Electric Network Frequency as Environmental Fingerprint for Metaverse Security: A Comprehensive Survey

Abstract

The rapid expansion of the Metaverse presents complex security challenges, particularly in verifying virtual objects and avatars within immersive environments. Conventional authentication methods, such as passwords and biometrics, often prove inadequate in these dynamic environments, especially as essential infrastructures, such as smart grids, integrate with virtual platforms. Cybersecurity threats intensify as advanced attacks introduce fraudulent data, compromising system reliability and safety. Using the Electric Network Frequency (ENF), a naturally varying signal emitted from power grids, provides an innovative environmental fingerprint to authenticate digital twins and Metaverse entities in the smart grid. This paper provides a comprehensive survey of the ENF as an environmental fingerprint for enhancing Metaverse security, reviewing its characteristics, sensing methods, limitations, and applications in threat modeling and the CIA triad (Confidentiality, Integrity, and Availability), and presents a real-world case study to demonstrate its effectiveness in practical settings. By capturing the ENF as having a unique signature that is timestamped, this method strengthens security by directly correlating physical grid behavior and virtual interactions, effectively combating threats such as deepfake manipulations. Building upon recent developments in signal processing, this strategy reinforces the integrity of digital environments, delivering robust protection against evolving cyber–physical risks and facilitating secure, scalable virtual infrastructures.

1. Introduction

The rapid evolution of the Metaverse as a decentralized and immersive digital ecosystem underscores the need for novel security strategies to protect virtual environments, augmented reality interfaces, and digital twins (DTs) [1]. The Metaverse, characterized by the integration of physical and virtual realities, relies extensively on fluid interactions among users, avatars, and critical infrastructure systems, such as smart grids, which are increasingly represented as cyber–physical systems (CPSs) [2]. These CPSs incorporate computational elements within physical structures, such as power grids, making them particularly vulnerable to advanced cyberattacks that exploit their interconnectedness. Electric Network Frequency (ENF) signals, originating from natural-variable power grid operations, present a robust solution to address security, operational, and privacy challenges [3]. Unlike conventional static authentication approaches, the ENF offers a dynamic, geographically distinct signature grounded in the physical environment, establishing a secure link between real-world infrastructure and virtual constructs [4].

Subtle frequency deviations in power grids generate the ENF, typically standardized frequencies of around 50 Hz or 60 Hz, caused by fluctuations in supply and demand [5]. These variations propagate through interconnected grids, creating a unique time-varying signal due to real-time physical dependencies. The ENF has validated multimedia content in digital forensics by aligning embedded ENF patterns with reference databases [3]. Applying this methodology to the Metaverse enables the ENF to serve as a physical anchor, securing communication protocols, and ensuring immutable audit trails that link virtual actions to actual physical events [6]. For example, within the DTs of the smart grid, virtual models used for monitoring and predictive analysis, the ENF verifies the authenticity of the data, ensuring an accurate representation of physical conditions [7]. This capability becomes increasingly critical as the Metaverse becomes a central platform for managing essential infrastructure such as energy grids, industrial operations, and financial networks. The ENF has demonstrated significant reliability in multimedia forensics, achieving 92% accuracy in audio authentication by aligning embedded frequency patterns with reference databases [7]. However, direct sensing through Phasor Measurement Units (PMUs) faces challenges, with detection accuracies ranging from 18% to 38% due to grid-induced variability [8,9]. These strengths and limitations highlight the potential of the ENF as a dynamic security tool in the Metaverse, complementing traditional methods such as biometrics and passwords in immersive, decentralized environments.

The increasing complexity of security threats targeting the Metaverse, including identity fraud, unauthorized access, and systemic disruptions, exposes the shortcomings of traditional security frameworks [10,11,12,13,14]. These attacks frequently exploit the dynamic interconnections between virtual and physical domains, which require adaptive security measures. ENF-based strategies, augmented by machine learning techniques for anomaly detection, provide proactive defenses by analyzing real-time frequency data to pinpoint indicators of malicious activities [15]. For example, irregularities in ENF signals can be used to detect tampering efforts, such as false data injection (FDI) attacks, a prevalent vulnerability within CPSs [16]. Moreover, the Metaverse’s extensive processing of sensitive information, including user identities and operational data, raises significant concerns regarding privacy. When integrated with privacy-enhancing technologies such as differential privacy and homomorphic encryption, the ENF protects data confidentiality without sacrificing analytical effectiveness [17]. These privacy-focused methodologies use advanced cryptographic research to ensure compliance with standards, such as the General Data Protection Regulation (GDPR), and enhance user trust.

