diff --git "a/2dAzT4oBgHgl3EQfRvud/content/tmp_files/2301.01221v1.pdf.txt" "b/2dAzT4oBgHgl3EQfRvud/content/tmp_files/2301.01221v1.pdf.txt" new file mode 100644--- /dev/null +++ "b/2dAzT4oBgHgl3EQfRvud/content/tmp_files/2301.01221v1.pdf.txt" @@ -0,0 +1,1341 @@ +Unlocking Metaverse-as-a-Service +The three pillars to watch: Privacy and Security, +Edge Computing, and Blockchain +Vesal Ahsani +Department of Electrical Engineering, +Sharif University of Technology, +Tehran, Iran +vesal.ahsani@ee.sharif.edu +Ali Rahimi +Department of Electrical Engineering, +Sharif University of Technology, +Tehran, Iran +a.rahimi@ee.sharif.edu +Mehdi Letafati +Department of Electrical Engineering, +Sharif University of Technology, +Tehran, Iran +mletafati@ee.sharif.edu +Babak Hossein Khalaj +Department of Electrical Engineering, +Sharif University of Technology, +Tehran, Iran +khalaj@sharif.edu +Abstract +In this article, the authors provide a comprehensive overview on three core pillars of metaverse-as-a-service (MaaS) platforms; +privacy and security, edge computing, and blockchain technology. The article starts by investigating security aspects for the +wireless access to the metaverse. Then it goes through the privacy and security issues inside the metaverse from data-centric, +learning-centric, and human-centric points-of-view. The authors address private and secure mechanisms for privatizing sensitive +data attributes and securing machine learning algorithms running in a distributed manner within the metaverse platforms. Novel +visions and less-investigated methods are reviewed to help mobile network operators and metaverse service providers facilitate +the realization of secure and private MaaS through different layers of the metaverse, ranging from the access layer to the social +interactions among clients. Later in the article, it has been explained how the paradigm of edge computing can strengthen different +aspects of the metaverse. Along with that, the challenges of using edge computing in the metaverse have been comprehensively +investigated. Additionally, the paper has comprehensively investigated and analyzed 10 main challenges of MaaS platforms and +thoroughly discussed how blockchain technology provides solutions for these constraints. At the final, future vision and directions, +such as content-centric security and zero-trust metaverse, some blockchain’s unsolved challenges are also discussed to bring further +insights for the network designers in the metaverse era. +Index Terms +Metaverse-as-a-service (MaaS), privacy and security, edge computing, blockchain +I. INTRODUCTION +What is the metaverse, exactly? The metaverse is a concept in the tech world that refers to a digital living environment +where conventional social structures are changed. It is a term that combines the concepts of the Greek1 prefix “meta,” which +means “more complete” or “transcending,” and the acronym “Verse” for “universe,” which signifies a space-and-time container2. +The idea of the metaverse was introduced in Neal Stephenson’s science fiction novel Snow Crash nearly 30 years ago [1]. +The rapid advancements of technologies like blockchain, virtual and augmented reality, gaming, artificial intelligence, and the +Internet of Things have made the metaverse one of the most buzzworthy terms in the tech world. Solutions and services are +being developed for virtual worlds to allow users to have fun, intelligently engage with their surroundings, and form deeper +connections with others [2]. Investment in the metaverse has grown significantly, with technology giants investing billions of +dollars in its development and many businesses putting together their own plans for the metaverse. A McKinsey & Company +report predicts that the metaverse will be valued at over $5 trillion by 2030 [3]. +What is Metaverse-as-a-Service (MaaS)? The phrase “as-a-Service” originally appeared in a 1985 file with the United States +Patent and Trademark Office (USPTO), and it gained popularity throughout the cloud computing era [4]. Everything that +may be considered as a service through a network can be referred to as XaaS [5]. Everything-as-a-service (XaaS) is a recent +development in the information and communication technology sector that enables the provision of scalable computing resources +on demand. Accordingly, the Metaverse can profit from “as-a-service” models, in which the key elements and technologies of +the Metaverse, such as platforms, infrastructures, software, and artificial intelligence (AI), could be provided as service models, +1https://en.wikipedia.org/wiki/Meta +2https://alldimensions.fandom.com/wiki/Category:Verse +arXiv:2301.01221v1 [cs.CR] 1 Jan 2023 + +i.e., Metaverse-as-a-service (MaaS). As tech giants entering the Metaverse space, including Microsoft, Samsung, NVIDIA, and +others, it won’t be long until there are many marketplaces with a Metaverse-as-a-Service (MaaS) offering allowing businesses +to profit from the technology with lowered entry barriers. Even though MaaS is still a relatively new technology area, there +are currently a number of vendors making headway there; such as Lovelace World, Propel MaaS, Touchcast, and MetaVerse +Books. +MaaS is characterized as an on-demand subscription solution that enables companies and/or operators to create and implement +different forms (such as existence, management, coordination, and implementation) in the Metaverse to support the processing +of Metaverse services, collaboration, company operations, and products, among other related scenarios. Everything in Metaverse +can be thought of as a delivery model that can be simply generated and/or modified as function modules, similar to XaaS in +cloud computing systems. See Figure 1. +The main benefits of using MaaS can be summarized as follows: +• Products for the Metaverse may be created by businesses without substantial digital experience or knowledge. Without +prohibitive capital requirements, even small to mid-sized firms can engage in the Metaverse economy. +• It promotes financial investment in a still-developing technology. The majority of systems are currently only designed +for consumer usage, and solutions like Microsoft Mesh and the Horizons app suite from Meta have not yet been widely +released. In this setting, MaaS enables businesses to make low-risk investments and profit from technology. This model +also reduces the time of programming, setting up and installing systems and brings the investor a profit sooner. +• MaaS may ultimately lead to industry standardization, with selected few businesses serving as Metaverse “brokers” to aid +in infrastructure creation. +• MaaS model is a win-win situation. The seller of this service does research, programming and implementation only once, +but can sell this service many times. On the other hand, the buyer of the service can also use Metaverse at a competitive +cost without getting into the technical, management and implementation complexities of Metaverse. +Physical world +Perception Layer +Users +IoT, sensors, … +Devices +Human-computer +interaction +AR/VR/… +Holography comm. +Service providers +Edge computing +layer +Edge servers +Computing +Storage +Communication +Shared, Scalable +5G/6G +Network Core +Cloud +infrastructure +Semantic comm. +Virtual world +Metaverse service +models +AIaaS, NSaaS, +BCaaS, … +SaaS, IaaS, PaaS, +DaaS, … +Application layer +Avatars +Digital twins +Virtual environment +Metaverse engine +Security & privacy +Modeling, +simulation, … +Human-centric +Learning-centric +Data-centric +Real-time +Interoperable +Management +Virtualization +Machine learning +Resource +integration +AI controller +Metaverse service +instantiation +resource pool +Distributed +management +Blockchain +Consensus +Smart contracts +Trading +Data management +Policies +Rules, SLAs, … +Regulation +Fig. 1. Structure and important modules in the Metaverse network. +Despite the outstanding advances in today’s existing MaaS platforms, the Metaverse would still need a more comprehensive +and robust collection of standards and protocols to embrace interoperability, much more comprehensively than what the +current internet includes, based on a set of rules and policies for communication, visuals, graphical demonstration, and data. +For instance, Fortnite runs on practically all popular platforms (such as iOS, Android, PlayStation, and Xbox) and supports +numerous identity/account systems and payment options, which forces rivals to cooperate (i.e., engage in interoperability). Web2 + +goliaths like Apple and Google employ similar technology today, but they are not built to integrate with one another. Building +a scalable Metaverse will depend more on interoperability than anything else. Companies might create their own Metaverse +campuses using established protocols with the support of Metaverse-as-a-Service (MaaS), and then start providing immersive +experiences that promote social interaction. The metaverse infrastructure should be supported by extensive accessibility to +mediate various aspects of human-beings’ lives, to be able to provide immersive experiences. This opens up new venues for +privacy and security risks. +The Internet-of-Things (IoT) plays an important role in digitalizing the physical world by utilizing pervasive sensors, cameras, +wearables, etc. Networking connectivity is then provided via wireless networks, while the computing and storage are provisioned +through cloud and edge computing. The IoT networks act as “bridges” between the physical world and its digital counterpart +[6]. The information flow between the two worlds, can help facilitate the decision making in both physical and digital worlds. +Users, who are mainly represented as avatars, can also produce and exchange digital contents across various platforms in the +metaverse. Mobile users interact with the digital world via their smartphones, wearable devices, and augmented reality/virtual +reality (AR/VR) helmets, to create and share contents and gain knowledge. +Information is the core resource of the metaverse. The data flow within the employed networks of metaverse has the key role +in realizing the integration of physical and digital worlds. To better understand the security and privacy needs in the metaverse, +we indicate that there exist two main sources of information in this era: The first one is the data gathered from and exchanged +by the real world that might be utilized and visualized digitally in the virtual space. The second source of information is the +output of the virtual worlds, e.g., the information generated by digital assets and services in a MaaS platform. At the same +time, artificial intelligence and machine learning (AI/ML) algorithms, performed in the computation layer or the digital twin +layer, help facilitate rendering and offering various services. Accordingly, it is crucial to safeguard the privacy and security of +data flows within the MaaS platforms, as well as the learning algorithms. +On the other hand, performing all metaverse computations on cloud servers is not necessarily the best option. In some +situations, using edge servers close to users can facilitate low-latency services, reduce network traffic, help manage data, +and improve user experience. In addition, the paradigm of edge computing is closely related to other technologies such as +artificial intelligence and IoT. For example, with the help of edge servers, it is possible to train neural networks locally with +the participation of end devices without having any internet connection. Also, edge servers enable secure management and +access control of IoT devices and sensors on-premise. +Last but not least, blockchain is regarded as one of the metaverse’s core infrastructures and helps supply the metaverse +with laws that are clear, open, effective, and trustworthy [7] because it can connect disparate minor sectors and create a solid +economic system,. It will be challenging to determine the worth of the commodities and resources traded in the metaverse +without the assistance of blockchain technology, especially when those digital components interact economically with the +real-world economy. Therefore, it would be wise to investigate blockchain technology for MaaS platforms along with the other +two pillars; privacy and security and edge computing. +In the following, we provide a comprehensive review on guidelines to safeguard the privacy and security of MaaS platforms +from different perspectives. Additionally, in order to help metaverse operators to identify an appropriate approach for using +edge computing in the metaverse, we have focused on the advantages and challenges of using the edge computing paradigm +in MaaS in two separate subsections. Moreover, we thoroughly discuss the requirments of using blockchain technology and +the actions MaaS developers should take to solve many technical challeneges by using blockcahin. Finally, we conclude the +paper with future visions and directions. +II. PRIVACY AND SECURITY +A. An Overview of Privacy and Security Challenges for the metaverse +Despite the ever-increasing advances in developing numerous services for the metaverse, privacy and security challenges are +considered as the main concerns that need to be properly understood. Considering the fact that realizing the concept of MaaS +requires an extremely wide variety of computation, communication, and networking, a wide range of security and privacy +threats also arise in the metaverse. See Figure 2. +A variety of recent technologies are integrated into the metaverse as its basis, hence, the intrinsic vulnerabilities of them +may be inherited by the corresponding MaaS platforms. Recently, numerous privacy and security flaws have been identified +for the emerging technologies, including the vulnerabilities of cryptography-based key management schemes against quantum +computers [8], [9], the privacy leakage of distributed learning networks against honest-but-curious servers or malicious clients +[10], and the misuse of massive private data by service providers [11], just to name a few. Such threats can be intensified in +the virtual world, while simultaneously other threats can also happen, which does not exist in the physical world, e.g., virtual +spying [11], [12]. Notably, massive amount of sensitive data are utilized to create a digital replica of the real world. This +opens up new challenges in terms of threats to privacy and security. Individuals use wearables or AR/VR devices with built-in +sensors collecting biometric data, brain patterns, users’ speech and facial expressions and poses, and also the surrounding +environment in order to render a high-quality immersive experience for the MaaS users. In addition, clients are under the threat + +Privacy and security +challenges for the +Metaverse +Data fusion of +multimodal +private data +Distinguishing +the realness +Highly +heterogeneous +virtual worlds +Securing the 6G- +enabled access to the +Metaverse +Secure wireless +secret key +generation +PHY +authentication +for the metaverse +access +Adversarial +learning for +wireless access +Data-centric privacy +and security +Data collection +Incognito mode +Authenticating +synthetic data +Secure and private +AI/ML for the +Metaverse +Privacy for +AI/ML +algorithms in the +Metaverse +Secure AI/ML +for the +Metaverse +Secure +aggregation +framework +Human-centric privacy +and security: privatized +social interactions in the +Metaverse +Social +applications +Biometric data +Privacy- +enhancing +protocols +Fig. 2. Privacy and security for the Metaverse. +of being illegally tracked through their VR headsets or wearables. Such a miscellaneous set of sensitive information flow in +the context of metaverse services provides adversaries with a broad attacking surface which requires proposing novel secrecy- +and privacy-preserving mechanisms for mobile network operators and service providers. +The main privacy and security challenges in the era of metaverse can be identified as follows. +• Data fusion of multimodal private data corresponding to interactions between users, avatars, and environments imposes +serious challenges for the MaaS platforms. Collection, communication, and processing of such sensitive information must +be taken into consideration in terms of privacy and security risks. +• The blurring boundary between the real world and its virtual counterpart causes serious confusions in terms of distin- +guishing the realness. Therefore, we should focus on providing stringent authenticity guarantees from the MaaS providers’ +perspective. +• Virtual worlds in the metaverse are highly heterogeneous in terms of hardware implementation, communication, and +learning interfaces. The beyond-the-fifth and the sixth generation (B5G/6G) of wireless communication networks, as +the main fabric to facilitate the communications within the metaverse, pose critical challenges for conventional security +solutions. This raises urgent needs for proposing context-aware and flexible security and privacy mechanisms [9], [13]. +We aim to carefully review these items in this paper. We provide readers with useful insights through a comprehensive overview +of security solutions across different layers of the MaaS platforms, ranging from the access layer to the social aspects. We +also review different solutions in the contexts of communication, computation, and learning. +B. Securing the 6G-enabled Access to the Metaverse +For the deployment of access layer to the metaverse, the sixth generation (6G) of wireless networking will play a pivotal +role. 6G technologies, such as pencil-sharp beams, edge AI, and integrated sensing and communications (ISAC), provides +users with perceived understanding of their surroundings. These technologies are the key enablers for implementing novel +PHY layer security (PLS) schemes as promising approaches to safeguard the security of the metaverse access, while offering +information-theoretic security guarantees [14]. Leveraging PLS, intelligent, flexible, and context-aware mechanisms are feasible +for detection and mitigation of privacy and security vulnerabilities [9], [13], [15], [16]. Considering the new era of truly end-to- +end (E2E) quality-of-experience provisioning for the MaaS products, service level agreements (SLAs) are expected to include +guarantees about the quality of security (QoSec) as well [13]. This will include addressing the required security level, adaptive, +risk-aware, and adjustable security solutions, just to name a few [9]. The evolution of the metaverse systems is expected to +introduce new means to harvest and interpret the “context” of the communication, where PLS techniques can be considered +as a promising QoSec-assuring approach. +In order to secure the metaverse access, we introduce the following security mechanisms, which can provide lightweight +and scalable solutions for securing the access to the MaaS platforms. + +1) Wireless Secret Key Generation and the Role of PHY Security: Smart wearable devices and cameras, together with +head-mounted displays, e.g., Oculus helmet and Vive Pro headsets, and handheld controllers are considered as the main +terminals for entering the metaverse. In this regard, key generation and management is essentially required for smart devices +to establish secure connections for accessing the metaverse, transmit sensory data, and receive immersive experiences. In [17] +and [18], intrinsic features of specific wearable devices, such as gestures and motions, are taken into account to propose key +agreement schemes. Researchers in [19] propose a heartbeat-based key generation scheme based on the measurements from +electrocardiography (ECG) and photoplethysmography (PPG) sensors as the source of common randomness. +While the security mechanisms of the 5G standard rely on cryptography-based keys, such as the elliptic curve cryptography +(ECC) to fulfill the confidentiality and authentication requirements, 6G networks with many peer-to-peer communications +undermine the performance of conventional solutions. Besides, assumptions on the security of cryptographic methods based +on the mathematical hardness of solving a certain problem will no longer hold with the advent of quantum computers. To +efficiently secure the metaverse era, data communications should to be proactively secured, where PLS has such capabilities +for quantum-resilient security [9]. As a promising framework to migrate from the conventional complexity-based solutions +towards lightweight security techniques, PHY layer secret key generation (PHY-SKG) has been envisioned to be employed +for 6G networks [15], [16]. PHY-SKG is realized through implementing lightweight mechanisms with minimum required +changes at the control plane. This can actually provide the access networks with different advantages. In particular, PHY- +SKG is thoroughly decentralized without relying on any infrastructure of a particular entity, which can substantially reduce +the required time for key agreement, making it suitable for extremely low latency applications in the metaverse. Moreover, +continuous update of the key is realizable thanks to the dynamic variations of the PHY attributes. The PHY-SKG protocol +under realistic assumptions of hardware impairments and adversarial attacks is developed in [15]. Notably, the PHY-SKG has +shown to be resilient against man-in-the-middle adversarial attacks in [16]. Furthermore, for establishing secure communication +between digital healthcare devices and the access point, a lightweight learning-based key agreement protocol is proposed by +the authors in [20], which can guide designers to bring device-level intelligence for securing the metaverse access. +2) Adversarial Learning for Wireless Access: Due to the proliferation of virtualized services in the metaverse era, malicious +actors may have access to the network functions, such as the open-source software of virtualized access networks, and can +exploit vulnerabilities of the metaverse access through adversarial machine learning (AML) [21]. Adversaries can poison the +wireless access of the metaverse, by manipulating the inputs to the learning algorithms employed for authorizing users [22], +leading to unauthorized access to the metaverse network due to the mislead access model. Malicious users can also infer +sensitive information about edge devices, users, and applications which have access to the network, through membership +attribute inference attacks [23] and model inversion attacks [24]. As a privacy threat to the MaaS access, data characteristics +such as device-level information may leak to adversaries. Malicious users can exploit this leaked information using membership +inference attacks [23] by building an inference model to determine whether a sample of interest (associated with a particular +device) has been used in the training data of the MaaS provider. To mitigate such attacks, generative adversarial networks +(GANs) are shown to be capable of detecting anomalies and mitigating wireless attacks for the next generation of communication +and networking services [25]. +Connection requests for the metaverse services come with quality of experience (QoE) requirements such as throughput and +latency, which can be fulfilled by assigning radio resource blocks and processing power to the requests. Although network +slicing can manage the MaaS QoE requirements, AML techniques have the potential to attack network resource management +and disrupt B5G network slicing, which can impose huge losses for the network operators, if not properly secured [22]. To +combat such attacks, different reactive and proactive defense mechanisms are proposed in [26], including: the induction of +randomness to the decision processes to reinforce robustness against malicious nodes. +So far, the security challenges for accessing the metaverse were addressed. In the following sections, we go deeper into the +privacy and security aspects inside the metaverse, in a data-centric manner. We also address synthetic and authentic realness, +envisioning the privacy enhancement within the metaverse platforms. +C. Data-centric Privacy and Security +Inside the metaverse, the scope of data collection and processing far exceed what is traditionally performed in mobile and +web applications. The technologies employed for the metaverse do not just track where users click, but where they go, what +they do, whom they interact with, what they look at, etc. Therefore, it is crucial to the users’ privacy and security to implement +data-centric privacy-preserving mechanisms and push for meaningful, sensible, and aggressive regulations in terms of data +privacy. +Recent studies demonstrate how easily the metaverse users’ privacy can be compromised [27]. It is shown that an adversarial +program can accurately infer over 25 personal data attributes. To deal with the unprecedented data-centric privacy risks in the +metaverse platforms, the idea of “incognito mode” has recently been developed for the metaverse as a novel and promising +solution for safeguarding VR users’ privacy [28]. To realize the incognito mode in the metaverse, client-level differential privacy +framework is proposed which has been shown to be capable of protecting a wide variety of sensitive data attributes in VR + +applications. It is show in [28] that by employing randomized mechanisms (depending on the format of the private attributes), +sensitive user data attributes can be obscured effectively [28]. +We are facing a new trend of intelligent systems that are empowered by synthetic realness, in which AI-generated data is +vastly exploited to reflect various aspects of the physical world [29], [30]. Inside the metaverse, consumers are not simply +targeted via pop-up ads, rather, they are provided with “immersive contents” in the form of virtual people and activities that +look realistic [31]. Nevertheless, we should concern about the authenticity of synthetic data at different levels. Authenticating +synthetic data can make AI algorithms more fair and secure through correcting data biases and safeguarding data privacy +[32]. Synthetic data is utilized for training AI models in ways that real-world data practically cannot or should not, while +protecting privacy and confidentiality. It is critically important for the network operators and MetaaS providers to leverage +generative AI in an authentic way to maintain trust for their customers [33]. To realize authentic realness, the provenance, the +policy, and the purpose of utilizing synthetic data should be taken into account by companies and service providers [34]. To +this end, distributed ledger technology (DLT) can be employed to verify the provenance of digital content, hence addressing +the authenticity. Moreover, network operators should clarify the purpose behind the exploitation of synthetic content and the +resulting advantage over non-synthetic content. +D. Secure and Private AI/ML for the Metaverse +In this section, we focus on challenges and solutions for the privacy and security of AI/ML implementations in the metaverse. +Although metaverse can bring amazing AI/ML-based services to individuals and enterprises, the utilized learning algorithms +are still vulnerable to privacy and security risks. There exist challenging bottlenecks for efficient deployment of AI/ML +techniques. Among these challenges, the big concern is that the learning algorithms might leak users’ private data [35]. +Besides, users might not be willing to share their private data in the metaverses, making it challenging for AI/ML algorithms +to perform a comprehensive data analysis for enhancing the learning-based services of the metaverse platforms. To address +the privacy risks of AI/ML algorithms, the integration of federated learning (FL) [36] and cross-chain technologies can be +considered as a promising solution to design and implement privacy-aware frameworks for the MaaS platforms [37]–[39]. For +instance, [39] proposes a hierarchical blockchain-based framework for decentralized FL. A main chain is employed to mimic +the parameter server (aggregator server) of the FL algorithm, and multiple subchains are implemented to manage local model +updates generated by smart devices (or their digital counterpart) acting as workers. +Despite the fact that FL algorithms leverage locally-trained models instead of the raw data of metaverse participants, sensitive +information can still be inferred by analyzing the model parameters uploaded by clients [24]. If the model updates are inspected +by bad actors, participant users’ privacy and security would be threatened. In addition, there might exist some malicious users, +who proactively upload adversarial data to the aggregator server(s) to mislead the training process, known as backdoor attacks +[35]. FL still has limitations in terms of security and privacy: First, malicious clients may inject mislabeled or manipulated +data to mislead the global model, which is known as data poisoning attacks [21], [40]; Second, during model update of AI/ML +algorithms, adversarial participants can save the model parameters and infer sensitive information, e.g., private attributes of +participants’ data. They might also be able to recover the original training samples of users [24], [39]. In the following, we +discuss the possible countermeasures to the abovementioned threats in more details. +We take into account the fundamental fact that local model updates in distributed AI/ML algorithms carry extensive +information about the local datasets owned by the metaverse clients. Hence, secure aggregation of local models are necessary +for the distributed learning entities in the metaverse. Moreover, there might exist some adversarial participants in the metaverse +platform who actively upload “poisoned data” to the servers (as aggregator entities) to mislead the training processes. Such +attacking venues can provide the surface for the so called “backdoor attacks” on the collaborative model training procedures +within the MaaS platforms [35]. As an example, virtual service providers and mobile network operators employ learning +algorithms as a service for their users [41]. Malicious users have the capability to eavesdrop on the models exchanged to infer +the membership of specific clients or undermine the performance of AI/ML services [23]. In addition, attackers can hack edge +devices