the machine learning model should be able to adjust itself to any change in the topology, which means that the network should train itself using online continuous feedback without any human intervention. Learning model parameters are fine-tuned using repetitive online training to adjust to a changing network environment [62,63]. Online learning and training can be used to perform packet routing, channel assignments, uplink/downlink scheduling and other wireless activities. In recent years, the concept of AI radio access networks (AI-RAN) has gathered significant interest, as it has potential for predicting traffic patterns that can be used to optimize the allocation of network resources in real time [64]. This prevents over-provisioning and enables efficient load balancing and the use of resources, which reduces network resource costs. Also, it helps to minimize downtime and improve 6G network availability.

2.1.1. Applications

AI and machine learning (ML) can be utilized to enable efficient and robust interworking between multiple emerging 6G technologies, such as THz, RIS, NTN, ISAC, etc., so as to tap the full potential of these technologies, which can significantly enhance capacity, performance, reliability, energy efficiency, and more. In the last decade, there has been significant progress in the AI and machine learning (ML) arena, which has advanced the transformation of the traditional wireless system into an intelligent system that has the ability to address various challenges [65,66], as discussed below:

(a)

AI for integrating various wireless technologies

AI has the potential to enable efficient interworking between various upcoming wireless technologies and enhance performance. In [67], federated learning (FL) has been integrated with RIS to improve the performance of drone networks. The framework is used in RIS-enabled drone swarms to create a distributed wireless network. The reflection coefficients of each drone’s RIS are optimized by a learning process that enables the drone swarm to enhance coverage and quality.

(b)

AI for adapting to ever-changing wireless environments

With the integration of non-terrestrial communication in 6G, both the user and base station (UAV) are mobile, which makes the wireless communication environment even more dynamic. The UAV in this case needs to dynamically adjust its transmitted power and precoding to minimize interference and maintain coverage. The use of the conventional analytical and numerical approach may not be well-suited for such a dynamic environment [68]. AI-based methods [19] like reinforced learning and deep reinforced learning (DRL) may be utilized in such dynamically changing environments. However, even with classical AI learning models/algorithms, the computational complexity may grow in proportion with the dimension of the input data, for example, channel state information, which may increase the number of layers and iterations. This increases the requirement for training time and introduces latency. Quantum-based AI can potentially attain faster training convergence and has been shown to deliver more precise prediction [66].

(c)

AI for channel estimation

Due to requirements for ultra-low latency, massive speed, ultra-high density, and high reliability in smart city applications, various emerging 6G technologies, such as THz, intelligent surfaces, ultra-massive MIMO, VLC, NTN, etc., will be utilized. The use of these technologies will make the radio communication channel more complex. It will be challenging to perform channel estimation with traditional mathematical approaches. AI-based methods can be used to perform channel estimation since a limited amount of information is required to provide estimates. Precise channel estimation can be achieved using a deep learning-based estimation process, in which firstly, the effect of the channel on some signals sent along with the pilot signal is extracted. Then, by using DL methods, channel characteristics are estimated using interpolated channels [69]. In [70], it is shown that for a given channel model, DL methods can be used to learn transmitter and receiver implementations by depicting transmitter–channel–receiver as a single deep neural network that can be trained as an auto-encoder. The concept has been further extended to multiple transmitter–receiver pairs.

(d)

AI for handling scalability issues

Several smart city applications, such IoV, IoD, IoMT, etc., are extensively distributed in nature. In addition, smart city applications pose stringent requirements for extremely low latency, superior reliability, security, and high energy efficiency. These requirements can be potentially addressed by having decentralized compute and storage capabilities and placing them closer to IoT devices. It is possible to deploy a distributed or decentralized AI that can offload signaling and computational bottlenecks by having a distributed framework [71]. The traditional centralized ML model incurs high costs due to data gathering and processing [72]. Among the various proposed approaches, decentralized machine learning enables local processing of data, which not only distributes the computing capacities but also ensures data confidentiality, as data are not transmitted but are processed locally. Among the decentralized learning approaches, federated learning (FL) is being considered as an innovative idea under the ambit of the ML paradigm [73]. In FL, participating units are decentralized, and they process the data locally to train a global model. After the model is trained, only the changes in the model are transmitted to a centralized parameter server [74], thus avoiding the transfer of data. Although FL is being considered as a promising model, it has issues of its own. Firstly, it is considered expensive since it involves multiple stages of communication during generation of the training model and transmitting it to the central server. Secondly, the training process for the FL model is too sluggish to meet the real-time requirement of customers. Both are efficiency-related issues that can be addressed by reducing the communication overhead at both the algorithm level and system level.

(e)

AI for modulation recognition

In a noisy environment, techniques like deep learning and CNN can be used for modulation recognition, which helps in signal decoding and demodulation [75].

(f)

AI for traffic prediction

By predicting traffic based on historical data, the network capacities and computing resources can be proactively managed. DL-based methods have shown potential accuracy in predicting traffic on a real-time basis by using different RNNs [76].

(g)

AI for radio resource management

Due to the smaller coverage footprint in 6G-THz cells, the small-cell density will increase, which will result in increased co-channel interference. Therefore, the optimal use of radio resources and the spectrum will require system-level management. Instead of using traditional heuristic-based methods, an ML/DL-based approach is employed. In a typical implementation, a DNN frame with multiple layers can be used in the optimization scheme [77] for tuning a specific parameter, for example, power optimization or sub-channel allocation.

(h)

AI for mobility management

Various AI techniques, such as ANN or LSTM along with RL, can be utilized to learn and predict the movement of users so as to efficiently allocate resources and remove congestion [78].

(i)

Deep learning-enabled 6G physical layer (PHY) architecture

In 6G, we expect to see end-to-end (E2E) optimization of given communication networks in complex scenarios, for example, in ultra-high mobility conditions and with unknown channel models. We call this the PHY revolution [79]. E2E optimization along-with DL methods depicts a drastic change that will completely transform 6G networks. Four primary and creative 6G concepts in PHY are (a) massive MIMO systems, (b) RIS-empowered systems, (c) advanced multicarrier waveforms, and (d) physical layer security.

For signal detection in MIMO systems, a DL-based detector named DetNet is explored in [80,81]. Another DL-based model-driven detector approach in [82] is used in a MIMO system, which is established using an iterative detection algorithm, providing additional trainable parameters to enhance BER performance. In [83], a DNN architecture is proposed using a pure data-driven approach to detect symbols. DL-based detectors using data-driven [84] and model-driven [85] approaches are envisioned to reduce the complexity of MIMO system models.

By integrating both the DL-based detectors and channel estimators, a single algorithm can be constructed to achieve an E2E intelligent communication network. In [86], an intelligent detector and receiver network named DeepSM is proposed for spatial modulation (SM) in time-varying dynamic channels. In [87], another approach, named OAMP-Net2, is proposed for joint channel estimation and symbol detection based on model-driven DL-based approach.

In [88], two separate DL-based receiver techniques for an uplink multi-user MIMO system are proposed. The first technique, named FullCon, which is a data-driven DNN-based algorithm, directly handles the detection function for the received signal without performing channel estimation. In contrast, the second technique, named MdNet, separates the symbol detection and channel estimation phases.

Another topic of interest is the DL-based constellation design because the constellation shape highly affects a communication system’s performance. Similarly, DL-based approaches have also been applied for precoding in massive MIMO [89,90].

(j)

AI-based data caching

Due to the rapid growth in data-intensive smart city applications such as interactive gaming, infotainment, VR, etc., it is becoming increasingly challenging to manage surging backhaul traffic and data storage requirements. AI-based edge caching is seen as a promising solution to meet the storage requirements of IoT devices [91]. This allows IoT devices to fetch data from edge resources in real time while transmitting data on the backhaul link. However, in a dynamic network environment, deep learning techniques will be required to design an optimal caching strategy for optimal performance in terms of throughput, cache hit rate, offloading, etc.

(k)

AI for energy management

In a smart city setup, there will be demand for a huge diversity of smart low-power devices. In addition to this, various airborne and maritime devices/sensors will need to be highly energy efficient. Due to these factors, among others, 6G networks will need to devise robust energy management strategies in a smart city setup. AI techniques can be leveraged to aid these devices and infrastructures to enhance their energy management strategies by intelligently managing their power consumption and using energy harvesting techniques [92].

2.1.2. Challenges

Although AI in the 6G arena has several promising features, there are several open issues that need to be looked into.

(a)

High demand of servers at edge—Firstly, artificial intelligence tasks require heavy computational processes that run on task-customized servers. In the decentralized learning approach, most of these tasks are performed on servers located at the network’s edge. To minimize the demand of servers at the edge, more effective machine learning algorithms need to be explored.

(b)

Energy consumption—Energy consumption in 6G IoT devices needs to be reduced. One way to address this issue is by devising an efficient energy supply method with an innovative signal processing model. Moreover, AI and ML algorithms are complex and computationally intensive and consume a lot of energy. So there is also a need to develop algorithms that are more efficient and suitable for small IoT devices.

(c)

Large training data requirement—Acquisition of a large training dataset is a prime prerequisite for an enhanced deep learning system. But fetching large-scale training samples poses challenge owing to the high cost of resources and time required for training datasets.

(d)

Security and privacy—Smart city networks demand reexamining the cyber-security framework to enable enhanced security in wireless networks, especially when sensitive data are involved [93,94]. The possibility of vulnerabilities in networks has increased with the phenomenal surge in the number of IoT devices [95], so more research efforts will be required to make AI-enabled 6G networks more secure from malicious attacks.

2.2. Role of Integrated Sensing and Communication (ISAC) in Smart City Concept

Smart cities now have a diverse range of sensors, including IOT devices, environmental sensors, cameras, and other data collection points. The objective is to collect real-time data on a vast range of services such as traffic management, energy consumption, air quality, waste management, etc. All these applications also demand highly precise sensing capability, in addition to high-performance wireless connectivity. Sixth-generation networks will extend beyond just providing communication services and will deliver ubiquitous sensing features that will be capable of imaging or measuring surrounding objects. Speculative technological trends show that forthcoming B5G and 6G wireless systems will overwhelmingly utilize new sensing functionality. This sensing functionality, with the network’s ability to collect sensory information on the surrounding environment, can facilitate learning and create intelligence that can be applied in several location/environment-aware use cases. While the communication function is focused on transmitting information and retrieving it from noisy environments, the sensing function conversely uses noisy observations and retrieves meaningful information from them [96].