in the physical world or impersonate benign avatars as a man-in-the-middle. They can then inject adversarial training +sample to spoof the learning services within the MaaS products. +In the following, we address secure aggregation framework to protect learning tasks from being eavesdropped by malicious +participants. +Secure Model Aggregation: Generally speaking, state-of-the-art secure aggregation (SA) protocols rely on two main +principles: i) pairwise random-seed agreement between clients to generate “masks” for hiding the metaverse users’ models; +and ii) secret sharing of the random seeds. A SA protocol enables secure computation of distributed learning models, while +ensuring that the aggregator entities do not infer information about the local models [42]. Taking into account the fact that +the MaaS users might be “honest-but-curious” during the learning process, a SA protocol must guarantee that nothing can +be learned beyond the aggregated learning model, even if up to T users cooperate with each other. This is called T-privacy +property of SA algorithms. In [42], a modular system design for SA is proposed, which can help develop lightweight SA +protocols for the learning infrastructure of the MaaS platforms. + +Backdoor Attacks: Malicious clients may inject mislabeled or manipulated data to mislead the learning models within +the highly-distributed and heterogeneous infrastructure of the metaverse platforms, which is known as data poisoning attacks +[21], [40]. In this context, backdoor attack is a type of “data poisoning” attacks that aims to tamper a subset of training +data via injecting adversarial triggers, such that AI/ML models trained on the manipulated dataset make (targeted) incorrect +prediction/decisions during the inference [40]. +Backdoor attack is aimed to mislead the trained model to infer a target label on any input data that has an adversarially- +chosen embedded pattern (a.k.a trigger). Distributed version of backdoor attacks, known as DBA, decomposes a global trigger +pattern into separate local patterns and “embed” them into the training set of different adversarial parties respectively. This +is in contrast to the conventional attacks on the surface of distributed AI/ML algorithms [43], where a unified global trigger +pattern is embedded for all malicious parties. The experiments in [44] show that the DBA method outperforms centralized +attack significantly when evaluated with the global trigger. +To overcome backdoor attacks, adversarial training (AT) is currently considered as an empirically strong defense mechanism +[40]. AT corresponds to the mechanism of training a model on adversarial examples, with the aim of making the learning +service more robust to attacks and reducing the inference error on benign samples. AT is comprehensively reviewed in a wide +variety of researches, including [40], [45], [46]. Recently, a generic framework to defend against the general type of data +poisoning attacks is proposed in [40] by desensitizing networks to the effects of poisoning attacks. To mention an application +of AT within the metaverse, an AT-based protection against facial recognition systems is proposed in [47], which can be +effectively utilized for the MaaS platforms. +E. Human-centric Privacy and Security: Privatized Social Interactions in the Metaverse +In the metaverse, human users socially interact with others through their avatars and experience MaaS platforms by leveraging +ubiquitous smart devices and network access in a real-time manner [48]. The metaverse as a “human-centric” virtual environment +is supported by massive human-centric social applications, such as holography AR, immersive VR gaming, and the tactile +Internet. Such immersive services can be realized through the interdisciplinary communication-computation-storage co-design +techniques, integrated with the concepts of human perception, cognition, and physiology for haptic communication [49]. Hence, +the most sensitive and personalized information will be exchanged among different parties, which significantly highlights the +essence of employing novel intelligent privacy- and secrecy-preserving algorithms for the information security and social +privacy of MaaS users [50]. +The biometric data plays a significant role in terms of social interactions inside the metaverse. To elaborate, human-driven +agents and avatars, with whom the users interact, are being developed based on users’ personal data [51]. Such sensitive data +can be fed into AI/ML algorithms to create personalized “interaction partners” and influence social interaction behaviors of the +metaverse users. By exploiting physical and mental traits inferred from biometric data, the interaction partners can maliciously +lead users into undesired situations and behaviors. Hence, the metaverse users can be manipulated not only by businesses and +institutions, but by other individuals in a human-centric manner. +To mitigate such human-centric threats, attention should be paid to the users themselves. Metaverse service providers should +make users aware of the implications of biometric data extracted by AR/VR devices. Developers can implement privacy- +enhancing protocols such as differentially private mechanisms as proposed in [52] and [53]. Users can also be provided with +secondary avatars, e.g. clone, to hide their actions in the metaverse [11]. The secondary avatars help users hide their real +behavior within the metaverse. Furthermore, users must be able to have an adjustable level of privacy [28] to dynamically +manage their personal space in the virtual world by choosing their desired privacy configuration. From a governance layer, +a modular bottom-up governance approach is proposed in [54], which can be adapted to different platforms and use cases. +Decentralized autonomous organizations (DAOs) are examples of governance systems that allow users to be involved in +decision-making processes for the MaaS platforms [55]. In this regard, organizations such as XR Safety Initiative (XRSI)3 +will play an important role in encouraging MaaS providers and network operators to help facilitate enhancing trust within the +MaaS implementations in terms of human-centric privacy and regulations. +III. EDGE COMPUTING +To unlock the full potential of the metaverse, edge computing represents a promising paradigm for technical development of +the metaverse. Edge computing is a well-known technology that allows operators to perform processing, storage, and network +functions close to the edge of the network, where users are located. By using this paradigm in the Metaverse, network efficiency +is improved, costs are reduced, and new features are added. +As an illustrative example, consider a smart hospital in the Metaverse. In this center, telemedicine facilities are implemented. +Artificial intelligence is used to diagnose diseases and suggest treatment, hospitalized patients are monitored in real time with +IoT sensors, and remote surgery can be done in this center. In this scenario, if all this data is to be sent directly to the +3https://xrsi.org + +cloud server for processing and storage in a traditional way significant network bandwidth will be occupied. In addition, with +this method, there will be a long delay in sending data, and it is possible that the confidentiality of patients’ data will be +compromised. This motivates the need for the edge computing paradigm that can process and store data near the edge of the +network where data is generated. In short, this design can be such that several servers are installed in the hospital that can +quickly collect and process large amounts of data, feed it to neural network models, then send the extracted information to the +cloud server. +Edge computing can be a lucrative opportunity for network operators. One-third of network operators’ costs are spent on +real estate to house infrastructure [56]. Converting these locations into edge computing sites is a huge revenue potential for +network operators. They can also use their extra processing and storage resources to empower network edge fascilities. On +the other hand, operators can have a successful entry in the development of metaverse. They can play key roles by using +the services they provide in the 5G and 6G mobile networks [57]. Consequently, providing MaaS is a smart way to boost +operator income and develop diverse applications. In the MaaS model, organizations can flexibly create their own virtual world +according to their needs and with the most recent versions of the tools. This is made possible by using operator edge facilities +and networks. In addition to the network operator that provides a lot of hardware and software infrastructure, vendors such +as application developers, edge computing equipment vendors, system integrators, and edge computing solution providers are +also active in this market and new revenue lines are open to them. Furthermore, various vertical markets such as automotive +vehicles, virtual content productions, robotics, eHealth, entertainment, smart manufacturing, smart grids, and many others are +added to this market. For a survey on the edge computing market, see [58]. +An edge computing system includes the internet gateway, edge servers, and perception layer [59]. Internet gateway is +responsible for connecting the edge system with the network backbone and cloud servers. Edge servers process, store, and +aggregate data. Perception layer includes edge devices such as IoT sensors and actuators, AR glasses, VR headsets, haptic +gloves, smartphones, wearable devices, personal computers, and holographic displays. The metaverse can benefit from edge +computing in different ways. See Figure 3. In the following, we explain some of them. In addition, edge computing also brings +challenges. After explaining the advantages, we explain the challenges of using edge computing in the metaverse. It should be +noted that we provide a high-level review of these items and delving into their technical details is beyond the scope of this +paper. +Latency reduction +• Physics emulation +• Graphic rendering +• Real-time straming +Performance optimization +• Resolving computation bottlenecks +• Network traffic control +• Higher variety of services +Data control +• Access control +• Scalability +• Ubiquitous access +Capacity enhancement: +• Network traffic reduction +• Free up cloud resources +• Orchestrate resource sharing +Computation offloading +• Enrichment of sensors and actuators +• Expanding into new markets +• Reducing energy consumption +Privacy preserving +• Strengthen data ownership and trust +• Leveraging encryption +• Training neural networks on local data +Monitoring & management +• Edge-assisted communication +• Dynamic efficient network +• Remote management +Reliability & availability +• Tactile Internet +• Hazardous environments +• Anomaly detection +Edge intelligence +• Edge computing for AI +• Edge AI for network management +• Distributed machine learning +Cost reduction +• Profitable edge computing +• Local solutions for local problems +• Resource sharing +Modern networking techniques +• Semantic communication +• Virtualization of real-world edge networks +• Taking advantage of multi-user scenarios +1 +2 +3 +4 +5 +6 +7 +8 +9 +10 +11 +Fig. 3. Metaverse requirements to be met by edge computing. +A. Metaverse requirements to be met by edge computing +Many applications in the metaverse need seamless synchronization of the real and virtual worlds. Some of these applications +are critical, such as remote control of a robotic assisted surgery system, and some of them, such as guiding an avatar in a +virtual environment, are related to the quality of the user experience (QoE). Therefore, it is necessary to provide a metaverse + +infrastructure to provide low-latency services. In traditional solutions, processing and storage are outsourced to cloud servers. +These servers can easily be located thousands of kilometers away from users, and this distance causes a lot of delays in +telecommunications. Edge computing, as an alternative solution, brings processing and storage closer to users and the place +where data is generated. Hence, in the literature, some metaverse processes that are very apt to run on edge servers and edge +devices such as physics emulation [60], [61], graphic rendering [62]–[65], and real-time streaming protocols [66] have been +investigated. +When edge servers are set up near users, it does not mean that the operator transfers some calculations from the cloud +server to the edge server forever. Instead, it flexibly uses edge facilities to optimize network performance depending on +different conditions. In other words, edge computing and edge storage capabilities add new dimensions to network performance +optimization problems and enlarge their feasible region. Therefore, the network has several more degrees of freedom. The +operator can dynamically decide, depending on the situation, how the edge servers come into play and bring advantages to +the metaverse network. For example, edge servers schedule computing tasks [67], provide better coverage [68], run load- +balancing algorithms [69], resolve computation bottlenecks [70], and provide heterogeneous services well addressed in the +MEC (Multi-access Edge Computing) standard [56]. +To make the metaverse network scalable, it is necessary to support issue tracking, data quality management, data protection, +and risk control functions. On the one hand, we know that the metaverse uses distributed systems. It also captures sensitive +data such as biometric data, eye movements, and financial data. In addition, the metaverse interacts with a wide range of +technologies from blockchain to artificial intelligence. For this reason, the impact of data control functions in the metaverse +is significant. On the other hand, we know that edge computing enables better and faster data control by keeping data in the +local network and close to the user. Edge computing makes it easier for organizations to control data access to a universal +metaverse [60] whilst manages data exchange in various ways such as W-LAN, cellular networks, and satellite communication +systems which are supported by 5G and 6G [71]. It is also possible to support distributed computing and use the processing +and storage resources of end-devices, which can help network scalability [72]. +Furthermore, small devices may not be able to run many heavy tasks in the metaverse like the inference of large neural +networks and real-time renderings. In this case, they should be able to outsource these tasks to powerful servers or perform the +tasks with the help of other devices in a distributed scheme [73]. This reduces telecommunications and thus reduces energy +consumption [74]. Edge servers can provide more performance, flexibility, and storage than edge devices. In addition, it is +usually easier to upgrade edge servers than end devices. This is because we can equip the edge server with more GPUs, RAM +cards, and storage devices, or change the previous components to increase the power of the edge server. But we may not be +able to disassemble all edge devices, including IoT modules, and change their features. So far, a lot of research has been done +on the outsourcing of computing of IoT devices to edge servers, in which its advantages, challenges, architectures, and various +approaches have been explained [75], [76]. Further, some IoT devices have limited batteries, and energy costs can be high. In +any case, optimizing energy consumption in the Metaverse network is crucial. A practical case study is presented in [77] that +explains the impact of edge computing on energy consumption in IoT devices. +Even with modest Metaverse assumptions, data usage could easily expand more than 20x during this decade [78]. If edge +servers perform storage and processing tasks for users, there is no need to send large amounts of raw data deep into the +network. For this reason, the telecommunication capacity of the network is less occupied. Hence, edge computing can replace +cloud computing at a lower cost in many applications. Additionally, since there are a lot of powerful edge devices in the +metaverse network, it is possible to encourage resource sharing among edge computing applications. Thus, edge devices can +serve as microcenters for storage and processing. In this case, the capacity of the metaverse network will increase without +occupying the resources of the cloud server [79]. The authors in [80] have presented an efficient method for sharing resources +at the edge of the network in the framework of the resource pool. For a survey of incentive mechanisms for sharing resources +at the network edge, see [81]. +Privacy protection is another challenge in the metaverse. Storing data on a local edge server (for example, at a university, +hospital, or airplane) can help keep users’ data confidential [82]. Edge servers can monitor the passage of sensitive data and +be equipped with hardware and software protection tools. Many IoT sensors and devices are not smart enough to not be +compromised. Sometimes their communication is not secure enough. Having an edge server near these devices and possibly +owned by the user can maintain confidentiality and bring more trust. Users connected to an edge server can use local area +network policies such as device and file sharing. On the other hand, authentication mechanisms can be implemented on edge +servers to secure the network against various types of hacker attacks [83]. Edge computing can come to the aid of end devices +and participate in cryptographic algorithm calculations [84] and training neural networks on local data [85]. +Edge servers make the network more self-aware and easy to manage and monitor. Spending on creating sites for edge servers +can be multi-purpose and simultaneously address several different needs. Edge servers provide monitoring of local network +information and its devices and help with higher-level configuration and customization. Edge computing is a key technology in +the achievement of 5G goals. It is also one of the enablers of 6G networks [86]. It enables distributed scenarios and low-latency +telecommunications which can help improve network traffic and be part of upcoming solutions in telecommunication issues + +such as channel selection. Furthermore, to choose the most appropriate coding and communication protocol, edge servers can +send network conditions to end devices and vice versa [87], [88]. Edge computing tools help end devices in many ways. In +addition to performing calculations, storage, caching, prefetching, and data traffic prediction of end devices, they can improve +network efficiency and dynamics. For example, they can have different communication channels and direct the information +packets of connected devices from the most efficient path [89]. Additionally, solutions have been offered to avoid the need for +physical presence to calibrate end devices with the help of edge computing [90]. Environmental threats such as temperature +rise, humidity, dust, and power outages are just as serious as cyber threats and can put sensitive edge devices out of service +[91]. The presence of processing and storage facilities at edge computing sites enables edge remote management solutions. +With the help of these solutions, we can take care of the critical facilities of the Metaverse network at the edge and reduce +the energy required for troubleshooting and repair. +Additionally, we want the metaverse network to be more reliable and available. Although there is no universal definition +of reliability and availability, and there is no standard way to measure them, we can talk about strengthening the reliability +and availability of some services compared to others. In some applications that are offered under MaaS, these two features +are prominent and should be considered in design, programming, and warranties given in agreements. In the event of a +network failure, edge computing infrastructure can still maintain the connection of the devices and perform some processing. +Additionally, it is possible to add redundancy to the system at edge computing sites for specific applications. By bringing +the processing tasks to the edge servers, the connection of the devices to the core network is reduced. This reduces the level +of vulnerability of the system to cyber threats. In [92], to increase the reliability of the Metaverse network, the authors have +presented a method based on distributed coded computing algorithms that can be implemented at the edge. Also, a growing +number of solutions are being offered for edge computing reliability in hazardous environments [93]. System reliability is +significantly affected by the detection and prediction of anomalies. Artificial intelligence can train machine learning models +on the large amount of local data that edge servers have access to and use them in real-time [94]. +3D reconstruction, character animation, detection of harmful content, machine vision, gesture recognition, facial expressions, +understanding emotions, body movement recognition, physical interactions, eye tracking, brain-computer interface, and many +more are activated by artificial intelligence in the Metaverse. Some strategies for using machine learning to strengthen security +in the Metaverse and to create recommendation systems are proposed in [95]. For various reasons, there is a rising desire to train +and execute machine learning models on edge servers, which is called Edge AI or Edge Intelligence. Many applications in the +metaverse such as autonomous vehicles, intelligent interaction of avatars with the environment in the virtual world, health care, +smart city, and industrial complexes need machine learning algorithms [96]. Having computing servers on the edge equipped +with artificial intelligence allows better network and telecommunication management. For example, in situations where we +have many sensors in an environment and we don’t want their massive data to leave the local network raw, we can reduce the +traffic volume with the help of an edge server that can process their data and reduce their size. In addition, in the edge server, +algorithms can be implemented to predict traffic and cache popular data, which increases the service quality. Today, machine +learning has been used in different telecommunication layers, from the physical layer and coding of telecommunication signals +[97] to the detection of cyberattacks in the application layer [98]. Designing distributed machine learning algorithms such as +federated learning [99], split learning [100] and gossip learning [101] is a hot research topic because it adds new capabilities to +the network. For example, with distributed learning, the neural network is trained without the data leaving the device and the +processing load is spread across multiple devices. In addition, in a local area network where Internet traffic is limited, devices +can continue to train neural networks with the help of each other without having a permanent connection to the Internet [102]. +As previously mentioned, edge computing is not necessarily a substitute for cloud computing. Rather, it is an option with its +advantages and challenges. The costs and desired quality of the various services provided in the metaverse should be examined +and the appropriate way to provide them should be chosen. In addition, attention should be paid to the interaction of cloud +servers and edge servers and how to divide their work. Due to the existence of various end devices and network equipment +that can be used in the metaverse, the importance of this review increases. Edge computing reduces network traffic load and +reduces costs. Additionally, it may be cheaper for some businesses to build edge servers than to outsource computing. For +example, in some regions, the cost of electricity makes it cost-effective to install edge servers. Cost-effective solutions for +edge servers are also provided. For example, the study [103] shows that micro data centers at the edge cost 42% less than +traditional centralized data centers. +In telecommunications, there are several techniques to reduce costs and increase capabilities. Some of them are in the +common domain of edge computing and the metaverse. Network designers can use these ideas. For example, semantic and +goal-oriented communication is provided to increase stability and efficiency [104], which can be used in the 6G network and +contribute to the metaverse. Edge servers can be equipped with various neural networks and perform semantic communication +tasks. On the other hand, the metaverse can help itself is by creating digital twins of edge devices. By monitoring the digital +twins of edge devices and edge infrastructures such as edge servers, I/O ports, RIS (re-configurable intelligent surface), UAV +(unmanned aerial vehicles), SAGIN (space-air-ground integrated network), and network equipment, the MaaS operator can +instantly control and calibrate these physical entities. + +B. Edge computing challenges +An operator who wants to use the edge computing paradigm in MaaS design must consider several challenges. Limitations of +edge servers is one of these challenges. Due to the limitations of end devices (power, bandwidth, computing, storage, etc.), edge +servers are proposed to assist end devices in tasks such as artificial intelligence processing and high-quality image rendering. +But it should be noted that the power of edge servers can also be limited due to hardware limitations or server room space. +In this case, the edge server cannot provide all the desired functions for a large number of users at the same time. +Another challenge to be considered is the resource allocation problem. Users in a region will likely make many requests +to the edge server at the same time and we will see network spikes. Managing these requests is a challenging issue. Another +issue is the management of faulty hardware. Cloud servers have more infrastructure compared to edge servers. The reason +why it is beneficial to have spare hardware on hand is so that in the event of a hardware failure, it is possible to replace it +quickly. But on edge servers, this issue is not economical. Suppose you own one device, and you buy a spare. The cost will +double. But suppose you have 100 pieces of equipment. If you buy a spare, your cost will be 101 instead of 100, which is +relatively not much. In addition, fairness is also essential in resource management. Because there is a lot of resource sharing +in the services provided at the edge, the metaverse operator at the edge needs to be very careful about the optimal scenarios. +For more information on resource management at the edge, see [105]. +Stragglers and malicious node are serious problems in outsourcing scenarios at the edge. When a computing load is distributed +among several servers or data is stored in several servers, it is possible that some of these servers, i.e., stragglers, will take +more time than usual to respond. This issue causes our entire work to be delayed on those servers. Malicious nodes in the +network are entities that seek to disrupt the functioning of the metaverse. There should be pre-prepared scenarios to deal with +stragglers and malicious nodes. There are various solutions to mitigate stragglers and malicious nodes in the network. For +example, in coded computing algorithms, we can divide a task into several parts and outsource it to several servers. If some +of these servers are stragglers or malignant, the computation’s final answer can still be recovered [106]. +There are also a couple of communication and networking challenges in using edge computing for the metaverse. Metaverse +services have diverse and strict requirements. For example, touch Internet requires high-reliability telecommunications, AR/VR +headsets require high bandwidth, and IoT devices require high coverage. A large amount of data in the network and its dynamic +nature should not cause friction and decrease the efficiency of the system in terms of reliability, delay, and rate. Increasing the +variety of services can make MaaS faster and more stable. For example, if different service modes and qualities are available, +for example, the video size can be changed or artificial intelligence can be adjusted with different accuracy levels, some of +the network challenges in overload conditions will be solved. +The distributed and heterogeneous nature of the metaverse network at the edge complicates its management too. In addition, +edge servers must be synchronized with end devices and cloud servers. In addition, various functions must be executed +simultaneously to maintain network performance, such as anomaly detection algorithms, resource provisioning, workload +prediction, and traffic routing. In places where edge servers are installed, such as a hospital, a network specialist is not always +available. For this reason, another challenge of network management at the edge is the high cost of admin intervention for +configuration, updates, and equipment maintenance. +Ownership considerations are another critical issue in this matter. Equipment, data, and software on edge servers can have +different owners. For example, in the case of equipment, some edge servers are privately owned, some are cooperatively owned, +some are rented out to generate revenue, and some are created with specific access requirements. This issue can make it difficult +for the Metaverse operator to enter into contracts and launch new services. +As the number of end devices, edge servers, and IoT modules increases, the metaverse network becomes challenged in various +sectors like wireless communication, queue management, and user authentication. For example, in wireless communication, +interference and congestion can occur. In queue management, there is a possibility of data being thrown away, and user +authentication can take a long time. We also know that edge servers hold sensitive data generated in places like hospitals, +financial institutions, and homes. But keeping data close to the user is not enough. Users and businesses should be given access +and data confidentiality guarantees. For more information on edge privacy solutions, see [107]. +In some applications, it is necessary to know who is responsible