The technological evolution path of ISAC as discussed in ref. [97] can be depicted in four levels as shown in Figure 6. At the lower most level (stage-1), efficient sharing of spectral resources between radar and communication (R&C) systems is to be achieved, without creating mutual interference. At the second level (stage-2), the same hardware platform is to be shared between R&C functions. At the third level (stage-3), a common waveform will be utilized to fully reuse wireless resources between R&C using a single transmitting hardware platform and a unified signal processing framework. At the fourth level (stage-4), a perceptive network will be created by using a common networking infrastructure for both R&C function, to perform both sensing and communication tasks [97].

ISAC processes involve both sensing and communication tasks, thus optimizing and sharing the usage of scarce resources, which in turn enhance the efficiency of the services [98]. There are different methods to facilitate such sharing, including coexistence, cooperation, and co-design [99]. In ISAC, communications and sensing tasks are carried out by using common resources such as time slots, hardware modules, or frequency bands. This not only optimizes the allocated resources but also contributes to better decision-making processes. An extensive overview on ISAC technology in [100] from the signal processing viewpoint summarizes three distinctive implementations: the radar-centric design, communication-centric design, and joint design. In the communication-centric design, the communication waveform is modified to accommodate radar detection function also, but in this case, the communication functionality has higher priority. For a radar-centric design, the original radar waveform is altered in order to modulate communication information into a radar waveform. Communication information can be embedded into radar’s frequency-modulated signal waveform, by using step frequency, phase, pulse width, or into the radar’s spatial domain [101]. The third implementation is the joint design, which becomes a favorable strategy to accommodate both communication and sensing performance and to optimize the ISAC-transmitted signal.

In sensing-assisted communication, sensing can provide environment knowledge that can enhance communication systems functionality. Sensing can not only enhance the channel estimation accuracy but also substantially reduce the overhead since the channel estimation process for sensing-based channel acquisition need not be repeated for individual links. In another example, the sensing-assisted beam alignment uses the environment map and users’ location information provided by sensing to help the base station accordingly align and adjust the beam and its power to maximize the throughput.

ISAC, in combination with various advanced communication technologies, would be a potential solution to lowering the cost of communication and achieving higher energy efficiency and higher spectrum efficiency in future smart cities.

2.2.1. Applications

Below are a few applications using 6G ISAC in smart city scenarios:

(a)

Localization applications: There are certain applications that demand high-precision localization accuracy, for example, hazardous or critical targets that may involve human lives, like pedestrians [96,102]. For such localization-intensive applications in smart cities, a base station can utilize its joint sensing and communication resources. With higher bandwidth, the 6G ISAC system can support better tracking and localization performance, and for outdoor use cases, it can provide localization accuracy down to the centimeter level. For vehicular applications, when two or more entities approach each another, it is important to have high accuracy in relative localization. In applications related to automatic warehousing, centimeter-level accuracy facilitates device-level placement, enabling efficient and accurate placement of components that have a minute form factor. Under these circumstances, localization has assumed a central role in the ISAC-based smart city arena, and can be accomplished using various techniques such as RSS methods, time difference of arrival (TDOA), and angle of arrival (AOA) [103,104].

(b)

Gesture and activity recognition: Another ISAC-based application is in gesture and activity recognition in the smart city scenario. ISAC systems utilize high bandwidth, which enables high resolution and clarity in capturing gestures, and creates a wide range of applications using gesture and activity capturing techniques. This technique has a particular advantage over contemporary camera-based techniques in protecting privacy and personal information, as it limits the exposure of humans involved. This application can also be used as a security feature in intrusion detection. Another use case can be in smart hospitals to detect medical issues using respiratory sensing without the use of camera devices [105].

(c)

Augmented sensing: ISAC can also be used in augmented sensing, which uses devices to sense the environment, which otherwise would not have been possible using human capabilities. This is because ISAC offers high-resolution imaging capabilities along with communication capabilities that can be used in intelligent smart city infrastructure. It can be used in monitoring air quality and pollution, the detection of fine particulate matter (PM10, PM2.5), gas sensing, explosive detection, security scans of baggage, etc.

(d)

Imaging, mapping, and surrounding reconstruction

Additionally, 6G-based high-accuracy, super-resolution sensing affords a wide range of prospects in 3D imaging and mapping. This feature can be used in applications for navigation, scene reconstruction, spatial localization, updating knowledge of surrounding objects, etc. In urban areas, ISAC and densely distributed base stations have a longer sensing distance, greater field of view, and higher resolution, which can potentially enable 3D localization and environmental reconstruction. The reconstructed map thus generated can be used for applications such as smart traffic flow monitoring, accident detection, and queue detection, which are vital use cases in the smart city setting.

2.2.2. Challenges

Although ISAC has been well investigated from various aspects in recent years, there are still several open challenges and areas that remain unexplored. Below are a few of the open problems:

(a)

Waveform design: Developing a new waveform design that can meet the requirement of both communication and sensing functions [106].

(b)

Performance tradeoff: Another significant challenge for ISAC is the beamforming due to conflicting requirements of communication and sensing systems. Whereas the communication systems prefer a narrowband beam, the sensing systems conversely favor a wideband arrangement so as to capture wide environmental information [107].

(c)

Clock synchronization requirements: For sensing functionality, the synchronization requirements are significantly more stringent than for communications. This is due to the resultant timing offset that may cause interference with the sensing functionality. In order to prevent phase-offset, ISAC systems need accurate clock synchronization between the receiver and transmitter [108].

(d)

Channel modeling and evaluation methodology: The channel model in 6G brings significant challenges, as it needs to consider both communication and sensing services. For example, the echo channel, which is one typical sensing channel, contains backscattering RCS characteristics generated from the surrounding objects. Such propagation channel types create new requirements that are currently not supported in prevailing communication channel models. Consequently, the prevailing channel modeling methodologies may need some rework and innovation.

Another challenge is in the evaluation of performance metrics that will be in accordance with new sensing requirements. As of now, any communication systems latency, throughput, and reliability are considered the main performance evaluation metrics. But for sensing applications, a new dimension of evaluation metrics, including detection probability, sensing resolution, accuracy, and update rate, needs to be considered.

2.3. IoT for Smart Cities with 6G

IoT seeks to establish a connected environment by linking a diverse range of electronic devices to the internet, where data are sensed, processed, analyzed and communicated, and all this happens automatically without human intervention [109]. A massive number of sensors are deployed to capture the details of the real world. This sensed dataset is processed by utilizing heavy computing resources to obtain high-level information and decisions that are fed back to the physical world to make it better. The architecture of IoT/CPS can be visualized in a three-layer structure [110,111], with sensors and devices placed in the bottom layer, the middleware in the center, and the networked applications at the top (Figure 7). The middle layer executes virtualization of the physical world into virtual space. The top-most layer performs intelligent tasks, such as delivering knowledge, performing prediction tasks, determining optimization requirements, etc., that are fed back for improvements to the actual world.

IoT can be applied in several domains, such as agriculture, transport, health care, household devices, the manufacturing industry, waste management, weather monitoring, etc., and 6G will play a significant role in the integration of functions such as sensing, analyzing, processing, and computation.

2.3.1. Characteristics of 6G-IoT

To support the futuristic demands of smart city applications in the 6G era, the IoT scenario will have key differences compared to earlier implementations, as described below:

(a)

Extremely massive numbers: The number of IoT devices are expected to show massive growth in the future. Some of the reasons are as follows: Firstly, several new device types are expected to emerge, such as extended reality smart devices, to deliver very high performance. Secondly, large number of sensors will be required in applications related to autonomous vehicles, industrial automation, e-health, transportation, etc. Thirdly, the advent of novel applications such as holographic communications will require high-end devices.

(b)

More sensitive and intensive: With the introduction of new applications such as immersive communication, virtual reality, augmented reality, etc., the futuristic 6G-enabled IoT will be data-exhaustive, compute-heavy, and security/privacy vulnerable.

(c)

Sensing and localization dependent: With applications like autonomous vehicles and holographic communication, there is a requirement to utilize sensing and localization functionality to map surrounding objects precisely. So, in addition to earlier communication, computing and caching functions, IoT will be expected to handle sensing and localization tasks as well [112].

(d)

Real-time communication: For applications such as in remote surgery, autonomous vehicles, industrial automation, etc., IoT will need to transmit data in real time with ultra-low latency.

(e)

Energy efficient: With the deployment of large-scale IoT, energy efficiency will be a dominant design metric for IoT-based 6G networks. In addition, focus will also be on employing energy harvesting techniques in the IoT setting.

2.3.2. Classification of IoT

Several IoT classifications are available in the literature, broadly based on connected object types and the types of transmitted data [113]. Some of the IoT types are as discussed below:

(a)

Internet of Robotic Things (IoRT): With immense progress in robotics technology and IoT, the combined use of both can considerably enhance automation and reduce cost. When integrated with robots, IoT devices augment robot functionality by fetching real-time sensing and monitoring data. Further, this sensed dataset can also be intelligently analyzed so that IoRT can attain efficient performance. IoRT can be utilized in applications related to warehousing, industrial automation, smart homes, etc.

(b)

Internet of Medical Things (IoMT): IoMT allows monitoring of patients remotely by using telehealth for screening and treatment and self-evaluation of health using smart wearables.

(c)

Industrial Internet of Things (IIoT): IIoT is utilized specifically for managing machine type communications for industrial systems. With the upcoming 6G technology, new emerging applications such as digital twins will demand 6G-enabled IIoT to provide super-high data rates. Similarly, industrial multi-robots using the 6G-IIoT system will require precise navigation, which makes localization and sensing a crucial KPI [114].

(d)

Internet of Everything (IoE): IoE has broadened the concept of IoT and includes networked environments of objects, files, humans and processes, as shown in Figure 8 [115]. It is projected by many that in the future, 99.4% of objects will be covered by the IoE concept [94]. IoE sensors can capture many parameters, such as temperature, pressure, bio-signals, velocity, etc., that are collected from massive numbers of intelligent devices used in industrial applications, health care, smart cities, etc. [116]. The sharp growth in the number of intelligent heterogeneous IoE devices, with each requiring massive data with low latency, creates a major hurdle for 5G networks. Despite these requirements, 6G networks can enable a smooth transition from IoT to IoE [117].