for each operation at the edge of the network. It should also +be possible to receive complaints and prevent manipulation, corruption, and abuse. Decision-making processes and functions +should be expanded to the extent that the privacy and property rights of individuals are not compromised. The participation of +different people and communities in monitoring the network not only improves the performance of edge servers in one place +but also creates knowledge that can be used throughout the metaverse network. +The needs of users are not the same in all regions. In addition, the data formats available in the network and devices connected +to the network are diverse. Therefore, the MaaS operator must design and install edge server equipment and software in a +flexible and compatible structure and configuration. Another issue is changes and developments in the network. The network +can change over time in terms of users or cloud servers. Accordingly, edge equipment should be able to meet these requirements +at a low cost. + +IV. BLOCKCHAIN TECHNOLOGY +To provide users with the best experiences possible, the metaverse gathers enormous amounts of private data. This information +is required by the businesses or programs in order to construct targeted systems successfully. With its authentication, access +control, and consensus methods, blockchain gives consumers total control over their data, protecting their personal information +[108]. +The blockchain utilizes hash algorithms and asymmetric-key encryption to protect data in the metaverse. Additionally, the +quality of the real-world data that users exchange is crucial to the construction of objects in the metaverse. This data is +collected from numerous applications, including those in entertainment, healthcare, and more. Blockchain enables individuals +and businesses to validate all transactions by providing comprehensive audit trails of all transactions. The metaverse’s data +quality will improve as a result [109]. +Blockchain +solutions +for MaaS +Data acquisition +Data storage +Data sharing +Data +interoperability +Data privacy +preservation +IoT +Digital twins +AI +Big data +Interactivity +Fig. 4. Blockchain solutions for MaaS. +On the other hand, the successful exchange of AR and VR data is essential to Metaverse. The smooth and safe data sharing +within the metaverse is made possible by the blockchain’s cutting-edge encoding information system. Moreover, stakeholders +in the metaverse must have access to and control over resources in various virtual environments. Due to the many settings +in which these virtual worlds are created, data interchange is constrained. A cross-chain protocol enables data interchange +between two or more blockchains that are present in different virtual worlds. In terms of integrity, data in the metaverse must +be constantly and accurately updated. Due to the immutability offered by the blockchain, the metaverse data is kept as a copy +in each block along the chain and cannot be changed or withdrawn without the agreement of a majority of the participants. +A number of vendors have already begun to enter the MaaS market; according to an announcement made by Propel, a +blockchain solutions platform, it would provide MaaS solutions for smart contracts, NFT utilities, and decentralized finance +(DeFi). Moreover, Lovelace, a different blockchain-based platform, offers a MaaS toolkit that gives users and developers +the technology required to build and trade NFTs, run smart contracts, monetize VR gaming, interact with other metaverse +platforms, and much more [110]. In the following, the authors concentrate on the steps that a MaaS developer should do and +utilize blockchain technology to address various difficulties associated with creating and developing the Metaverse platform. +See Figure 4. + +A. Open challenges for MaaS +Data Acquisition: The metaverse will generate large amounts of unstructured, real-time data through decentralized appli- +cations, but acquiring this data can be difficult. Building applications in the metaverse, such as recommender systems, will +require high levels of data integrity. The use of virtual reality and increased streaming in the metaverse will further strain data +acquisition systems [111], [112]. The quality of the data may also be impacted by the acquisition of duplicative or incorrect +data [113]. +Data Storage: Once the metaverse is fully functional, the physical world’s ability to store data may be strained to its +breaking point. This could create significant challenges for the metaverse [114]. If the metaverse relies on a central storage +system, there is a risk of data leakage, manipulation, or loss. The potential of the metaverse to offer biometric data, voice +inflections, and vital signs that depend on sensitive data is also jeopardized by the likelihood of data loss and corruption in +centralized applications [115], [116]. +Data Sharing: Data sharing on centralized platforms carries the risk of exposing private and sensitive information [117], +[118]. Additionally, due to data mutability, there is a risk of high latency and reduced data availability [119]. This is particularly +relevant in the metaverse, where many applications will generate large amounts of real-time data. When demand for real-time +data increases, data flexibility can become a problem. +Data Interoperability: The metaverse will be created through the merger of many digital domains, but these domains are +currently fragmented and disorganized. This can make it difficult for users to engage with multiple virtual worlds, as they must +set up separate accounts, avatars, hardware, and payment infrastructure for each one [114]. There are also few methods for +users to transfer their digital assets between different digital environments. In order for the metaverse to be truly interoperable, +digital world apps must be able to easily exchange information with one another, regardless of their location or the technology +being used. The conventional approach to interoperability is inadequate for the metaverse, so new solutions are needed [120]. +Data Privacy Preservation: In the early stages of the metaverse, attackers may be able to deceive users and steal important +data. This could be particularly dangerous if attackers use artificial intelligence bots, as users may not be aware that they are +not speaking with a real person. The metaverse also raises concerns about the confidentiality of personal data, particularly +personally identifiable information (PII) [121]. Finally, as the metaverse grows and more validity information is included, +managing the large amounts of data will become increasingly difficult. +IoT: The metaverse will have a large number of interconnected IoT sensors, which raises concerns about IoT security +and storage. Real-time analysis of unstructured IoT data is also challenging [122]. When storing data across virtual worlds, +a centralized solution is not ideal, as tampering with even one piece of data could compromise the entire set of findings. +Additionally, data sharing across virtual worlds will depend on the cross-platform capabilities of IoT devices [123]. Finally, +IoT data tracking is necessary for safety and legal compliance. +Digital Twins: The quality of the data used to build digital twin models is important for their accuracy, so the information +provided by the source must be accurate and of high quality [124]. Digital twins from different sectors, such as healthcare and +finance, must be able to communicate and connect with one another. To improve the accuracy and consistency of communication, +digital twins should be able to detect and correct faults. However, data security can be a challenge when using a range of +devices and sensors to create digital twin models that utilize real-time data. This can be particularly vulnerable to botnets and +other viruses [125]. +AI: The ownership of AI-powered content in the metaverse is difficult to determine, as users have no way to distinguish +between communicating with a real person and an avatar created by a computer. This could lead to users using AI technology +to exploit other users or resources in the metaverse, such as by cheating at games or stealing from other users’ accounts [126]. +Additionally, AI may make mistakes, which could lead to people losing trust in the metaverse. Another challenge is the use +of a similar blockchain across different AI applications in the metaverse. +Big Data: One of the main challenges is the sheer volume and rate of data production in the metaverse, which can be +difficult to keep up with, even with advances in data storage technology. Another challenge is the variety of data produced +by metaverse apps, which can make it difficult and time-consuming to collect and organize the necessary data for consumers. +Additionally, the rapid development of big data technology [127]–[132] can make it challenging to stay up to date with technical +advancements in the metaverse. +Interactivity: The metaverse, which is a virtual world created through the use of technology like holographic telepresence and +augmented reality, offers immersive, realistic experiences by combining audio, video, cognition, and other elements. However, +the use of XR technology in the metaverse also creates challenges related to data storage, data sharing, and data interoperability. +For example, the businesses can create recommendation systems using data from XR technology, but this data must be stored +securely and shared transparently with stakeholders. Additionally, the metaverse must be able to handle the exchange of data +between virtual worlds in an interoperable way in order to provide users with a seamless experience [133]. + +B. Blockchain solutions for MaaS +Data Acquisition: With the use of blockchain technology, it will be simpler to gather reliable data in the metaverse for +uses like social networking. Blockchain’s distributed ledger will make it possible to trace data in the metaverse and validate +transaction records [134], [135]. Because the majority of nodes in the ledger must consent before any modifications to the +data in the metaverse can be made, data collection is therefore resistant to attacks [136]. A blockchain-specific validation +process that is driven by consensus mechanisms is applied to all data collected in the metaverse [137], [138]. Every action is +documented as a transaction on a blockchain, and each block includes a cryptographic hash of the one before it, as well as the +metadata, a date, and the activity [139]. As a result, changing the data in one block will change the data in all the other blocks +as well. Any block’s data is impervious to manipulation [140]. There will be no repetition in the data collecting process since +the likelihood of producing a duplicate block is almost zero. Data obtained by blockchain enabled acquisition mechanisms in +the metaverse will be trustworthy since every block is approved on the blockchain [141]. +Data Storage: The metaverse storage is impermeable to hacking since a new block is generated for each transaction [142]. +As a result, data is stored across the chain as a copy of the original blocks, increasing data dependability and transparency +in the metaverse [143]. If the centralized data store is hacked, the metaverse applications, which include anything from real +estate to digital things, would be very vulnerable [144]. Utilizing blockchain technology will lead to a large number of blocks +contributing to data distribution, enhancing data accessibility in applications like vital monitoring and life support alerts in +the metaverse. Blockchain technology’s decentralized nature enables data scientists in the metaverse to work together and on +data cleaning, which will greatly minimize the time and expenses involved with labeling data and getting datasets ready for +analytics [145]. +Data Sharing: Blockchain technology has the potential to increase the accuracy and transparency of transactions in the +metaverse for applications such as education and cryptocurrency trading [118]. Stakeholders would be able to access a +decentralized, unchangeable record of all transactions created by applications like governance and finance. Therefore, increased +data openness will be advantageous to the metaverse’s stakeholders [146]. Users’ confidence will increase as a result of being +able to comprehend how third-party programs like Thunderbird, the Bat, and Pegasus manage data thanks to blockchain +technology, which can also reduce grey market transactions [147]. The owner of the data will also have total control over +the data. Distributed ledger technology can also be useful for data audits. Blockchain thus saves time and money by reducing +the need for data validation [148]. The flexibility of data sharing will be increased through smart contracts. Usually, they are +used to automate the execution of a contract so that all parties may be sure of the result right away, without the need for an +intermediary or a waste of time. The diverse programming of smart contracts is made possible by blockchain. As a result, +programs like Nmusik, Ascribe, Tracr, UBS, and Applicature will benefit [149]. +Data Interoperability: A cross-chain protocol is the ideal approach to guarantee interoperability across virtual worlds in +the metaverse [150], [151]. This enables the transfer of goods like avatars, NFTs, and money between virtual worlds. This +protocol will lay the foundation for broad acceptance of the metaverse. Cross-blockchain technology will make it possible for +virtual worlds to communicate with one another, doing away with the necessity for middlemen in the metaverse [152]. In the +metaverse, connecting users and apps will be made simple via blockchain. +Data Privacy Preservation: Through the use of private and public keys, blockchain technology enables users of the metaverse +to govern their data, effectively giving them ownership over it. Third-party intermediates are prohibited from misusing or +obtaining data from other parties in the blockchain-enabled metaverse. Owners of personal data stored in the blockchain- +enabled metaverse will have control over when and how third parties can access that data [153]. Blockchain ledgers come +with an audit trail as a standard, guaranteeing that the transactions in the metaverse are comprehensive and consistent. Zero- +knowledge proof has been used on the blockchain, giving people easy access to the identification of crucial data in the metaverse +while keeping their privacy and control over their belongings. Blockchain technology uses zero-knowledge proofs as a method +for users to convince apps of something without having to provide the information [154]. +IoT: Through cross-chain networks, which are created by blockchain technology, IoT devices in the metaverse may exchange +data and create tamper-proof records of shared transactions in virtual worlds. Applications and individuals will be able to +exchange and access IoT data thanks to blockchain technology without the requirement for centralized administration or +control [155]. Each transaction is documented and validated in order to reduce disputes