(e)

Internet of Space Things (IoST): IoST is considered to be an extension of IoT. Since there are limitations to the extent of area under terrestrial coverage, communication with remote areas that cannot otherwise be covered due to economic reasons can be made possible through the use of satellite-based communication harnessing the high speed, low latency and high bandwidth of 6G mobile technology [118].

(f)

Internet of Nano Things (IoNT): IoNT is the next-generation IoT, comprising minute nano-sized IoT devices with sizes varying from 0.1 to 100 nm [119]. Depending on the use case, these devices can be of various types, for example, nanomachines, nanorobots, nano-sensors, etc., which can transfer data amongst themselves or with external systems. These devices will find possible applications in healthcare, industrial processes, the monitoring of toxic gases, agriculture, etc. For specific healthcare applications, the Internet of Bio-Nano-Things (IoBNT) will use nano-sensors inside the human body to collect data for medical applications.

2.4. Blockchain (BC) and 6G-Enabled Smart Cities

Smart cities have massively interconnected IoT devices and sensors that form a large-scale network to exchange data, enable the monitoring and management of city infrastructure, and thus improve the living standards of citizens. The 6G-enabled smart city architecture will be increasingly decentralized and softwarized, so it is of the utmost importance to ensure a robust mechanism for safety and privacy in this complex heterogeneous architecture. The coalition of Blockchain and 6G-enabled smart city applications [120] can solve these concerns by establishing trust in distributed platforms. Blockchain, a type of distributed ledger technology, was originally designed for cryptocurrency, but now, it finds applications in wireless communications [121], Internet of Vehicles (IoV) [122,123], smart grids [124], and smart manufacturing [125].

Blockchain works by using a distributed ledger model. This allows a decentralized, unalterable arrangement so that parties can independently transact and validate the data on the ledger, and this takes place without any requirement of intermediaries. Blocks are interconnected in a blockchain by the use of smart contracts that are self-executable and rely on cryptographic protocols [126]. Records, once accepted and added in the distributed ledger, cannot be deleted or altered and are immutable, thus ensuring data integrity and security. The blockchain structure maintains both traceability anonymity, which is accomplished by interconnecting different blocks by the use of hashes. Collections of transactions are arranged in each block in a Merkle tree format [127]. The Merkle tree allows authenticity against the known origin by efficient validation and verification of each transaction, thus facilitating traceability in the blockchain.

Smart city applications are related to citizens’ wellbeing, so any cybersecurity vulnerability, for example, Denial-of-Service (DoS), alerting, message tampering, spoofing, malware, and eavesdropping attacks, can severely compromise the integrity of communicated data, resulting in hazardous situations.

Similarly, user privacy and confidentiality are also supreme. Due to the heterogeneous and complex nature of 6G networks, ensuring confidentiality becomes challenging, more so when other technologies such as AI, the cloud, etc., are also deployed alongside. Anonymization techniques are also utilized to protect the user data by truncating personally identifiable information (PII) before data are transmitted over the network. The use of Blockchain in smart cities would enable greater controlled governance since the intermediaries will not control the data [128]. Blockchain also allows the option of encryption and traceability of transmitted data, while maintaining anonymity at the same time. Blockchain helps to achieve smart city objectives, for example, secure and fast data exchange, the reduction of fraud, regulatory processes and audits, and the management of energy consumption. Below, shown in

Table 7. Blockchain applications depicted in recent papers.

Ref.	Year	Application Domain of Smart Cities	Technologies Used	Areas/Topics Covered
[123]	2024	V2X	6G, Blockchain, Federated learning, Fog Computing	Comprehensive V2X security analysis. Future research direction for privacy in XR, secure SDN, physical layer security in THz.
[129]	2024	Smart Traffic Management	Edge Computing, Blockchain, Reinforced learning	Traffic optimization is achieved by decentralized integration of IoT sensors on vehicles and traffic signals and edge devices and the use of BC rules for real-time decisions.
[130]	2024	Supply Chain Management	Blockchain, IoT, Edge Computing	A Blockchain-based and IoT-enabled transparent and secure supply chain management framework is proposed for public emergency services in smart cities.
[131]	2023	Intelligent Transport System (ITS)	Blockchain	An ITS cross-domain data interaction framework between devices and agencies is proposed to achieve secure and efficient cross-chain communication.
[36]	2023	IoT	Blockchain, Big Data, AI	Framework and architecture based on Blockchain, AI and Big Data.
[39]	2023	Industrial Applications	6G, Blockchain, IoT	Case study of smart supply chain. Benefits and challenges of BT and 6G-IoT
[132]	2023	IoD (Internet of Drones)	6G, Blockchain	Analysis of multilayered Blockchain-IoD novel Global Compliance System (GCoS) and Swarm Security (Sse) system
[133]	2023	IoT-Blockchain efficiency	6G, IoT-oriented Blockchain	Improves Blockchain-IoT performance by targeted optimization to improve low power efficiency and slow ledger synchronization.
[134]	2022	IoV	6G, Blockchain	A survey paper for BC in IoVs sharing underlying 6G technology. Explores how privacy and security issues in IoVs can be tackled using BC technology.
[135]	2022	Food Supply Chain Management	IoT, Blockchain	Blockchain enables traceability of food supply from factories/fields to the customer’s table. IoT devices probe food condition.

, are some examples of Blockchain implementation in smart city applications in the recent literature.

Challenges

Despite the large number of benefits, integrating Blockchain into the smart city also has some challenges that need to be overcome:

(a)

Blockchain substantially increases requirements for computational and storage capacity. This becomes a pressing issue in a smart city environment that has large number of devices generating voluminous data. With Blockchain integration, there is an increase in system complexity and also a substantial increase in the signaling overhead.

(b)

Energy consumption by IoT devices will increase due to increased computation. Developing energy-efficient algorithms, leveraging technologies like edge computing, and optimization of resource utilization can help mitigate this challenge.

(c)

Industry-wide protocols and standards will be required for seamless interoperability between Blockchain and a diverse range of 6G-enabled smart city IoT devices and sensors.

2.5. Terahertz (THz) Communication

Owing to its ability to attain ultra-high (up to 1Tbps) data rates and wide bandwidth, Terahertz is regarded as the key enabler of 6G technology [136,137]. Two main reasons behind the need for the THz band for 6G are as follows: Firstly, in the sub-6 GHz band, spectrum scarcity and non-continuity present a significant challenge in attaining higher data rates. Secondly, the mmWave technologies below 100 GHz provide insufficient bandwidth [138]. In THz, the availability of large, accessible bandwidth ranging from tens up to hundreds of gigahertz, affords the potential for delivering enormously high data rates that overcome the capacity constraints of 5G networks.

The THz band is situated in between the mmWave and infrared frequency bands and uses the frequency band between 0.1 and 10 THz, for an equivalent wavelength range of 0.03 mm to 3 mm [139]. For 6G usage, the Federal Communications Commission (FCC) has considered the frequency band between 95 GHz and 3 THz [14]. Communication using THz frequencies requires the use of smaller cells that can have a coverage footprint of only a few tens of yards [140]. To ensure efficient beamforming at this higher frequency range, massive arrays of antennas will be required for the THz band [141]. THz communication can be utilized along with reconfigurable intelligent surfaces (RISs) to achieve NLOS propagation and enhance the coverage footprint [142,143]. RISs can be made to smartly steer the incident beam of radio waves in desired directions by performing phase shifts using a smart controller. This enables an RIS to convert a LOS THz link into a virtual LOS link. Thus, the use of RISs mitigates the blockage sensitivity of THz signals, which improves THz communication system reliability.

Sixth-generation networks using THz frequencies are envisioned to deliver centimeter-level precision [144], which conventional GPS techniques are unable to provide. High-frequency THz localization techniques utilize the concept known as simultaneous localization and mapping (SLAM), whereby the location accuracy is improved through the fusion of high-resolution surrounding images with the estimated range of the user. Additionally, THz signals have also been conventionally utilized in wireless sensing and imaging in various applications, for example, in food safety, quality control, and security.

2.5.1. Applications/Use Cases

Below are descriptions of a few use cases for the THz band that can also potentially find applications in smart cities.

(a)

V2X communications: The road safety applications in smart cities have stringent requirements for the ultrafast, low-latency, and ultra-reliable exchange of data. These stringent requirements can be potentially fulfilled by implementing the THz band in 6G V2X communications. Highly directed THz links may be utilized for data exchange among vehicles in fleets of autonomous vehicles executing cooperative driving.

(b)

Nano-networks in the medical domain: Due to the very small wavelengths, antennas with tiny dimensions can be designed, paving way for novel applications, for example, in nanoscale communications for nanomachines, the Internet of Nano-Things (IoNT), and on-chip communications [145]. For in vivo body-centric monitoring, THz waves can support nano-communication [146] and also enable unique advantages for imaging use cases.

(c)

Short-range communications: The THz band can be used for close-proximity communications involving LoS and a short distance (<1 m), which include applications like kiosk downloads for scenarios like high-data rate file exchange between kiosks and user devices found in places such as train stations, airports, shopping malls, etc.

(d)

High-speed train (HST) communication: High-speed broadband communication access for HST users is anticipated to be implemented by the next generation of mobile communications. 6G along with THz are envisioned to support “smart rail mobility” that demands seamless wireless connectivity along with ultra-high data rates for five scenarios: train-to-infrastructure (T2I), train-to-train (T2T), inside station, intra-wagon, and infrastructure-to-infrastructure (I2I) [147].

(e)

WLAN: The THz band can be used in typical indoor communications scenarios such as conference rooms and office rooms. But for such indoor settings, there is a need to consider LoS, scattering, and link blockage. For these reasons, it will be mainly used to interconnect access points and enable user access with fixed locations.

(f)

THz in health care: THz radiation has several spectral features, such as the following: it is non-ionizing, non-invasive, offers spectral fingerprinting, good resolution of <1 mm, polar substance phase sensitivity, coherent detection, and penetration capabilities that make it a promising technology for spectroscopy, sensing, and imaging in healthcare applications [148,149].

(g)

Terahertz imaging: The use of THz in imaging has many technical benefits over visible light and microwaves. Due to its small wavelength and wide bandwidth, THz imaging provides high spatial resolution [150]. Additionally, THz exhibits better penetration performance as compared to visible light and infrared. In the smart city context, there can be several security screening applications at critical locations such as airports, banks, border crossings, etc., to perform imaging of parcels for any hidden objectional objects.

2.5.2. Challenges

(a)

High propagation loss, atmospheric attenuation and blockage and absorption [151].

(b)

The cell size of THz communication will be small, so network architecture with higher cell density will be required.

(c)

Further research on accurate channel modeling and its characteristics for propagation scenarios containing ground, space, and nano-scale communication [151].

(d)

Hardware challenges due to high frequency in the development of THz antennas, amplifiers, and mixers, including the challenge of high bandwidth design of super heterodyne transceivers [152].

2.6. Quantum Communication (QC)

In the quest to meet the rapidly surging demand for intelligent, fast, reliable, secure, and green communications, the urge for higher systems computational power has also risen rapidly. The inherent parallelism provided by the basic concepts of quantum mechanics and its potential as illustrated via recent results for QC technology clearly reveal definite prospects to outperform traditional computing systems. Quantum communication (QC) uses the quantum key technique, which is based on the quantum no-cloning theorem and the principle of uncertainty. Absolute randomness is introduced by the use of the quantum nature of information, which in turn provides security [153]. QC-assisted communications is another new research area considered to hold potential for attaining ultra-high data rates and ensuring link security in upcoming 6G. By using quantum communication in 6G wireless networks, a strong security communication system will be established. QC has the ability to detect eavesdropping [154]. In addition to providing protection against cyber-attacks, quantum communication has the ability to achieve an extremely high data rate [153]. In 6G, quantum- and ML-based cyber-security encryption can be used in communication links, which may further enhance security and privacy. Quantum key distribution (QKD), which is a quantum communication protocol, uses quantum mechanics for establishing a key between two communicating parties [155]. QKD has potential for use in 6G undersea communication, satellite-based communication, and THz communication systems. Quantum communication, when used in futuristic 6G communication networks, can increase the channel capacity and facilitate quantum cryptography and quantum teleportation [153].

In the communications and computing domains, the currently used protocols can be improved with more effective algorithms by utilizing the physical phenomena of the quantum world in quantum principles and tools. Quantum principles can yield significant advantages in communication networks, including increased channel capacity and potential for transmitting unknown quantum states (quantum teleportation) and providing secure information (quantum cryptography), which will otherwise be impossible with traditional techniques [153].

In contrast to the traditional binary-based communications networks, quantum communications has the potential to yield absolute randomness, security, higher information-carrying capacity, and considerable improvements in transmission quality. Additionally, quantum-based techniques have ultra-fast task execution potential that is beyond the classical systems’ capability [156]. To broaden the prospects for integrating traditional data transmissions into a system that is quantum-based, a few quantum communication protocols are available, which are quantum teleportation [157], quantum key distribution (QKD) [158,159], and dense coding [160]. Multiple access techniques, such as wavelength division and frequency division, can be used in quantum networks, just as in the case of classical communication systems. For other multiple access methods, photon orbital angular momentum and coherent states are utilized in QC networks. The use of quantum communication can substantially enhance the energy efficiency, spectral efficiency, and security of 6G wireless systems [161].

2.6.1. Applications

Some of the major applications of QC in 6G are discussed below:

(a)

Quantum algorithms in 6G

Complex optimization requirements in 6G networks that otherwise are difficult to handle with classical algorithms can be efficiently resolved by using quantum algorithms. For network design of 6G systems, the use of quantum computing algorithms can assist in managing complicated optimization issues related to user scheduling, resource allocation, network topology design, data detection, routing, and beamforming design [162]. In ref. [163], encouraging results were shown by the use of quantum algorithms to strengthen MIMO data detection. A few more-recent papers have also analyzed quantum-inspired algorithms and quantum-annealing algorithms for near-optimal MIMO data detection [164].

(b)

Quantum computing and machine learning in 6G

In wireless communication networks, combinations of quantum computing and machine learning (QML) are regarded as key 6G enablers [153]. Regarding its projected potential with new findings on quantum mechanics, for example, more qubits integration, inherent parallelism, and quantum algorithms, quantum computing substantially outperforms traditional computing systems when compared in terms of computational capability. Accordingly, joint machine learning and quantum computing provides a sturdy solution by exploiting their joint benefits in wireless systems deployment. The entanglement concept, parallelism concepts of qubit, and superposition can manage voluminous data with large-dimensional vectors and produce statistical patterns of data for machine learning methods [162].

(c)

Quantum sensing in 6G

Due to the distinctive characteristics of quantum mechanics, such as squeezing and entanglement [165], high-accuracy measurements for quantum sensing can be achieved, which otherwise are not feasible using classical measurement methods. Quantum sensing has the capability to considerably enhance 6G wireless network performance in terms of enhanced localization, timing synchronization, and navigation functions [166].

(d)

Quantum Blockchain in 6G

Due to certain limitations in classical blockchain technology, such as low transaction speeds, privacy breaches, and security threats [167], quantum-secured blockchain technologies might be required in upcoming 6G networks that will deliver the security ensured by post-quantum cryptography and QKD protocols [168]. Additionally, to design an effective blockchain transaction algorithm, entanglement can be utilized as a technique to decrease the communication overhead, in turn also increasing transaction speed [166].

2.6.2. Challenges

While 6G networks are expected to materialize within this decade, quantum communication and quantum internet at this point are still in the nascent stages. To realize quantum communication practically in future 6G communications systems, extensive research work is still required. Quantum communications present two main issues in terms of designing and developing new communication protocols.

(a)

The first main challenge is its implementation in building quantum internet with network entities, such as, quantum routers, switches and repeaters, which, due to no-cloning theorem, is difficult [169].

(b)

Measures of capacity are not fully understood for quantum communication channels. This is because there can be different possibilities with regard to delivering information, including entanglement-assisted information and quantum information [170,171].

(c)

A major milestone to conceive quantum internet and quantum communication will be the establishment of quantum channels over long distances, which can be established by the physical platform of photons. But due to the loss of photons, which is a known research problem, quantum repeaters will be utilized to address this drawback.

2.7. Immersive Communication (IC)

Immersive communication and its supporting technologies enable its users to interact via 3D audio-visual means and obtain real-life like experiences in the virtual world, physical world, or both. The use of immersive communication can find many potential applications in the smart city scenario. But even with the substantial advancement in network capabilities and roll-out of 5G systems, immersive communication still has many challenges in several aspects of networking, communications, and computing. Researchers are expecting breakthroughs in the realization of full-scale immersive communication in the upcoming 6G era.

2.7.1. Types of Immersive Communication

There are three main types of immersive communication, i.e., (a) XR, (b) haptic communication, and (c) holographic communication.

(a)

Extended reality (XR): XR includes technologies like virtual reality (VR), augmented reality (AR), and mixed reality (MR). Through the use of XR devices, users in the physical world can interact with their virtual avatar. Depending on the scale of virtuality, XR can be categorized as AR and VR. AR, with a lower scale of virtuality, deals in creating objects in the virtual world that are similar to those available in the physical world. VR uses a higher level of virtuality by creating entire virtual scenery. The XR implementation process can be broadly outlined in three steps: (a) content transmission, (b) rendering, and (c) feedback collection. In content transmission, VR content generated from content servers of VR content providers is transmitted. Transmitting ultra-high 12 K resolution videos from content servers demands ultra-high data rates. In the second step of content rendering, transmitted VR videos are stitched together for VR devices and projected as a 3D stereoscopic video. After playing the video, cameras or sensors available at the user end capture the users’ actions and motions. For accurate and smooth content delivery based on feedback on user motions, there are stringent latency and high data-rate requirements in XR. Also, high computing capabilities are required of both user devices and network servers for the smooth operation of interactive XR applications.

(b)

Haptic communication: Haptics refer to interactions of the form dealing with the sense of touch. In the first step, haptic sensors acquire haptic information that includes tactile and kinesthetic information. In the second step, the volume of haptic data is reduced without degrading the user’s immersive experience. In the third step, haptic data are transmitted over the communication network, creating a haptic data-stream. In the final step, when the haptic data are received by the haptic interface (HI) receiver, realistic haptic sensations are created by haptic actuators [172].

(c)

Holographic communication (HC): HC is a further evolution of immersive 3D images and video that demands extremely high data rates with ultra-low latency to transmit a life-size interactive hologram. With the considerable progress in holographic display technology in the last few years, for example, Microsoft’s HoloLens [173], this application will become a reality in the coming years. The massive data rates required, even after performing data compression, may range from a few tens of Mbps to 4.3 Tbps [174]. Similarly, the latency requirement for a seamless 3D holographic experience is less than sub-milliseconds. In order to create a reconstruction of a near-real hologram, HC requires the use of multiple cameras and sensors [175] to capture the image. The captured images are compressed and transmitted on a communication channel. On the received end, the images are decompressed and then projected using laser beams. In 6G, motion and audio data along with images will also be transmitted. Along with high data rates, high computational power is also required for such applications. In the future, holographic communication will thus provide real-time, in-person communicating experiences, enabling virtual meetings, virtual concerts, and more-interactive remote education.

2.7.2. Use Cases for Immersive Communication

In the smart city context, immersive communication can have several potential use cases, some of which are given in

Table 8. Potential use cases for IC.

Use Case	Description
Remote Surgery	▪ Remotely placed robotic arms are used by surgeons to operate. ▪ Robotic arms can cancel the surgeon’s hand tremors [176]. ▪ Helpful in delicate surgical procedures with minimal insertion [177]. ▪ Surgeons use 3D-video with depth information. ▪ AR can be used to superimpose ultrasound and CT scan images [178]. ▪ Low-latency, high-reliability, high-speed requirements.
Holographic Teleconferencing	▪ HC enables life-like 3D projected images as holograms of remote users [179]. ▪ Multiple sensors capture the user’s 3D visual and associated audio information, which is transmitted and then reconstructed at the receiver end as a hologram to project a 3D audio-visual display for interactions [175]. ▪ Along with AV information, haptic information can also be added to obtain a sense of physical contact along with the hologram [172].
Immersive Gaming	▪ Interaction among players is in a life-like environment without sensing the barrier between the virtual and physical worlds [180]. ▪ Player is displayed in the game in the virtual world using XR devices, and actions are captured to enable interaction with the virtual world.
Metaverse	▪ Users can have their digital representations or digital avatars in virtual and imaginary worlds in games or virtual cities. ▪ User activity in the physical world is captured by sensors and actuators and reflected through their avatar in the virtual world. ▪ To enhance human-centric immersive social experiences, a metaverse can leverage AI, blockchain, big-data analysis, and advanced networking technologies [181].