and boost user confidence throughout +the metaverse. IoT-enabled blockchain in the metaverse makes it possible to store data in real-time. Due to the immutability of +blockchain transactions, all stakeholders can depend on the information and respond quickly and effectively [156]. By allowing +stakeholders to manage their IoT data records in shared blockchain ledgers, blockchain technology can assist in resolving +problems in the metaverse. +Digital Twins: Digital twins are attack-resistant because to blockchain’s encryption capabilities and historical data openness, +which also allow for safe data sharing [157] across many virtual worlds. With the use of an intelligent distributed ledger, data +may be exchanged between digital twins in virtual environments. Using an intelligent distributed ledger, real-world items will +be saved on the blockchain and synced to digital twins in the metaverse. The implementation of digital twins on a blockchain +will also help to resolve problems with data security and privacy [158]. Tracking sensor data and creating high-caliber digital + +twins in the metaverse will be possible by combining blockchain and AI. Every digital twin activity in the metaverse will be +documented as a transaction on the blockchain, which is unchangeable and requires consensus to modify [157]. +AI: Blockchain-based encryption gives users of the metaverse total control over their data and makes it easy to transfer +ownership of AI consent to another entity. Through the use of zero-knowledge proofs, users may convince apps and other +parties that certain information about them is true without divulging this information to the applications themselves, granting +the authority to utilize data for AI model training. Blockchain ledgers frequently offer an audit trail that may be used to +verify the legitimacy of any transactions that take place in the metaverse. People can locate crucial metaverse facts via a +zero-knowledge evidence system while still maintaining their privacy and control of their resources against deepfakes [159]. +By doing this, AI will be stopped from wasting resources in the metaverse. +Big Data: By assisting in the collecting of data from reliable data sources, blockchain technology will help to reduce the +quantity of inaccurate data received. Data modification by other parties will be prohibited, and the data owners will have +complete control over their data. This guarantees high-quality data flows across the metaverse [135]. Data scientists in the +metaverse will be able to communicate and work together on data cleaning thanks to the decentralized nature of blockchain +technology. This will greatly cut down on the time and costs involved in categorizing data and building datasets for analytics +applications, as well as the danger of data contamination. Since the data will be copied across the network and the blockchain +is immutable, it will be impossible to alter it [160]. Data accessibility for metaverse stakeholders will therefore be improved. +Interactivity: A blockchain-based distributed ledger would make it possible to validate the records of holographic telep- +resence and other XR applications in the metaverse and track the origin of inaccurate data. As a result, a more precise +recommendation system will be created. The zero-trust mechanism and cross-chain technology of the blockchain will make +it simpler for holographic telepresence and other XR applications to safely transmit data between virtual worlds [133]. Data +integrity is guaranteed for XR apps and holographic telepresence by the interplanetary file system offered by the blockchain. The +consensus technique used by these devices will make the data they gather and store on a blockchain unchangeable. Blockchain +supports confidence among AR/VR stakeholders by facilitating transparent ownership transfer and asset verification [161]. +V. FUTURE VISIONS AND DIRECTIONS +A. Content-centric Metaverse +With the ever-growing increase in using MaaS platforms, UGC is expected to be increasingly generated and transmitted +through the metaverse networks. Current IP-based host oriented content transmission protocols will face critical challenges for +securing UGC dissemination by heterogeneous end devices over the metaverse platforms. To address this challenge, content +centric networking (CCN) can be employed to rethink the current Internet architecture. According to CCN, contents will +be routed directly by their naming information instead of IP addresses [162]. For data and content sharing in CCN-based +metaverse, users request the desired UGC via sending an interest message to any CCN based node that occupies the matched +content. The main idea for securing CCN is to directly safeguard the security of every single content/data itself, rather than +securing the “pipe” or the communication channels/links [163]. In this way, flexible and content-centric secure metaverse can +be realized. Due to the inherent attributes of the CCN architectures, CCN-based metaverse can explicitly cause new security +concerns as well. For instance, content poisoning and network monitoring would be two security issues in the CCN-based +metaverse. Specifically, malicious users might inject poisoned UGCs, resulting in delayed or failed valid UGC delivery, e.g., +through flooding. A curious CCN node might observe the sensitive content disseminated by CCN users by directly monitoring +the network traffic. Therefore, further research on privacy and security protection for CCN-based metaverse is required [164]. +B. Edge computing +Hospitals, production lines, residential complexes, offshore equipment and game centers have different requirements. Hence, +solutions based on edge computing in the Metaverse should be adapted for different applications. This work requires a +methodical, repeatable and well-reasoned approach from the Metaverse operators. Otherwise, the cost of maintaining and +managing edge servers will increase and many problems will arise for users and the network. At the same time, the variety of +services and flexibility in service level agreements should still exist so that users have a better user experience. +For various applications, edge computing can be used to increase reliability, reduce latency, and offload tasks in 5G networks +[165]. More interestingly, edge computing is one of the main enablers of the 6G network because it can be used for reliable +low-latency communications, AI-empowered capabilities, and increasing energy efficiency [166]. In addition, edge computing +plays a vital role in the new generations of the Internet of Things [167]. Hence, we expect edge computing to progress along +with related technologies and add new capabilities to the metaverse. +In designing edge computing solutions, various tradeoffs must be considered. Processing power, storage volume, telecommu- +nication link bandwidth, spare parts, emergency power, security systems and many other things must be selected depending on +the needs of users. In addition, it is possible that the needs of users change over time. A scalable and flexible design approach +can help with this. In this type of design, it is possible to increase and decrease each of these facilities, depending on the +conditions. For example, it is not necessary for the Metaverse operator to have installed a lot of storage space on the edge + +sites from the beginning. However, it should have provided the possibility of increasing storage space on the edge servers. +This would enable the network to meet this critical requirement in the shortest possible time with the increase in users’ needs. +C. Blockchain’s still unsolved challenges +Despite many challenges which were discussed in this paper for MaaS developers and the solutions that blockchain technology +can provide, there are still more constraints that should be taken into consideration. There is still work to be done on creating a +robust blockchain for the metaverse to overcome unsolved problems. In the following, the authors mention some other unsolved +challenges and make some recommendations as well. +• Blockchain can be slow because of its complexity and distributed nature. +• On a blockchain, transactions can take a very long time to execute. +• Data in the metaverse will be more able to withstand copying and manipulation with the help of a consensus-based +distributed ledger, but since any new data must be duplicated throughout the entire chain, more study is needed to +overcome the latency problem. +• The number of blocks must grow along with the number of users in the metaverse, requiring the employment of enormous +computational resources [168]. As a result, users will pay a higher transaction cost for the verification of shared transactions. +For effective data sharing in the metaverse, next-generation blockchains must overcome this problem [169]. +• The existence of numerous public blockchains in various virtual reality environments that do not speak the same language +presents the biggest obstacle to cross-blockchain enabled the metaverse interoperability. It will be challenging to adjust +because different platforms will offer different degrees of smart contract capabilities. Furthermore, these virtual worlds +use a wide range of transaction architectures and consensus mechanisms, which limits interoperability [170]. +• If a small number of miners control the majority of the network’s overall mining hash rate, blockchains are susceptible. +• Due to the anonymity offered by blockchain technology, it is challenging to track down all IoT transactions involving +illicit services in the metaverse. +• To carry out the metaverse’s expansion, the blockchain needs to be regularized. +• For blockchain to be successfully used in digital twin applications in the metaverse, challenges like standardization, +privacy, and scalability must all be resolved. +• The quality of digital twins in the metaverse will increase as a result of the integration of blockchain, XAI, and federated +learning methodologies. +VI. CONCLUSIONS +This article provided an overview on privacy and security aspects of the metaverse, from different perspectives, including +the wireless access, learning algorithms, data access, and human-centric interactions. New directions towards realizing privacy- +aware and secure metaverse-as-a-service (MaaS) platforms were addressed, and less-investigated methods were reviewed to +help mobile network operators and service providers facilitate the realization of secure and private MaaS through different +layers of the metaverse, ranging from the access layer to privatizing the social interactions among clients. Additionally, edge +computing, which is one of the key enablers of the metaverse, has been discussed, along with the advantages and challenges +associated with its use in the metaverse. Later in this work, a comprehensive investigation and analyses of challenges for MaaS +developers and the blockchain’s solutions for MaaS platforms were provided. At the final, future vision, unsolved challenges, +and some recommendations were also discussed to bring further insights for the network operators and engineers in the era of +metaverse. +REFERENCES +[1] J. Joshua, “Information bodies: Computational anxiety in neal stephenson’s snow crash,” Interdisciplinary Literary Studies, vol. 19, no. 1, pp. 17–47, +2017. +[2] B. +Marr, +“The +10 +best +examples +of +the +metaverse +everyone +should +know +about.” +https://www.forbes.com/sites/bernardmarr/2022/05/16/ +the-10-best-examples-of-the-metaverse-everyone-should-know-about/, 2022. +[3] McKinsey, “Value creation in the metaverse: The real business of the virtual world,” tech. rep., McKinsey and Company, https://www.mckinsey.com/ +business-functions/growth-marketing-and-sales/our-insights/value-creation-in-the-metaverse, June 2022. +[4] X.-T. Nguyen and J. A. Maine, “Taxing facebook code: Debugging the tax code and software,” Buff. L. Rev., vol. 60, p. 1, 2012. +[5] P. Banerjee, R. Friedrich, C. Bash, P. Goldsack, B. Huberman, J. Manley, C. Patel, P. Ranganathan, and A. Veitch, “Everything as a service: Powering +the new information economy,” Computer, vol. 44, no. 3, pp. 36–43, 2011. +[6] F. Hu, Y. Deng, W. Saad, M. Bennis, and A. H. Aghvami, “Cellular-connected wireless virtual reality: Requirements, challenges, and solutions,” IEEE +Communications Magazine, vol. 58, no. 5, pp. 105–111, 2020. +[7] Q. Yang, Y. Zhao, H. Huang, Z. Xiong, J. Kang, and Z. Zheng, “Fusing blockchain and ai with metaverse: A survey,” IEEE Open Journal of the +Computer Society, vol. 3, pp. 122–136, 2022. +[8] P. Porambage, G. G¨ur, D. P. M. Osorio, M. Liyanage, A. Gurtov, and M. Ylianttila, “The roadmap to 6g security and privacy,” IEEE Open Journal of +the Communications Society, vol. 2, pp. 1094–1122, 2021. +[9] A. Chorti, A. N. Barreto, S. K¨opsell, M. Zoli, M. Chafii, P. Sehier, G. Fettweis, and H. V. Poor, “Context-aware security for 6g wireless: the role of +physical layer security,” IEEE Communications Standards Magazine, vol. 6, no. 1, pp. 102–108, 2022. +[10] S. Kadhe, N. Rajaraman, O. O. Koyluoglu, and K. Ramchandran, “Fastsecagg: Scalable secure aggregation for privacy-preserving federated learning,” +arXiv preprint arXiv:2009.11248, 2020. + +[11] B. Falchuk, S. Loeb, and R. Neff, “The social metaverse: Battle for privacy,” IEEE Technology and Society Magazine, vol. 37, no. 2, pp. 52–61, 2018. +[12] T. Hunter, “Surveillance will follow us into ‘the metaverse,’ and our bodies could be its new data source.” https://www.washingtonpost.com/technology/ +2022/01/13/privacy-vr-metaverse/, 2022. +[13] M. Letafati, H. Behroozi, B. H. Khalaj, and E. A. Jorswieck, “On learning-assisted content-based secure image transmission for delay-aware systems +with randomly-distributed eavesdroppers,” IEEE Transactions on Communications, vol. 70, no. 2, pp. 1125–1139, 2021. +[14] M. Letafati, A. Kuhestani, K.-K. Wong, and M. J. Piran, “A lightweight secure and resilient transmission scheme for the internet of things in the +presence of a hostile jammer,” IEEE Internet of Things Journal, vol. 8, no. 6, pp. 4373–4388, 2020. +[15] M. Letafati, H. Behroozi, B. H. Khalaj, and E. A. Jorswieck, “Deep learning for hardware-impaired wireless secret key generation with man-in-the-middle +attacks,” in 2021 IEEE Global Communications Conference (GLOBECOM), pp. 1–6, IEEE, 2021. +[16] M. Letafati, H. Behroozi, B. H. Khalaj, and E. A. Jorswieck, “Hardware-impaired phy secret key generation with man-in-the-middle adversaries,” IEEE +Wireless Communications Letters, vol. 11, no. 4, pp. 856–860, 2022. +[17] F. Sun, W. Zang, H. Huang, I. Farkhatdinov, and Y. Li, “Accelerometer-based key generation and distribution method for wearable iot devices,” IEEE +Internet of Things Journal, vol. 8, no. 3, pp. 1636–1650, 2020. +[18] Z. Chen, W. Ren, Y. Ren, and K.