.

2.8. Visible Light Communication (VLC)

VLC offers numerous advantages, such as energy efficiency, cost-effectiveness, an un-licensed spectrum, no electromagnetic interference, secure access technology, and large bandwidth [182,183]. For VLC with a wavelength range of 380–780 nm, high data throughput of the order of Gbits/s is achieved due to the large bandwidth, which is about 1000 times larger than that of RF transmissions [184]. VLC has gathered tremendous interest among researchers who aim to exploit the use of ordinary LEDs to enable high-speed, pervasive and reliable wireless communications [185,186]. Recent studies using fast LED sources and receivers, has shown very high data rates varying between several hundred Mbps to Gbps [187], although most of these studies have been conducted in lab-controlled environments. Many investigations of VLC implementation in 5G are also available in the literature [188,189]. Although VLC systems offer remarkable advantages, there are many implementation hurdles, the foremost being the limited 3 dB bandwidth of LED, which reaches only a few MHz, resulting in poor throughput [190]. To overcome this limitation, several techniques have been proposed, among which are the use of laser diodes as VLC sources, the use of blue-optical filters at the receiver end, or the use of white-light phosphor laser [191].

2.8.1. Free-Space Optics (FSO)

Recently, FSO systems have also emerged as appealing technology applications in metropolitan networks such as backhauls for wireless systems, inter-building communication, fiber backup, indoor links, satellite communications, and also for military purposes [192]. FSO transmission demands stringent alignment tolerances and suffers severe attenuation of the light beam due to impairments caused by weather conditions such as rain, fog, turbulence, haze, and dust [193]. Additionally, when FSO is installed on skyscrapers, the building sway causes misalignment losses, which can be addressed by using an automatic pointing and tracking system [194].

2.8.2. Fiber-Wireless System (FiWi)

The evolution of RAN architecture is leading to the convergence of radio and fiber system interfaces known as fiber-wireless (FiWi) systems [195]. For an optical-wireless convergence network, the cloud-RAN (C-RAN) deployment is very important. In traditional distributed-RAN (D-RAN) networks, the baseband units (BBUs) and remote radio Units (RRUs) are co-located at a Base Station (BS). However, in a C-RAN setup as shown in Figure 8, the baseband processing is centralized by moving it to a Central Office (CO), and the RRU remains at the antenna location [196]. The centralized architecture brings several advantages, such as operational and management (O&M) simplification, infrastructure reuse, multiple technologies coexistence, reduced energy consumption, and reduced operating expense (OPEX) and capital expense (CAPEX). Fiber-radio integration in a FiWi system is achieved by implementing the radio-over-fiber (RoF) technology, which is further classified into two types, namely, analog radio over fiber (A-RoF) and digital RoF (D-RoF), with both schemes having benefits and drawbacks of their own. The fiber/wireless system shown in Figure 9 uses the C-RAN architecture, in which the implementation of a front-hall (FH) link can be achieved using different types of technologies, such as, analog radio-over-fiber (A-RoF), digital RoF (D-RoF), radio link, or by using free-space optics (FSO). Multiple COs are connected to the core network using a backhaul (BH) optical link.

2.8.3. Power over Fiber (PoF)

Using PoF technology, electrical power can be transported over an optical fiber link, which provides immunity from magnetic fields, creates superb electrical isolation, and increases reliability and safety. PoF systems consist of three main units: (1) a high-power laser diode (HPLD) for producing optical power at the transmitter, (2) fiber-optic cable used as transmission medium, and (3) a photovoltaic power converter (PPC) at the receiver, which converts optical to electrical power. By using PoF, the simultaneous transmission of data and power is possible, with the delivered power levels now having reached over 40 W [197]. With the tremendous increase in cell densification in 6G, it will be important to find solutions to provide a consistent power supply for ensuring network stability. PoF can play a pivotal role in meeting both the power and communication requirements of future base stations. However, there are a few drawbacks as well. Firstly, due to the use of shorter wavelengths in PoF systems (e.g., 808 nm and 980 nm), the feasibility of long-distance front-haul is not available due to the higher attenuation loss, which reduces the power transmission distance to a few kilometers. Secondly, the PoF system’s power transmission efficiency (PTE) and maximum power delivered are low, limiting it to use in powering small cells that require low power. But it is expected that when the technology matures in the future, these parameters can be improved to meet the power requirements.

2.8.4. Challenges

Currently, VLC technology is considered to be in the research phase. There are several open issues, some of which are highlighted below:

(a)

Further work is required for VLC Integration with currently operating communication protocols like Wi-Fi, etc.

(b)

The bandwidth of presently used LED is low, less sensitive as receiver, and also has non-linear performance. Laser diodes require strict transmitter–receiver alignment. VLC systems for 6G networks will require new, enhanced optical hardware that can support wide bandwidth and high sensitivity and that, at the same time, is inexpensive.

(c)

To reduce the impact of noise introduced by ambient light, new channel coding schemes needs to be looked into for more effective reduction of noise than possible with currently used techniques.

(d)

For VLC system design, there is a need for a recognized and realistic channel model considering mobility, indoor/outdoor, dimming control, LOS/NLOS scenarios, etc.

2.9. Mobile Edge Computing (MEC)

The prime purpose of 6G networks in the smart city context is to provide ultra-high speed, support ultra-massive density of users, provide ultra-high bandwidth, and ensure reliability, energy efficiency and security. Sixth-generation networks are intended to seamlessly and fully integrate various applications such as IoT, drones (aerial networks), underwater communication, and satellite access. All these integrated applications will massively rely on advanced AI and ML technologies. This will lead to ultra-massive computationally exhaustive applications and services. Edge computing has gained tremendous attention as a computation offloading and caching technique and is considered as an in-built capability in 6G. As shown in Figure 10 [25], Edge computing deploys computing resources in close proximity to the end-user. By doing so, there is a reduction in delay and latency, an enhancement of processing speed, and also the on-premise security is ensured. Mobile edge computing is an architectural standard given by ETSI for edge computing, which allows deployment flexibility at aggregation points, radio nodes, or at the edge of core nodes to ensure efficiency, reusability and extensibility.

Applications

In the smart city concept, a few applications of MEC are discussed below:

(a)

Video-based applications such as smart traffic monitoring, environmental monitoring or smoke detection capture voluminous video data. In such scenarios, video analytics performed at the edge locations can have a crucial role in distributing the computational load. In surveillance video applications, instead of transferring the massive data, computations can be performed at edge, and only the actionable data can be transmitted for a quicker response.

(b)

MEC in an intelligent transport system (ITS) has the capability to ensure that the transportation system is safe, sustainable and efficient. An IoT-based ITS is a combination of data-exhaustive, complex, uncertain and dynamic operations like resource allocation, task management, and security, for which edge computing can be effectively used. In any typical traffic scenario, there can be a wide range of different end devices in vehicles requiring high bandwidth and low latency. The computational device with required power can be placed closer to the end-user, for example, at a roadside unit (RSU) or a base station at the edge, as shown in Figure 11 This affords a significant decrease in end-to-end latency as compared to the centralized cloud [198]. ITS can be broadly classified into three service areas: (1) safety services to minimize accidents, (2) non-safety services to minimize congestion, and (3) infotainment services such as content sharing [199].

(c)

MECs use integrated aerial and terrestrial networks that are essential parts of 6G networks [200]. Unmanned aerial vehicles and drones constitute the aerial network and are widely used in surveillance, search-and-rescue and wildlife conservation applications. These applications are based on video analytics that are real-time. MEC is used in such video applications, which helps in computational offloading and latency reduction, but certain issues still need to be addressed, including (1) offloading computational tasks to another MEC when a drone flies away from the coverage area, and (2) limited drone battery power that restricts computational capacity.

2.10. Reconfigurable Intelligent Surfaces (RISs)

RISs constitute a planar surface with an array of passive elements whose characteristics can be altered dynamically using electronic circuits to reflect the incident EM waves [201]. Sixth-generation (6G) THz communication, due to its high frequency, is highly sensitive to obstructions in the environment, so the line of sight (LoS) signal is blocked easily. NLOS propagation also suffers refraction, reflection, scattering and diffraction at THz frequencies. To improve upon these limiting factors, existing studies have focused only on optimizing the transmitter and receiver side of wireless communication with an assumption that the wireless propagation environment is uncontrollable. However, the idea with RISs is to improve the coverage and signal quality by controlling the propagation environment. The concept of regarding the environment as an optimized variable is called smart radio environment (SRE), or more recently termed as intelligent radio environment(IRE) [201]. An RIS is a plane surface having a large number of reflecting passive elements that have the ability to reflect the signal wave received from a transmitter towards the designated receiver in a controlled fashion. Since an RIS can be considered as a complementary element in wireless systems, integrating it in wireless networks does not demand changing their standardization. Therefore, deploying RISs in wireless systems can be transparent to users, thus yielding excellent compatibility and high flexibility. Reflection beamforming by using RIS supports strengthening the signal-to-interference-plus-noise ratio (SINR) for users located near the RIS, thus improving the performance [202].

Since an RIS is a considered an almost passive reflective element, there is a large reduction in power consumption. Also, an RIS uses a simple configuration and has low manufacturing and deployment costs. Due to all these benefits, RIS is found to be an appealing technology for 6G to augment communication [202] and sensing [203]. Since RIS can intelligently alter the outbound reflected signal waves to focus on target receiver position, it finds application in high-precision positioning [204,205], where it can be used to create a virtual LoS link to cope with NLOS issues. Figure 12 illustrates how RIS can enable communication even in NLOS scenarios.

The majority of available work on RIS has concentrated on reflective RIS, assuming (a) all of the incident waves are reflected and are not able to penetrate the surface, and (b) both transmitter and receiver are on the same side of the RIS. To further advance the capability of RIS, refractive RIS is proposed in [206], wherein all incident waves can penetrate and pass through the RIS surface. In [206], the concept of refractive RIS is used to enhance QoS and reduce the power of the base station (BS) and satellite in a hybrid satellite–terrestrial relay network (HSTRN) setup. In [207], a hardware and channel model of the combined use of both reflective and refractive RIS is proposed.

Beyond ground-based applications, RIS can be used in conjunction with unmanned aerial vehicles (UAVs) that can efficiently enhance network coverage. Due to high mobility, UAVs can be positioned to deliver LoS dominant links for users on the ground. On the other hand, RIS can smartly alter its reflecting elements to realize passive beamforming. Also, RIS, with its inexpensive meta-surfaces, has the ability to mimic MIMO capability [208]. So, just by altering the phase shift of meta-surfaces, higher gain can be achieved, which does away with the requirement of using multiple antennas in a UAV. Another advantage of integrating RIS with UAVs is in minimizing the UAV trajectory, as the users located closer to the RIS can be served with coverage from the RIS rather than an UAV moving from one point to another. Figure 13 shows three possible scenarios of RIS integration with UAVs.

RIS can also be used to improve the energy harvesting performance of simultaneous wireless information and power transfer (SWIPT) technology [209]. The SWIPT technique is effectively used to transmit information and power simultaneously over a radio link, which means that the received signal can be utilized for information decoding and energy harvesting. RIS-assisted SWIPT can achieve both signal enhancement and efficient energy management. However, RIS-enabled SWIPT systems can also be compromised by eavesdroppers and pose a security risk. Several works in RIS-aided networks have been conducted to look into secrecy transmission designs, for example, artificial noise-assisted secure transmission [210], using an alternating optimization approach and successive convex approximation [211], etc.

In addition to the mentioned applications, RIS also has an important role to play in the range of applications in 6G networks, such as in ubiquitous connections, smart connections, and holographic connections.

Challenges

RIS placement: The optimal positioning and placement strategy for RIS poses significant influences on its performance. Several factors need to be carefully analyzed, such as user distribution, availability of location, cost, etc. Moreover, the maximum number of RISs to be deployed in the area should be analyzed, so that they do not interfere with each other.

Control protocol standardization: A control protocol needs to be standardized for communication between RIS controllers and base stations. This is typically important for a complex scenario where several RISs from multiple vendors and multiple operators are deployed.

“Multiplicative-fading” effect: In a scenario in which a direct link from the base station is strong, very low-capacity gain is attained using RIS. This is because of what is termed “multiplicative fading”, in which path loss on the Tx-RIS link is multiplied (instead of summed) with the path loss on the RIS-Rx link [212]. This creates a huge path loss, which has to be compensated by RIS having thousands of elements.

2.11. Non-Terrestrial Networks (NTNs)

NTNs can be classified into airborne (UAVs/drones), low-Earth orbit (LEO) satellites, medium-Earth orbit (MEO) satellites, and geostationary Earth orbit (GEO) satellites. Among these Airborne devices, LEO satellites are relatively closer to the Earth’s surface and so offer faster response time (smaller round-trip delay). However, since they are close to the Earth’s surface, they have a smaller coverage footprint. So, to offer a global coverage footprint, mega-constellations of LEO satellites will be required. Figure 14 shows a diagrammatic representation of terrestrial and airborne networks and LEO, MEO and GEO satellites.

2.11.1. Airborne Base Stations (ABS), UAVs, and Drones Uses in a 6G Smart City

Applications/Benefits

The use of the Terahertz frequency band will necessitate that the 6G base be placed in close proximity to the end user, which will lead to high 6G base-station density. In this scenario, the use of drones/UAVs can be a game changer due to its below-listed benefits and applications in the smart city arena:

(a)

Airborne base stations will provide line-of-sight (LoS), high-quality air-to-ground 6G coverage. They also extend the coverage footprint of terrestrial base stations.

(b)

Airborne base stations can be used to provide additional high capacity at dense urban hotspots that otherwise would not have been possible with a terrestrial base station.

(c)

UAV-BSs can be used to handle sudden surges in traffic during disaster recovery [213,214] or during large public gatherings such as sports events, concerts, etc.

(d)

In remote or rural areas that are lacking in telecom infrastructure, UAV-BS can be used as a low-cost solution to provide connectivity.

Challenges

There are several key challenges that need to be addressed in UAV-assisted base station setups.

(a)

Optimal placement and movement of UAV-BSs in the sky is required so as to have efficient network performance in dynamically changing network topology. Conventional heuristic and numerical methods may not be well suited for such dynamic environments [68]. AI-based methods [19] such as reinforced learning and deep reinforced learning (DRL) can be utilized in such dynamically changing environments.

(b)

Frequent radio link failure and ping-pong handover events triggered by fragmented reference signal receive power (RSRP) are another major problem [215]. In order to provide seamless UAV connectivity, optimal handover rules based on reinforced learning are a potential solution. The RL-based solution shown in Figure 15 uses a base station as agent and uses a combination of cost of handover and RSRP as reward. Based upon the agent’s reward from its action in the current state, it moves to a new state after interacting with the environment.

(c)

Limited energy availability in drones curtails their hovering time, which is a big hindrance in their adoption. However, a few techniques, such as the grasping capability [216] of drones, can improve their flying/hovering time.

2.11.2. Satellite Communication

The objective of 6G technology is to provide a 3-dimensional coverage footprint in space-air-sea-ground with integrated sensing, communication and computing. In order to provide global seamless coverage, the use of satellite communication becomes increasingly beneficial. For achieving ubiquitous communication, 6G-supported low-earth orbit (LEO) satellite constellations is the next frontier. LEO-6G-enabled satellite networks have lower construction costs, lower latency, seamless global coverage, much stronger signals, and higher user density [217]. LEO-based 6G networks have fueled great interest in spite of the higher Doppler effect. To deal with the high-speed mobile requirements, the traditional OFDM technique that deals in the time-frequency domain is proposed to be replaced with orthogonal time frequency space (OTFS) [218], which deals in the delay-Doppler domain. The use of the OTFS technique in 6G networks can address issues with high-speed communication applications such as UAV networks and airborne communication.

With the introduction of ISAC in 6G-enabled LEO, high-precision navigation and positioning for space-air-ground operations is enabled [219] since integration of communication and navigation (ICAN) is an integral part of ISAC. Therefore, 6G-LEO networks along with ICAN can efficiently deliver high-performance communication as well as high precision location-based services.

Additionally, in order to have seamless 6G coverage in space-air-ground (Figure 16), a careful design is required for efficient integration and interoperability of the aerial systems, satellite systems and terrestrial network segments since the resources in these are limited and unbalanced [220]. A space-air-ground integrated network (SAGIN) encompasses the challenge of interworking these three network elements, namely, aerial systems, satellite systems and terrestrial networks. It also involves the challenge of seamless call handovers between satellite segments, air segments and terrestrial base stations [221,222].

Based on above discussion,

Table 9. Concise summary of potential 6G technologies.

Tech.	Applications/Benefits	Challenges
AI	▪ Facilitates integration of various wireless technologies [67]. ▪ Channel estimation [69,70]. ▪ Handling scalability issues [71,72,73,74]. ▪ Modulation recognition [75]. ▪ Traffic prediction [76] ▪ Radio resource management [77]. ▪ Mobility management [78]. ▪ Enabling 6G PHY architecture [79,80,81,82,83,84,85,86,87,88,89,90]. ▪ Data-caching [91]. ▪ Energy management [92].	▪ High computing capacity required at edge. ▪ Energy consumption ▪ Large training data requirement. ▪ Security and privacy [93,94,95].
ISAC	▪ Localization applications [103,104]. ▪ Gesture and activity recognition [105]. ▪ Augmented sensing. ▪ Imaging and mapping	▪ Waveform design [106]. ▪ Performance tradeoff [107]. ▪ Clock Sync. requirements [108]. ▪ Channel modelling and evaluation method.
THz	▪ V2X comm. ▪ Nano-networks in medical domain [145,146]. ▪ Short-range comm. ▪ Health care [148,149]. ▪ Imaging [150]	▪ High propagation loss [151]. ▪ Small cell size (so, high network cost). ▪ Channel modelling requirements [151]. ▪ Hardware challenges [152].
BC	▪ Smart traffic management [129,131]. ▪ Supply chain management [130]. ▪ Industrial applications [39]. ▪ V2X [123]	▪ Computing and storage capacities. ▪ Energy consumption. ▪ Industry-wide protocol standardization.
QC	▪ Quantum algorithms in 6G [162,163,164]. ▪ QC and ML in 6G [153,162]. ▪ Quantum sensing [165,166]. ▪ Quantum Blockchain in 6G [166,167,168]	▪ Building quantum routers, switches and repeaters [169]. ▪ Measures of capacity not fully understood [170,171]. ▪ Establishment of quantum channels on long-distance.
NTN	▪ Airborne BS provides high-quality air-to-ground coverage. ▪ Used in disaster recovery [213,214]. ▪ Can augment capacity in urban hot-spots. ▪ Can provide coverage in remote areas with minimal cost.	▪ Optimal placement and movement of UAV in dynamic environment. ▪ UAV’s limited energy availability. ▪ Seamless call handover between space-air-ground segments.
MEC	▪ In ITS applications [198,199]. ▪ Video-based applications. ▪ Use in integrated aerial and terrestrial networks [200].	▪ Energy consumption at edge nodes. ▪ Handling of heterogeneous user applications and end-devices. ▪ Need for facilitating user mobility. ▪ Operations and remote management.
RIS	▪ NLOS applications [201]. ▪ Enhanced coverage [202] ▪ Sensing [203] and high-precision positioning [204,205]. ▪ Energy harvesting and power transfer [209]	▪ RIS optimal placement and positioning. ▪ Control protocol standardization. ▪ “Multiplicative-fading” effect
IC	▪ Remote surgery [176,177,178]. ▪ Holographic teleconferencing [175,179]. ▪ Immersive gaming [180]. ▪ Metaverse [181].	▪ Development of multi-sensory stimulation, e.g., olfactory. ▪ High cost of peripherals. ▪ Synchronization of multi-modal comm.
VLC	▪ Vehicle-to-vehicle comm. ▪ Under-water comm. ▪ Deployment in electromagnetic fragile zones like hospitals, aircraft, etc. ▪ LED-based WLAN setup.	▪ VLC integration with other comm. protocols. ▪ Next-gen. optical hardware requirement. ▪ Channel coding schemes. ▪ Recognized channel models

provides a concise summary of applications and challenges of 6G potential technologies.