-K. R. Choo, “Lirek: A lightweight and real-time key establishment scheme for wearable embedded devices by gestures +or motions,” Future Generation Computer Systems, vol. 84, pp. 126–138, 2018. +[19] J. Zhang, Y. Zheng, W. Xu, and Y. Chen, “H2k: A heartbeat-based key generation framework for ecg and ppg signals,” IEEE Transactions on Mobile +Computing, 2021. +[20] M. Letafati, H. Behroozi, B. H. Khalaj, and E. A. Jorswieck, “Wireless-powered cooperative key generation for e-health: A reservoir learning approach,” +in 2022 IEEE 95th Vehicular Technology Conference:(VTC2022-Spring), pp. 1–7, IEEE, 2022. +[21] J. Geiping, L. H. Fowl, G. Somepalli, M. Goldblum, M. Moeller, and T. Goldstein, “What doesn’t kill you makes you robust (er): How to adversarially +train against data poisoning,” 2021. +[22] Y. Shi and Y. E. Sagduyu, “Adversarial machine learning for flooding attacks on 5g radio access network slicing,” in 2021 IEEE International Conference +on Communications Workshops (ICC Workshops), pp. 1–6, IEEE, 2021. +[23] Y. Shi and Y. Sagduyu, “Membership inference attack and defense for wireless signal classifiers with deep learning,” IEEE Transactions on Mobile +Computing, 2022. +[24] J. Geiping, H. Bauermeister, H. Dr¨oge, and M. Moeller, “Inverting gradients-how easy is it to break privacy in federated learning?,” Advances in Neural +Information Processing Systems, vol. 33, pp. 16937–16947, 2020. +[25] E. Ayanoglu, K. Davaslioglu, and Y. E. Sagduyu, “Machine learning in nextg networks via generative adversarial networks,” IEEE Transactions on +Cognitive Communications and Networking, 2022. +[26] Y. Shi, Y. E. Sagduyu, T. Erpek, and M. C. Gursoy, “How to attack and defend 5g radio access network slicing with reinforcement learning,” arXiv +preprint arXiv:2101.05768, 2021. +[27] V. Nair, G. M. Garrido, and D. Song, “Exploring the unprecedented privacy risks of the metaverse,” arXiv preprint arXiv:2207.13176, 2022. +[28] V. Nair, G. M. Garrido, and D. Song, “Going incognito in the metaverse,” arXiv preprint arXiv:2208.05604, 2022. +[29] S. +Castellanos, +“Fake +it +to +make +it: +Companies +beef +up +ai +models +with +synthetic +data.” +https://www.wsj.com/articles/ +fake-it-to-make-it-companies-beef-up-ai-models-with-synthetic-data-11627032601, 2021. +[30] S. Solaiyappan and Y. Wen, “Machine learning based medical image deepfake detection: A comparative study,” Machine Learning with Applications, +vol. 8, p. 100298, 2022. +[31] S. J. Nightingale and H. Farid, “Ai-synthesized faces are indistinguishable from real faces and more trustworthy,” Proceedings of the National Academy +of Sciences, vol. 119, no. 8, p. e2120481119, 2022. +[32] P. +Tiwald, +“Tackling +ai +bias +at +its +source +with +fair +synthetic +data +(fairness +series +part +4).” +https://mostly.ai/blog/ +tackling-ai-bias-at-its-source-with-fair-synthetic-data-fairness-series-part-4/, 2020. +[33] P. Daugherty, M. Carrel-Billiard, and M. Biltz, “Tackling ai bias at its source with fair synthetic data (fairness series part 4).” https://www.accenture. +com/us-en/insights/technology/technology-trends-2022, 2020. +[34] L. B. Rosenberg, “Regulating the metaverse, a blueprint for the future,” in International Conference on Extended Reality, pp. 263–272, Springer, 2022. +[35] H. Wang, K. Sreenivasan, S. Rajput, H. Vishwakarma, S. Agarwal, J.-y. Sohn, K. Lee, and D. Papailiopoulos, “Attack of the tails: Yes, you really can +backdoor federated learning,” Advances in Neural Information Processing Systems, vol. 33, pp. 16070–16084, 2020. +[36] T. Li, A. K. Sahu, A. Talwalkar, and V. Smith, “Federated learning: Challenges, methods, and future directions,” IEEE Signal Processing Magazine, +vol. 37, no. 3, pp. 50–60, 2020. +[37] L. Jiang, H. Zheng, H. Tian, S. Xie, and Y. Zhang, “Cooperative federated learning and model update verification in blockchain empowered digital +twin edge networks,” IEEE Internet of Things Journal, 2021. +[38] J. Kang, D. Ye, J. Nie, J. Xiao, X. Deng, S. Wang, Z. Xiong, R. Yu, and D. Niyato, “Blockchain-based federated learning for industrial metaverses: +Incentive scheme with optimal aoi,” in 2022 IEEE International Conference on Blockchain (Blockchain), pp. 71–78, IEEE, 2022. +[39] Y. Lu, X. Huang, K. Zhang, S. Maharjan, and Y. Zhang, “Low-latency federated learning and blockchain for edge association in digital twin empowered +6g networks,” IEEE Transactions on Industrial Informatics, vol. 17, no. 7, pp. 5098–5107, 2020. +[40] L. Lyu, H. Yu, and Q. Yang, “Threats to federated learning: A survey,” arXiv preprint arXiv:2003.02133, 2020. +[41] G. Singh, “Best machine learning as a service platforms (mlaas) that you want to check as a data scientist.” https://neptune.ai/blog/ +best-machine-learning-as-a-service-platforms-mlaas, 2022. +[42] J. So, C. J. Nolet, C.-S. Yang, S. Li, Q. Yu, R. E Ali, B. Guler, and S. Avestimehr, “Lightsecagg: a lightweight and versatile design for secure +aggregation in federated learning,” Proceedings of Machine Learning and Systems, vol. 4, pp. 694–720, 2022. +[43] E. Bagdasaryan, A. Veit, Y. Hua, D. Estrin, and V. Shmatikov, “How to backdoor federated learning,” in International Conference on Artificial Intelligence +and Statistics, pp. 2938–2948, PMLR, 2020. +[44] C. Xie, K. Huang, P.-Y. Chen, and B. Li, “Dba: Distributed backdoor attacks against federated learning,” in International Conference on Learning +Representations, 2019. +[45] I. Goodfellow, P. McDaniel, and N. Papernot, “Making machine learning robust against adversarial inputs,” Communications of the ACM, vol. 61, no. 7, +pp. 56–66, 2018. +[46] A. Kurakin, I. Goodfellow, and S. Bengio, “Adversarial machine learning at scale,” arXiv preprint arXiv:1611.01236, 2016. +[47] V. Cherepanova, M. Goldblum, H. Foley, S. Duan, J. Dickerson, G. Taylor, and T. Goldstein, “Lowkey: Leveraging adversarial attacks to protect social +media users from facial recognition,” arXiv preprint arXiv:2101.07922, 2021. +[48] H. Duan, J. Li, S. Fan, Z. Lin, X. Wu, and W. Cai, “Metaverse for social good: A university campus prototype,” in Proceedings of the 29th ACM +International Conference on Multimedia, pp. 153–161, 2021. +[49] K. Antonakoglou, X. Xu, E. Steinbach, T. Mahmoodi, and M. Dohler, “Toward haptic communications over the 5g tactile internet,” IEEE Communications +Surveys & Tutorials, vol. 20, no. 4, pp. 3034–3059, 2018. +[50] L. E. Buck and B. Bodenheimer, “Privacy and personal space: Addressing interactions and interaction data as a privacy concern,” in 2021 IEEE +Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW), pp. 399–400, IEEE, 2021. + +[51] L. Buck and R. McDonnell, “Security and privacy in the metaverse: The threat of the digital human,” 2022. +[52] J. Wei, J. Li, Y. Lin, and J. Zhang, “Ldp-based social content protection for trending topic recommendation,” IEEE Internet of Things Journal, vol. 8, +no. 6, pp. 4353–4372, 2020. +[53] K. Schmidt, G. Munilla Garrido, A. M¨uhle, and C. Meinel, “Mitigating sovereign data exchange challenges: A mapping to apply privacy-and authenticity- +enhancing technologies,” in International Conference on Trust and Privacy in Digital Business, pp. 50–65, Springer, 2022. +[54] N. Schneider, P. De Filippi, S. Frey, J. Z. Tan, and A. X. Zhang, “Modular politics: Toward a governance layer for online communities,” Proceedings +of the ACM on Human-Computer Interaction, vol. 5, no. CSCW1, pp. 1–26, 2021. +[55] N. Reiff, “Decentralized autonomous organization (dao): Definition, purpose, and example.” https://www.investopedia.com/tech/what-dao/, 2022. +[56] D. Sabella, A. Reznik, and R. Frazao, Multi-access edge computing in action. CRC Press, 2019. +[57] T. Loozen and A. Baschnoga, “Seven ways telecom operators can power the metaverse,” tech. rep., EY - Global, https://www.ey.com/en gr/ +telecommunications/seven-ways-telecom-operators-can-power-the-metaverse, May 2022. +[58] Fact.MR, “Edge computing market.” https://www.factmr.com/report/4761/edge-computing-market, 2022. +[59] K. A. Kumari, G. S. Sadasivam, D. Dharani, and M. Niranjanamurthy, Edge Computing: Fundamentals, Advances and Applications. CRC Press, 2022. +[60] T. Kechadi, L. Chen, N. Aung, H. Ning, L. Atzori, et al., “Edge-enabled metaverse: The convergence of metaverse and mobile edge computing,” 2022. +[61] W. Y. B. Lim, Z. Xiong, D. Niyato, X. Cao, C. Miao, S. Sun, and Q. Yang, “Realizing the metaverse with edge intelligence: A match made in heaven,” +IEEE Wireless Communications, 2022. +[62] M. Xu, W. C. Ng, W. Y. B. Lim, J. Kang, Z. Xiong, D. Niyato, Q. Yang, X. S. Shen, and C. Miao, “A full dive into realizing the edge-enabled +metaverse: Visions, enabling technologies, and challenges,” arXiv preprint arXiv:2203.05471, 2022. +[63] A. Mehrabi, M. Siekkinen, T. Kamarainen, and A. yski, “Multi-tier cloudvr: Leveraging edge computing in remote rendered virtual reality,” ACM +Transactions on Multimedia Computing, Communications, and Applications (TOMM), vol. 17, no. 2, 2021. +[64] R. Xie, J. Fang, J. Yao, X. Jia, and K. Wu, “Sharing-aware task offloading of remote rendering for interactive applications in mobile edge computing,” +IEEE Transactions on Cloud Computing, 2021. +[65] R. Xie, J. Fang, J. Yao, K. Liu, X. Jia, and K. Wu, “Qos-aware scheduling of remote rendering for interactive multimedia applications in edge +computing,” IEEE Transactions on Parallel and Distributed Systems, 2022. +[66] Y. Pan, C. Wang, Y. Liu, C. Xu, Y. Liu, and L. Zhang, “5g mobile edge assisted metaverse light field video system: Prototype design and empirical +evaluation,” International Journal of Human-Computer Studies, 2022. +[67] S. Chen, Q. Li, M. Zhou, and A. Abusorrah, “Recent advances in collaborative scheduling of computing tasks in an edge computing paradigm,” Sensors, +vol. 21, no. 3, p. 779, 2021. +[68] M. Kamel, W. Hamouda, and A. Youssef, “Uplink coverage and capacity analysis of mmtc in ultra-dense networks,” IEEE Transactions on Vehicular +Technology, vol. 69, no. 1, pp. 746–759, 2019. +[69] T. N. G. Alliance, “6g applications and use cases.” https://nextgalliance.org/white papers/6g-applications-and-use-cases/, 2022. +[70] A. Mijuskovic, A. Chiumento, R. Bemthuis, A. Aldea, and P. Havinga, “Resource management techniques for cloud/fog and edge computing: An +evaluation framework and classification,” Sensors, vol. 21, no. 5, p. 1832, 2021. +[71] A. Al-Ansi, A. M. Al-Ansi, A. Muthanna, I. A. Elgendy, and A. Koucheryavy, “Survey on intelligence edge computing in 6g: Characteristics, challenges, +potential use cases, and market drivers,” Future Internet, vol. 13, no. 5, p. 118, 2021. +[72] F. Liu, G. Tang, Y. Li, Z. Cai, X. Zhang, and T. Zhou, “A survey on edge computing systems and tools,” Proceedings of the IEEE, vol. 107, no. 8, +pp. 1537–1562, 2019. +[73] Y. Jiang, J. Kang, D. Niyato, X. Ge, Z. Xiong, and C. Miao, “Reliable coded distributed computing for metaverse services: Coalition formation and +incentive mechanism design,” arXiv preprint arXiv:2111.10548, 2021. +[74] C. Preist, D. Schien, and P. Shabajee, “Evaluating sustainable interaction design of digital services: The case of youtube,” in Proceedings of the 2019 +CHI conference on human factors in computing systems, pp. 1–12, 2019. +[75] I. A. Elgendy and R. Yadav, “Survey on mobile edge-cloud computing: A taxonomy on computation offloading approaches,” in Security and Privacy +Preserving for IoT and 5G Networks, pp. 117–158, Springer, 2022. +[76] H. Lin, S. Zeadally, Z. Chen, H. Labiod, and L. Wang, “A survey on computation offloading modeling for edge computing,” Journal of Network and +Computer Applications, vol. 169, p. 102781, 2020. +[77] J. Mocnej, M. Miˇskuf, P. Papcun, and I. Zolotov´a, “Impact of edge computing paradigm on energy consumption in iot,” IFAC-PapersOnLine, vol. 51, +no. 6, pp. 162–167, 2018. +[78] C. Suisse, “Metaverse: A guide to the next-gen internet.” https://www.credit-suisse.com/media/assets/corporate/docs/about-us/media/media-release/2022/ +03/metaverse-14032022.pdf, 2022. +[79] I. Petri, O. F. Rana, J. Bignell, S. Nepal, and N. Auluck, “Incentivising resource sharing in edge computing applications,” in International Conference +on the Economics of Grids, Clouds, Systems, and Services, pp. 204–215, Springer, 2017. +[80] J. Liu, K. Luo, Z. Zhou, and X. Chen, “Erp: Edge resource pooling for data stream mobile computing,” IEEE Internet of Things Journal, vol. 6, no. 3, +pp. 4355–4368, 2018. +[81] X. Huang, B. Zhang, and C. Li, “Incentive mechanisms for mobile edge computing: Present and future directions,” IEEE Network, 2022. +[82] W. C. Ng, W. Y. B. Lim, J. S. Ng, Z. Xiong, D. Niyato, and C. Miao, “Unified resource allocation framework for the edge intelligence-enabled +metaverse,” arXiv preprint arXiv:2110.14325, 2021. +[83] J. Singh, Y. Bello, A. R. Hussein, A. Erbad, and A. Mohamed, “Hierarchical security paradigm for iot multiaccess edge computing,” IEEE Internet of +Things Journal, vol. 8, no. 7, pp. 5794–5805, 2020. +[84] K. A. Kumari, A. Sharma, C. Chakraborty, and M. Ananyaa, “Preserving health care data security and privacy using carmichael’s theorem-based +homomorphic encryption and modified enhanced homomorphic encryption schemes in edge computing systems,” Big Data, vol. 10, no. 1, pp. 1–17, +2022. +[85] Q.-V. Pham, X.-Q. Pham, T. T. Nguyen, Z. Han, D.-S. Kim, et al., “Artificial intelligence for the metaverse: A survey,” arXiv e-prints, pp. arXiv–2202, +2022. +[86] P. Dong, Q. Wu, X. Zhang, and G. Ding, “Edge semantic cognitive intelligence for 6g networks: Novel theoretical models, enabling framework, and +typical applications,” China Communications, vol. 19, no. 8, pp. 1–14, 2022. +[87] R.-S. Cheng, C.-M. Huang, and H.-H. Chen, “Md2do: A channel selection method for device to device communication using the mobile edge computing +(mec) paradigm,” Mobile Networks and Applications, vol. 27, no. 1, pp. 15–32, 2022. +[88] L. Chang, Z. Zhang, P. Li, S. Xi, W. Guo, Y. Shen, Z. Xiong, J. Kang, D. Niyato, X. Qiao, et al., “6g-enabled edge ai for metaverse: Challenges, +methods, and future research directions,” Journal of Communications and Information Networks, 2022. +[89] M. Marvi, A. Aijaz, and M. Khurram, “Toward an automated data offloading framework for multi-rat 5g wireless networks,” IEEE Transactions on +Network and Service Management, vol. 17, no. 4, pp. 2584–2597, 2020. +[90] Q. Wang, H. Li, H. Wang, J. Zhang, and J. Fu, “A remote calibration device using edge intelligence,” Sensors, vol. 22, no. 1, p. 322, 2022. +[91] V. Avelar, “Digital remote monitoring and dispatch services’ impact on edge computing and data centers.” https://www.se.com, 2022. + +[92] Y. Jiang, J. Kang, D. Niyato, X. Ge, Z. Xiong, C. Miao, and X. Shen, “Reliable distributed computing for metaverse: A hierarchical game-theoretic +approach,” IEEE Transactions on Vehicular Technology, 2022. +[93] L. edg, “Reduce your data center cooling costs by 96 percent with servers built for the heat.” https://lanternedge.com/, 2022. +[94] X. Yu, X. Yang, Q. Tan, C. Shan, and Z. Lv, “An edge computing based anomaly detection method in iot industrial sustainability,” Applied Soft +Computing, p. 109486, 2022. +[95] N. Ghantous and C. Fakhri, “Empowering metaverse through machine learning and blockchain technology: A study on machine learning, blockchain, +and their combination to enhance metaverse,” ScienceOpen Preprints, 2022. +[96] J. Mendez, K. Bierzynski, M. Cu´ellar, and D. P. Morales, “Edge intelligence: Concepts, architectures, applications and future directions,” ACM +Transactions on Embedded Computing Systems (TECS), 2022. +[97] M. V. Jamali, H. Saber, H. Hatami, and J. H. Bae, “Productae: Toward training larger channel codes based on neural product codes,” in ICC 2022-IEEE +International Conference on Communications, pp. 3898–3903, IEEE, 2022. +[98] J. G. Almaraz-Rivera, J. A. Perez-Diaz, and J. A. Cantoral-Ceballos, “Transport and application layer ddos attacks detection to iot devices by using +machine learning and deep learning models,” Sensors, vol. 22, no. 9, p. 3367, 2022. +[99] H. G. Abreha, M. Hayajneh, and M. A. Serhani, “Federated learning in edge computing: a systematic survey,” Sensors, vol. 22, no. 2, p. 450, 2022. +[100] C. Thapa, P. C. M. Arachchige, S. Camtepe, and L. Sun, “Splitfed: When federated learning meets split learning,” in Proceedings of the AAAI Conference +on Artificial Intelligence, vol. 36, pp. 8485–8493, 2022. +[101] J. Tu, J. Zhou, and D. Ren, “An asynchronous distributed training algorithm based on gossip communication and stochastic gradient descent,” Computer +Communications, 2022. +[102] C. P. Filho, E. Marques Jr, V. Chang, L. Dos Santos, F. Bernardini, P. F. Pires, L. Ochi, and F. C. Delicato, “A systematic literature review on distributed +machine learning in edge computing,” Sensors, vol. 22, no. 7, p. 2665, 2022. +[103] V. Avelar, “Cost benefit analysis of edge micro data center deployments,” Schneider Electric, 2017. +[104] X. Luo, H.