3. Applications of 6G in Smart Cities

Technological solutions play a vital role in addressing urban challenges and transforming the city environment into a resilient, efficient and sustainable setup. The 6G technologies have the potential to not only to address the current stringent requirements of smart cities but can also open new possibilities with the collaborative use of various emerging 6G technologies. Future smart cities are anticipated to deploy large-scale IoE devices. The generated data require efficient and secure transmission, storage, intelligent processing and analysis so as to support a wide range of smart city applications spread across several domains, for example, intelligent transport systems, e-health, logistics, energy management, resource management, e-education, manufacturing, etc. Given below are a few major applications that can be implemented using potential 6G technologies:

3.1. Industrial Automation and Smart Manufacturing

With the introduction of Industry 4.0, every sector of industry is focusing on fully automating their manufacturing processes so as to gain more efficiency and reduce human errors. 6G promises a massively ultra-low latency network, supports enormous IoT devices, incorporates AI capabilities, and enables extremely high data rates with high capacity. All these features are crucial for the operation of thousands of systems and devices in a full automation process. Additionally, the compounding effect of 6G along with other emerging technologies, like AI, cloud computing, Big Data, Blockchain, digital twins, etc., will have an inevitable effect on the development of Industry 4.0 [223,224].

In Industry 4.0, the concept of smart manufacturing aims to promote processes and production flows that are more adaptable and resilient to fast-changing market demands [225]. It covers three parts—logistics, demand–supply, and smart factories. In a smart factory, two types of collaborative robots, namely the transportation robots and production robots, play key roles [226], thus forming a dynamic complex robotic ecosystem enabled by AI [227]. There is a requirement for real-time operation of these robots and sensors in the smart factory environment, which demands that communication be highly reliable, with high data rates and ultra-low latency. This is because the decision-making is based on real-time Big Data collected from this networked multi-robot system in a smart factory. Key technologies to meet these stringent requirements are 6G mobile communications, Big Data, robotics and AI. Figure 17 shows a smart factory architecture [228]. The smart factory can be visualized as four areas: Robots/agents, RAN part, core network and machine learning network.

Hexa-X, a flagship project by the European Union, envisages 6G as a convergence of three worlds, namely the physical world, the digital world and the human world. More importantly, it will seamlessly integrate humans. In the 6G era, the human inclusion will be the center of the system in all use cases and applications, forming a global Internet-of-Everything-and-Everyone. This vision, when attained, will set off a new revolution in the current Industry 4.0 concept. Humans will be integrated into the industrial system, not just as an external entity who is controlling it but as integral parts comprising invisible, organic systems [229]. As a 6G enabler, the digital twin (DT) technology can also be used to meet the challenges posed by Hexa-X [230]. Virtual replicas of physical systems or objects can be used in digital twin technology to simulate, optimize, and monitor their performance and behavior. In fact, 6G, I4.0 and DT are the three mutually integrated concepts that play a vital role in the industrial digital transformation.

3.2. Vehicle-to-Everything (V2X) Technology in Smart Cities

To align with the United Nations’ Sustainable Development Goals (SDGs) and 2030 Agenda [231], which are related to sustainable development, 6G networks will be focused to ensure a basic quality of life for citizens through the use of sustained mobility in future mega-cities.

In the smart city context, intelligent transport systems (ITSs) characterize the evolving topics of connected automated driving, connected cars, and vehicular communication systems. The goals of intelligent road transport (ITS) are not limited only to road safety, traffic efficiency, reduced emissions, affordability, accessibility, sustainability, etc., but also include CCAM (Cooperative, Connected and Automated Mobility) and social inclusion. Both ITS and V2X connectivity contribute to accomplish these targets.

To achieve the full potential of autonomous vehicles (AVs), there has to be a robust automation mechanism for the vehicle’s communication with objects in its surrounding environment. High-precision sensors are required that can sense the objects based on physical stimuli, which are categorized as internal and environmental [232]. Internal sensors capture parameters such as the vehicle’s speed, dynamic state, acceleration, and braking. Environmental sensors are used to monitor external objects, for example, pedestrians and road signs. Various radar-based sensors, proximity sensors, exteroceptive sensors, and cameras are used to perform these tasks.

The data thus collected are transmitted using a range of communication devices. Communication quality is the foremost factor for ITS functioning. Reliable, real-time and high-speed data flow with ultra-low latency and storage is essential for vehicular communication systems. 6G-V2X communication technology in the THz band can support hyper-fast, low-latency, and ultra-reliable communication compared to 5G-NR [233].

Different types of communication used in ITSs include vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2V), vehicle-to-pedestrian (V2P), and vehicle-to-RSU (V2R). For data management, data centers and cloud computing (vehicle-to-cloud (V2C)) for vehicular applications are implemented [234]. Various tasks such as remote vehicle diagnostics, complex computational tasks, and software updates are commonly executed on the cloud [235].

The evolution of autonomous vehicles heavily depends on the use of artificial intelligence (AI). Algorithms based on artificial intelligence and machine learning are implemented to make decisions and direct the vehicle’s activities. With the advent of deep learning (DL), several challenging problems such as path navigation and obstacle detection can be handled. An AI-based AV performs the below-mentioned actions [236]:

Perception: The surrounding environment is continuously scanned, and tracking is performed by the AV using various types of sensors, such as cameras, radar, or LiDAR, so as to emulate human vision. Convolutional neural networks (CNNs) can be used to detect road signs, other vehicles, trees, etc. [237].

Localization and mapping: In autonomous driving, localization is important since the reliability of several tasks depends on it. The task of navigation is simplified if the priori map of the environment is matched by the AV with the perceived surrounding features using sensors. By using the simultaneous localization and mapping (SLAM) method, the AV can ascertain its location in an unspecified area with no a priori map available.

Decision-making: This includes maneuvering through the traffic, path planning, and avoiding obstacles, automated parking, etc., without any human driver intervention.

In smart cities, with the exponential rise of inter-connected devices, voluminous data are generated and transported to cloud infrastructure for analysis and storage. This can introduce performance issues related to latency, bandwidth and disruption in AV functioning. The highly demanding performance requirements of autonomous vehicles have resulted in a confluence of artificial intelligence and edge computing, leading to the emergence of a model termed Edge-Intelligence (EI or Edge AI). Thus, the computing performance of AVs can be greatly enhanced by implementing an EI model so as to improve latency and accuracy [238].

Use Cases of V2X

Through the amalgamated use of edge and cloud computing, Big Data and AI for V2X in 6G networks, the following use cases in the smart city paradigm can be considered:

(a)

Smart routing—Vehicular traffic congestion in urban settings causes delays and increases fuel consumption and pollution. Smart routing can be used to guide drivers to follow the most efficient routes, which are less congested. This is achieved through the use of AI-based routing algorithms [239] that are based on real-time data collected from IoT sensors, vehicle ad-hoc networks and pedestrian activity. This is particularly useful for emergency vehicles like ambulances, fire brigades, etc. Also, through its use, the traffic is evenly distributed across all possible routes. Additionally, it affords better fuel efficiency and reduced emissions [240].

(b)

Smart parking—In any urban environment, parking problems have increased with the increase in vehicle density and limited parking spaces. So, the key to solve this issue is to optimally utilize the available spaces. A smart solution to the problem can be to use sensors to capture the occupancy status of a parking lot, and an application can be used to display that status. This will not only optimally utilize the parking spaces but will also reduce the time and fuel consumed by vehicle drivers in searching for parking spaces.

(c)

Speed harmonization—By shaping the speed of vehicles as per determined recommendations, smooth and safe traffic flow can be achieved. By doing so, frequent acceleration and deceleration is avoided, which yields better fuel efficiency and reduced emissions. Another approach is using green light coordination, in which multiple green lights work in a coordinated manner so that a smooth traffic flow is ensured. AI with cloudification is a technology enabler to implement this use case. For faster communication, localized V2X should be implemented between vehicles whose speed harmonization is required.

(d)

Green driving—In order to reduce the environmental impact of vehicles in critical areas such as near hospitals, schools, etc., a green driving strategy to reduce emissions is used. One way to achieve this is to devise a traffic management strategy to reduce the number of vehicles in the area during high pollution periods. Another way is to reduce the speed of vehicles and hence their fuel consumption and emissions.

Table 10. Summary of V2X use cases.

Application (Use Case)	Benefits	Devices/Tech Used
Smart Routing	Avoidance of traffic congestion. Useful for emergency vehicles. Traffic balancing on roads. Reduction in emissions [240] Reduce delays.	IOT sensors. Vehicle ad-hoc networks. AI real-time routing algorithms [239]. Cloud and edge computing for data processing and analysis.
Smart Parking	Contribution to sustainability. Optimal utilization of parking spaces. Reduced time for drivers to search for parking spaces.	V2V and V2I communication. Use of sensors for indicating parking status. AI and cloud computing.
Speed Harmonization	Reduces frequent need for acceleration and deceleration. Continuous traffic flow. Reduces emissions. Safe travel.	AI and cloudification. Green-light coordination.
Green Driving	Reduction of fuel consumption. Reduction of pollution near critical areas like hospitals.	Collection of pollution data by roadside sensors. Data transfer to centralized cloud. Traffic management decision based on AI algorithm. On-road displays for flashing traffic management decisions.
Coordinated Maneuvers	Smooth traffic flow. Emission reduction.	V2I information exchange among vehicles and RSU [241]. Low-latency, low-delay transmission. Advanced AI implemented at edge for delay-free decisions.

provides summary of use cases of V2X.