-H. Chen, and Q. Guo, “Semantic communications: Overview, open issues, and future research directions,” IEEE Wireless Communications, +2022. +[105] J. Zhou and X. Zhang, “Fairness-aware task offloading and resource allocation in cooperative mobile-edge computing,” IEEE Internet of Things Journal, +vol. 9, no. 5, pp. 3812–3824, 2021. +[106] T. Jahani-Nezhad and M. A. Maddah-Ali, “Berrut approximated coded computing: Straggler resistance beyond polynomial computing,” IEEE +Transactions on Pattern Analysis and Machine Intelligence, 2022. +[107] P. Ranaweera, A. D. Jurcut, and M. Liyanage, “Survey on multi-access edge computing security and privacy,” IEEE Communications Surveys & +Tutorials, vol. 23, no. 2, pp. 1078–1124, 2021. +[108] H. Guo and X. Yu, “A survey on blockchain technology and its security,” Blockchain: Research and Applications, vol. 3, no. 2, p. 100067, 2022. +[109] T. R. Gadekallu, T. Huynh-The, W. Wang, G. Yenduri, P. Ranaweera, Q.-V. Pham, D. B. da Costa, and M. Liyanage, “Blockchain for the metaverse: +A review,” arXiv preprint arXiv:2203.09738, 2022. +[110] I. Calzada, “Data co-operatives through data sovereignty,” Smart Cities, vol. 4, no. 3, pp. 1158–1172, 2021. +[111] L. Brunschwig, R. Campos-L´opez, E. Guerra, and J. de Lara, “Towards domain-specific modelling environments based on augmented reality,” in 2021 +IEEE/ACM 43rd International Conference on Software Engineering: New Ideas and Emerging Results (ICSE-NIER), pp. 56–60, IEEE, 2021. +[112] J.-B. Jeong, S. Lee, and E.-S. Ryu, “Rethinking fatigue-aware 6dof video streaming: Focusing on mpeg immersive video,” in 2022 International +Conference on Information Networking (ICOIN), pp. 304–309, IEEE, 2022. +[113] W.-L. Shiau and L.-C. Huang, “Scale development for analyzing the fit of real and virtual world integration: an example of pok´emon go,” Information +Technology & People, 2022. +[114] Y. Bian, J. Leng, and J. L. Zhao, “Demystifying metaverse as a new paradigm of enterprise digitization,” in International Conference on Big Data, +pp. 109–119, Springer, 2021. +[115] X. Wang, X. Liu, C.-T. Cheng, L. Deng, X. Chen, and F. Xiao, “A joint user scheduling and trajectory planning data collection strategy for the +uav-assisted wsn,” IEEE Communications Letters, vol. 25, no. 7, pp. 2333–2337, 2021. +[116] L. V. Kiong, Metaverse Made Easy: A Beginner’s Guide to the Metaverse: Everything you need to know about Metaverse, NFT and GameFi. Liew +Voon Kiong, 2022. +[117] L. Liu, J. Feng, Q. Pei, C. Chen, Y. Ming, B. Shang, and M. Dong, “Blockchain-enabled secure data sharing scheme in mobile-edge computing: an +asynchronous advantage actor–critic learning approach,” IEEE Internet of Things Journal, vol. 8, no. 4, pp. 2342–2353, 2020. +[118] B. Egliston and M. Carter, “Critical questions for facebook’s virtual reality: data, power and the metaverse,” Internet Policy Review, vol. 10, no. 4, +2021. +[119] K. Yu, L. Tan, M. Aloqaily, H. Yang, and Y. Jararweh, “Blockchain-enhanced data sharing with traceable and direct revocation in iiot,” IEEE transactions +on industrial informatics, vol. 17, no. 11, pp. 7669–7678, 2021. +[120] S. Mystakidis, “Metaverse,” Encyclopedia, vol. 2, no. 1, pp. 486–497, 2022. +[121] I. Hughes, “The metaverse: Is it the future?,” ITNOW, vol. 64, no. 1, pp. 22–23, 2022. +[122] L. Zhang, F. Li, P. Wang, R. Su, and Z. Chi, “A blockchain-assisted massive iot data collection intelligent framework,” IEEE Internet of Things Journal, +2021. +[123] Y. Hajjaji, W. Boulila, I. R. Farah, I. Romdhani, and A. Hussain, “Big data and iot-based applications in smart environments: A systematic review,” +Computer Science Review, vol. 39, p. 100318, 2021. +[124] C. Zhuang, J. Gong, and J. Liu, “Digital twin-based assembly data management and process traceability for complex products,” Journal of manufacturing +systems, vol. 58, pp. 118–131, 2021. +[125] L. U. Khan, W. Saad, D. Niyato, Z. Han, and C. S. Hong, “Digital-twin-enabled 6g: Vision, architectural trends, and future directions,” IEEE +Communications Magazine, vol. 60, no. 1, pp. 74–80, 2022. +[126] B. K. Wiederhold, “Ready (or not) player one: initial musings on the metaverse,” 2022. +[127] V. Ahsani, A. Sharma, C. Hegde, S. Knickerbocker, and N. Hawkins, “Improving probe-based congestion performance metrics accuracy by using change +point detection,” Journal of Big Data Analytics in Transportation, vol. 2, no. 1, pp. 61–74, 2020. +[128] V. Ahsani, M. Amin-Naseri, S. Knickerbocker, and A. Sharma, “Quantitative analysis of probe data characteristics: Coverage, speed bias and congestion +detection precision,” Journal of Intelligent Transportation Systems, vol. 23, no. 2, pp. 103–119, 2019. +[129] A. Sharma and V. Ahsani, “Assessing the impact of game day schedule and opponents on travel patterns and route choice using big data analytics,” +2019. +[130] P. Chakraborty, Y. O. Adu-Gyamfi, S. Poddar, V. Ahsani, A. Sharma, and S. Sarkar, “Traffic congestion detection from camera images using deep +convolution neural networks,” Transportation Research Record, vol. 2672, no. 45, pp. 222–231, 2018. +[131] S. Poddar, K. Ozcan, P. Chakraborty, V. Ahsani, A. Sharma, and S. Sarkar, “Comparison of machine learning algorithms to determine traffic congestion +from camera images,” in Transportation Research Board 97th annual meeting, Washington, DC, pp. 7–11, 2018. +[132] A. Sharma, V. Ahsani, S. Rawat, et al., “Evaluation of opportunities and challenges of using inrix data for real-time performance monitoring and +historical trend assessment,” 2017. + +[133] P. Bhattacharya, D. Saraswat, A. Dave, M. Acharya, S. Tanwar, G. Sharma, and I. E. Davidson, “Coalition of 6g and blockchain in ar/vr space: +Challenges and future directions,” IEEE Access, vol. 9, pp. 168455–168484, 2021. +[134] A. Islam and S. Y. Shin, “Buav: A blockchain based secure uav-assisted data acquisition scheme in internet of things,” Journal of Communications +and Networks, vol. 21, no. 5, pp. 491–502, 2019. +[135] N. Deepa, Q.-V. Pham, D. C. Nguyen, S. Bhattacharya, B. Prabadevi, T. R. Gadekallu, P. K. R. Maddikunta, F. Fang, and P. N. Pathirana, “A survey +on blockchain for big data: approaches, opportunities, and future directions,” Future Generation Computer Systems, 2022. +[136] C. Xu, Y. Qu, T. H. Luan, P. W. Eklund, Y. Xiang, and L. Gao, “A lightweight and attack-proof bidirectional blockchain paradigm for internet of +things,” IEEE Internet of Things Journal, vol. 9, no. 6, pp. 4371–4384, 2021. +[137] S. Bouraga, “A taxonomy of blockchain consensus protocols: A survey and classification framework,” Expert Systems with Applications, vol. 168, +p. 114384, 2021. +[138] B. Lashkari and P. Musilek, “A comprehensive review of blockchain consensus mechanisms,” IEEE Access, vol. 9, pp. 43620–43652, 2021. +[139] Y. Luo, Z. Su, W. Zheng, Z. Chen, F. Wang, Z. Zhang, and J. Chen, “A novel memory-hard password hashing scheme for blockchain-based cyber-physical +systems,” ACM Transactions on Internet Technology (TOIT), vol. 21, no. 2, pp. 1–21, 2021. +[140] L. Zhang, Z. Zhang, W. Wang, Z. Jin, Y. Su, and H. Chen, “Research on a covert communication model realized by using smart contracts in blockchain +environment,” IEEE Systems Journal, 2021. +[141] J. Guo, X. Ding, and W. Wu, “Reliable traffic monitoring mechanisms based on blockchain in vehicular networks,” IEEE Transactions on Reliability, +2021. +[142] W. Liang, Y. Fan, K.-C. Li, D. Zhang, and J.-L. Gaudiot, “Secure data storage and recovery in industrial blockchain network environments,” IEEE +Transactions on Industrial Informatics, vol. 16, no. 10, pp. 6543–6552, 2020. +[143] H.-j. Jeon, H.-c. Youn, S.-m. Ko, and T.-h. Kim, “Blockchain and ai meet in the metaverse,” Advances in the Convergence of Blockchain and Artificial +Intelligence, vol. 73, 2022. +[144] D. Yang, J. Zhou, R. Chen, Y. Song, Z. Song, X. Zhang, Q. Wang, K. Wang, C. Zhou, J. Sun, et al., “Expert consensus on the metaverse in medicine,” +Clinical eHealth, vol. 5, pp. 1–9, 2022. +[145] L. +Inks, +“Can +blockchain +do +ai +and +data +labeling. +effect.ai +thinks +they +can, +and +will.” +https://medium.com/@lawrence.effect/ +can-blockchain-do-ai-and-data-labeling-effect-ai-thinks-they-can-and-will-9ea94229dbf1, 2019. +[146] A. Rashid, A. Masood, H. Abbas, and Y. Zhang, “Blockchain-based public key infrastructure: a transparent digital certification mechanism for secure +communication,” IEEE Network, vol. 35, no. 5, pp. 220–225, 2021. +[147] N. Vashistha, M. M. Hossain, M. R. Shahriar, F. Farahmandi, F. Rahman, and M. M. Tehranipoor, “echain: A blockchain-enabled ecosystem for +electronic device authenticity verification,” IEEE Transactions on Consumer Electronics, vol. 68, no. 1, pp. 23–37, 2021. +[148] T. Min and W. Cai, “Portrait of decentralized application users: an overview based on large-scale ethereum data,” CCF Transactions on Pervasive +Computing and Interaction, pp. 1–18, 2022. +[149] O. Ali, A. Jaradat, A. Kulakli, and A. Abuhalimeh, “A comparative study: Blockchain technology utilization benefits, challenges and functionalities,” +IEEE Access, vol. 9, pp. 12730–12749, 2021. +[150] R. Belchior, A. Vasconcelos, S. Guerreiro, and M. Correia, “A survey on blockchain interoperability: Past, present, and future trends,” ACM Computing +Surveys (CSUR), vol. 54, no. 8, pp. 1–41, 2021. +[151] M. Madine, K. Salah, R. Jayaraman, Y. Al-Hammadi, J. Arshad, and I. Yaqoob, “appxchain: Application-level interoperability for blockchain networks,” +IEEE Access, vol. 9, pp. 87777–87791, 2021. +[152] R. Jabbar, N. Fetais, M. Krichen, and K. Barkaoui, “Blockchain technology for healthcare: Enhancing shared electronic health record interoperability +and integrity,” in 2020 IEEE International Conference on Informatics, IoT, and Enabling Technologies (ICIoT), pp. 310–317, IEEE, 2020. +[153] P. Kumar, R. Kumar, G. Srivastava, G. P. Gupta, R. Tripathi, T. R. Gadekallu, and N. N. Xiong, “Ppsf: a privacy-preserving and secure framework using +blockchain-based machine-learning for iot-driven smart cities,” IEEE Transactions on Network Science and Engineering, vol. 8, no. 3, pp. 2326–2341, +2021. +[154] J. Sedlmeir, F. V¨olter, and J. Str¨uker, “The next stage of green electricity labeling: using zero-knowledge proofs for blockchain-based certificates of +origin and use,” ACM SIGENERGY Energy Informatics Review, vol. 1, no. 1, pp. 20–31, 2021. +[155] U. Majeed, L. U. Khan, I. Yaqoob, S. A. Kazmi, K. Salah, and C. S. Hong, “Blockchain for iot-based smart cities: Recent advances, requirements, +and future challenges,” Journal of Network and Computer Applications, vol. 181, p. 103007, 2021. +[156] A. Dorri, F. Luo, S. Karumba, S. Kanhere, R. Jurdak, and Z. Y. Dong, “Temporary immutability: A removable blockchain solution for prosumer-side +energy trading,” Journal of Network and Computer Applications, vol. 180, p. 103018, 2021. +[157] D. Lee, S. H. Lee, N. Masoud, M. Krishnan, and V. C. Li, “Integrated digital twin and blockchain framework to support accountable information +sharing in construction projects,” Automation in construction, vol. 127, p. 103688, 2021. +[158] W. Shen, T. Hu, C. Zhang, and S. Ma, “Secure sharing of big digital twin data for smart manufacturing based on blockchain,” Journal of Manufacturing +Systems, vol. 61, pp. 338–350, 2021. +[159] A. A. Hussain and F. Al-Turjman, “Artificial intelligence and blockchain: A review,” Transactions on emerging telecommunications technologies, vol. 32, +no. 9, p. e4268, 2021. +[160] D. M. Gligor, K. G. Pillai, and I. Golgeci, “Theorizing the dark side of business-to-business relationships in the era of ai, big data, and blockchain,” +Journal of Business Research, vol. 133, pp. 79–88, 2021. +[161] S. Kumar, A. K. Bharti, and R. Amin, “Decentralized secure storage of medical records using blockchain and ipfs: A comparative analysis with future +directions,” Security and Privacy, vol. 4, no. 5, p. e162, 2021. +[162] R. Tourani, S. Misra, T. Mick, and G. Panwar, “Security, privacy, and access control in information-centric networking: A survey,” IEEE communications +surveys & tutorials, vol. 20, no. 1, pp. 566–600, 2017. +[163] C. Ghali, G. Tsudik, and C. A. Wood, “(the futility of) data privacy in content-centric networking,” in Proceedings of the 2016 ACM on Workshop on +Privacy in the Electronic Society, pp. 143–152, 2016. +[164] Z. Yang, X. Li, L. Wei, C. Zhang, and C. Gu, “Sgx-icn: A secure and privacy-preserving information-centric networking with sgx enclaves,” in 2020 +3rd International Conference on Hot Information-Centric Networking (HotICN), pp. 142–147, IEEE, 2020. +[165] R. W. Coutinho and A. Boukerche, “Design of edge computing for 5g-enabled tactile internet-based industrial applications,” IEEE Communications +Magazine, vol. 60, no. 1, pp. 60–66, 2022. +[166] A. Bandi, “A review towards ai empowered 6g communication requirements, applications, and technologies in mobile edge computing,” in 2022 6th +International Conference on Computing Methodologies and Communication (ICCMC), pp. 12–17, IEEE, 2022. +[167] E. Commission, “The next generation internet of things.” https://digital-strategy.ec.europa.eu/en/policies/next-generation-internet-things, 2022. +[168] Y. Gao, W. Wu, P. Si, Z. Yang, and F. R. Yu, “B-rest: Blockchain-enabled resource sharing and transactions in fog computing,” IEEE Wireless +Communications, vol. 28, no. 2, pp. 172–180, 2021. +[169] A. Rahimi and M. A. Maddah-Ali, “Multi-party proof generation in qap-based zk-snarks,” IEEE Journal on Selected Areas in Information Theory, +vol. 2, no. 3, pp. 931–941, 2021. + +[170] S. Wibowo and T. Sandikapura, “Improving data security, interoperability, and veracity using blockchain for one data governance, case study of local +tax big data,” in 2019 International Conference on ICT for Smart Society (ICISS), vol. 7, pp. 1–6, IEEE, 2019. +