3.3. Smart Healthcare

By 2050, the human population above 60 years of age will increase to 2.1 billion from a figure that was 1 billion in 2020 according to the WHO [242]. Such a phenomenal increase in the older population in the future will overwhelm the traditional methods of medical care. Therefore, smart remote medical care or telemedicine systems through the use of IoT devices will play a crucial role in remote patient monitoring, evaluation and even treatment [243,244]. During the last few years, there has been tremendous growth in the range of IoT wearable/implantable sensors and devices. These IoT medical devices can be categorized based on the location they are placed. On-body devices are the ones that are placed in the clothing or on the skin of the patient. Off-body devices are the ones that communicate with external devices outside the body of the patient while the in-body type are those devices that are implantable, injectable or ingestible [245,246]. To monitor and obtain real-time data on the patient’s health using these medical sensors and devices, they need to be connected wirelessly using a wireless body area network (WBAN) [247]. Treatment can be given remotely to patients based on symptoms and the captured data. So, by using a smart healthcare system, the detection of disease, its diagnosis, analysis, prevention, prediction and treatment can be achieved in a more efficient, evidence-based and effective manner. However, the huge real-time heterogeneous data collected from a range of medical sensors, devices, electronic health records, genomic data and medical imaging need to be captured, stored and analyzed; this is known as a Big Data challenge due to the unstructured and complex nature of such data [248]. By leveraging the use of 6G technologies, along with machine learning and advanced statistical models, more personalized and accurate treatments can be delivered. By doing so, patient wait time is reduced, efficiency of medical staff is enhanced, and cost-saving opportunities are afforded. Future 6G networks are targeted to achieve a more harmonious human–machine interconnection. Also, 6G can deliver the delay-sensitive extreme ultra-reliable requirement of health care applications. Recently, the concept of Internet of Medical Things (IoMT) was introduced; this concept involves IoT devices related to the health care sector and performing the function of collecting medical data and running medical software applications. IoMT devices are capable of predicting risk, making fast decisions, and taking suitable actions. As relatively newer concepts inspired by progress in nanotechnology, the Internet of Nano Things (IoNT) and Internet of Bio-Nano Things (IoNT) have ushered in a new dimension in health care Systems. Inside the body, these bio-nano machines communicate with one another using molecular communication, which is a bio-inspired technology where information is exchanged through molecules [249].

Challenges

(a)

Major challenges in the healthcare system include security and privacy. Failing to satisfy these challenges may lead to erroneous treatment decisions and catastrophic results.

(b)

Further study and research are required in molecular communication in eHealth systems, leading to the practical applicability of IoBNT. Further insight into biological channel environments and system architectures is required for efficient coding, modulation and detection [249].

3.4. Smart Grid

One of the important pillars of sustainability in the smart city is in efficient energy/power generation, its distribution and its transmission. Earlier energy grids were based on centralized power generation and the unidirectional flow of energy [250]. A smart grid now involves the implementation of ICT and 5G/6G technologies along with AI [251], cloud computing and massive machine-to-machine communication to make it a more efficient energy distribution system. Smart grid management (SEGM) systems enable the bidirectional flow of energy with more advanced features like real-time monitoring, self-healing and awareness, security, scalability, resilience, automatic decision-making, etc. [252].

Essential components of a smart grid are demand–response (DR) systems, advanced metering infrastructure (AMI) [253], grid automation, distributed energy resources (DERs), and energy management systems (EMSs). In addition, a smart grid seamlessly integrates renewable energy sources (RESs) such as wind and solar energy into the grid [254]. The smart grid network comprises three components, which are generation, transmission, and energy consumption [252], as depicted in Figure 18 [255].

Among a few challenges in smart grid implementation, firstly, the massive data collected from meters, sensors, and control systems need to be processed, analyzed and actionable, and insights need to be derived so as to optimize the network [256]. Secondly, within the grid, there is an issue of slow connectivity between devices and sensors, resulting in a delayed response [257]. These challenges can be addressed by leveraging the features and capabilities offered by 6G, which enables faster, reliable data exchange with low latency and can support massive numbers of sensors/devices [255]. In addition, 6G wireless networks can further seamlessly integrate advanced technologies such as edge computing [258], cyber twin and Blockchain to augment the reliability and resilience of smart grid systems. Blockchain technology can ensure security, trustworthiness and transparency of smart grid transactions, thus ensuring peer-to-peer energy trading and data integrity [259].

Challenges

(a)

Cybersecurity: It is crucial to safeguard the integrity of smart systems and their data, as any laxity can lead to catastrophic results. Massively connected cyber physical systems (CPS) like a smart grid, which works on 6G networks and uses software-defined IoTs, can be vulnerable to cyber-attacks [260]. Technologies that can be leveraged to significantly reduce the risk are Blockchain [261], network slicing, edge computing, DSA, and cyber twins.

(b)

Interoperability and standards: In a smart grid, a massive number and wide range of smart sensors and devices are interconnected and should seamlessly communicate with one another. Open architectures facilitate interoperability between devices and systems. Further, AI and SDN can help to meet the interoperability challenges.

3.5. Smart Waste Management

Sustainability is one of the important goals of 6G communication, under which it addresses the urgent demand for environment friendly networks. Smart waste management (SWM) in the smart city environment is an important field for the accomplishment of sustainable development. With the rise in urbanization, the amount of waste generated globally is estimated to surge to 2.2 billion tons by year 2025 [262]. SWM involves efficient management and monitoring of waste collection, processing and disposal.

SWM systems can be implemented by using massive numbers of smart IoT sensors and devices that are capable of measuring, processing, analyzing, storing and communicating the information [263]. SWM systems gather and analyze data from smart garbage bins (SGBs) equipped with sensors. Based on the analysis, the SWM system performs the classification, segregation and processing of solid waste and the management of waste trucks for disposal. In ref. [264], a waste management system is proposed with three subsystems: (a) smart garbage bins using IoT devices, (b) smart garbage collecting trucks with GPS for city route selection, and (c) the collection of user (drivers, dispatchers, citizens) information on a real-time basis. SGB levels are captured, based on which an estimated waste collecting schedule is compiled to helps drivers and dispatchers make more informed decisions, thus saving time and fuel. Energy saving at the SGB level is further demonstrated in [265], in which a low-energy adaptive clustering hierarchy method is used as an optimization process.

Waste classification is an important step performed to identify biodegradable and non-biodegradable waste materials. The waste classification can be performed at the smart garbage bin (SGB) by using heterogeneous sensors, actuators, or image capturing for identifying the trash. Once the data are captured, there are several methods available that can assist in classifying the waste using image processing, artificial intelligence, deep learning [266,267,268] and computer vision [269].

With the emergence of 5G/6G technologies and intelligent products, for example, wireless sensor nodes and Internet of connected Vehicles (IoCV), the classification, segregation and disposal of urban waste is now more efficient and intelligent. Planning for waste collection vehicles and the optimization of their service routes based on data transmitted by smart waste bins can be achieved on a real-time basis [270]. In ref. [271], the garbage collection vehicle routing problem is addressed with meta-heuristic algorithms to minimize the costs of operation and minimize the environmental impact by applying the Industry 4.0 concept. In another approach, the shortest path of the garbage collection vehicles toward the waste centers is estimated using an IoT-based optimal routing path protocol that considers distance, energy, delay, and weight of the waste [272]. Ref. [273] investigates the relation between waste bin fill level and the overall efficiency of waste collection. Efficiency in terms of transport fleet usage, optimal traveled distance, and amount of waste collected is shown to be achieved by setting a threshold waste level (TWL) parameter between 70 and 75%.

4. Conclusions, Open Challenges and Possible Future Research

The integrated use of 6G in smart cities can bring a transformative change that will benefit the way we live and work in the urban environment. In this article, we have highlighted the 6G next-generation technologies that can be used in the smart city arena. The interoperability of these technologies needs to be achieved efficiently so as to enable seamless smart city applications. The article has also discussed important applications using 6G technologies. The journey toward 6G-enabled smart cities will not be without challenges. These challenges, which are listed below, can give direction to future research.

▪

Further work is required in developing energy-efficient and green technology solutions that can not only be environmentally beneficial but also reduce capital and operational power expenditures. Since 6G devices will operate in a higher frequency band, they are expected to demand higher energy requirements [274]. Techniques like energy-aware task offloading, efficient resource management, energy-aware architecture, the use of RIS [275], energy harvesting, and energy cooperation [276] need to be implemented to enable energy efficiency. Also, the convergence of 6G and AI will potentially address the shortcomings of network topology and find an optimized path toward a sustainable ecosystem [277,278].

▪

Robust measures are required to address security and privacy concerns, which are heightened due to hyper-connected devices beyond 5G-advanced networks. Advanced threat-detection mechanisms and quantum-safe cryptography need to be in place and may become central themes in protecting the integrity of 6G smart city networks.

▪

For THz use in 6G, further research in accurate channel modeling is required. Also, the hardware challenges in the development of high-frequency THz antennas, amplifiers, and mixers, including the challenge of designing high-bandwidth super heterodyne transceivers, need to be addressed [61,152].

▪

With the use of NTN, fast moving UAVs and satellites will be major challenges because their relocation has an impact on network topology and handover. This will make mobility scenarios for 6G much more complex since both ground users and aerial base stations are mobile.

▪

The TCP/IP protocol is often implemented in computer networks, but its use in satellite-based networks will be inefficient due to long delays and higher BER. So, in order to have smooth interoperability between terrestrial and non-terrestrial networks, a new or improved protocol might be needed. Several improved protocols are used in satellite communication, such as TCP Reno (1990), TCP Vegas (1994), TCP Westwood (2001), MPTCP (2011), Cross-layer, Novel-ECM, etc. During the 6G evolution process, several versions of the TCP enhancement technologies may coexist during the evaluation stage to meet the mega-constellation’s link criteria [279].

▪

There is a need for the development of a smart city digital technology platform that can converge technical solutions that currently prevail across multiple cities. This will help in promoting data interoperability and technology neutrality [280]. The platform should have the capability to bring comprehensive convergence and generativity spread across the smart city innovation ecosystem [281,282]. A recent survey paper [283] on middleware smart city solutions performed an assessment of the functional and non-functional requirements of 20 different middleware solutions, like AMF-CPS [284], CityPulse [285], FIWARE [286], S2NetM [287], etc. But there are still open challenges in using smart city middleware, such as interoperability, security issues amidst big data, scalability, context management, energy efficiency, reliability, and QOS. Strong collaborations between various smart city stakeholders, such as computer scientists, social scientists, urban planners, and other specialists, can lead to greater understanding and solutions.

▪

A 6G-enabled smart city should address the issue of the digital divide, which means that its network services should be available and accessible to all citizens. Although 6G is aligned to address this issue through the combined use of several technologies, such NTN, RIS, AI, Cloud, VLC, etc., there may be challenges in generating cost-viable integrated solutions so that smart city applications are affordable for all citizens.

▪

During the transition to new 6G technology, the existing devices/sensors in smart cities will need to be replaced with 6G-enabled smart devices. This will involve massive economic costs, among other challenges. There is a need to harmoniously combine multiple technologies to mitigate these costs.