Combining the ENF with distributed technologies such as blockchain and federated learning (FL) further extends its potential for decentralized governance within the Metaverse [18]. Blockchain technology facilitates immutable logging of ENF data, reinforcing accountability in grid operations [19]. Currently, FL supports collaborative anomaly detection in multiple locations without compromising local data ownership [20]. These innovations resonate with the decentralized principles of the Metaverse, addressing its inherently diverse and scalable architecture. Traditionally, ENF analysis has used signal processing methods, such as the Fourier transform, to isolate frequency elements from noisy signals [21]. Contemporary techniques, including wavelet transforms and spectrogram analysis, enhance resolution and enable real-time applicability, positioning the ENF as an adaptable security tool for virtual settings [22]. This paper reviews cutting-edge applications, examines privacy-preserving techniques, and suggests future research paths, emphasizing the critical role of the ENF in Metaverse security. The ENF outlines a pathway toward secure, reliable virtual ecosystems amid a rapidly evolving digital context by anchoring digital interactions to tangible physical events.

The emergence of the Metaverse as a foundational infrastructure for next-generation applications, ranging from virtual asset management and industrial automation to telemedicine and digital education, requires security models that are context-aware and intrinsically tied to physical reality [1]. Unlike traditional cyberspace, where data integrity can often be enforced through static verification, the Metaverse demands continuous, real-time virtual interaction validation that reflects underlying physical states [23]. The ENF addresses this gap by offering an environmental fingerprint that cannot be easily forged or separated from real-world conditions, thus enabling a new class of security primitives grounded in physical laws. As adversaries increasingly employ AI-driven tools to fabricate virtual identities, alter data provenance, and bypass authentication mechanisms [24], the ENF emerges as a uniquely resilient signal, difficult to replicate, temporally dynamic, and inherently decentralized. This makes it especially suitable for securing the Metaverse’s layered architecture, where physical infrastructure, DTs, and immersive applications must interact reliably across geographically distributed networks.

The remainder of this paper is structured as follows. Section 3 establishes the foundation for understanding the characteristics of ENF signals and Metaverse security requirements. Section 4 presents ENF sensing methods, signal characteristics, and, most importantly, reliability limitations. Section 5 discusses threat modeling for Metaverse-integrated systems, ENF-driven security mechanisms that address the CIA triad (Confidentiality, Integrity, and Availability), and comprehensive mitigation strategies. Section 6 presents a detailed case study to demonstrate the practical application of the security framework. Section 7 demonstrates how the ENF principles apply beyond smart grids to virtual reality (VR) training, financial systems, and healthcare applications. Section 8 presents research opportunities that specifically address ENF reliability issues alongside emerging application areas. Finally, Section 9 acknowledges both the potential of the ENF and its current limitations, offering a realistic assessment of its role in Metaverse security.

Contributions and Novelty

This survey makes several key contributions to the fields of ENF-based forensics, smart grid security, and Metaverse applications:

• Unified analysis of ENF in emerging contexts: We provide a comprehensive review of the ENF’s strengths, limitations, and quantitative trade-offs (e.g., 95–99% detection accuracy in anomaly detection), extending beyond traditional multimedia forensics to Metaverse-integrated infrastructures like DTs in smart grids.
• Novel applications and frameworks: We explore the ENF’s role in Metaverse security, including confidentiality via dynamic encryption, data provenance assurance, and real-world anchoring for event validation, with detailed case studies such as the ANCHOR-Grid framework for countering deepfake attacks.
• Forward-looking challenges and recommendations: We outline research directions for hybrid integrations (e.g., with quantum neural networks and homomorphic encryption), addressing scalability, privacy, and sociotechnical implications in decentralized virtual environments.
• Empirical and visual enhancements: Through tables (e.g., performance comparisons) and figures (e.g., ENF correlation visualizations), we offer actionable insights, including mitigation strategies to improve ENF reliability by 10–20% in noisy scenarios.

Compared to previous surveys, this work systematically compares the reliability of ENF sensing methods across a range of Metaverse scenarios, integrates empirical findings from a real-world case study, and provides actionable guidance for addressing persistent challenges, such as privacy, interoperability, and cross-domain standardization. As the field rapidly evolves, a fresh synthesis is needed to guide researchers and practitioners facing novel cyber–physical threats in decentralized virtual environments. In addition, our work advances the literature by bridging gaps in interdisciplinary applications. For example, previous surveys on the ENF in multimedia forensics [18,25] focus primarily on extraction techniques and detection of tampering in audio/video, achieving high accuracy but without extending to virtual ecosystems. Surveys on smart grid security [26,27] emphasize cyber threats and defenses but rarely incorporate physical signals, such as the ENF, for authentication. In Metaverse security, recent overviews highlight privacy risks but overlook environmental fingerprints that ground virtual actions in physical reality [28]. Our survey uniquely integrates these domains, justifying its novelty through the ENF’s application as a dynamic, reliable anchor for Metaverse-scale challenges, with quantitative progress not covered in prior works, as shown in Figure 1.

2. Related Works

This section reviews the existing literature on Electric Network Frequency (ENF) applications, IoT/smart grid security, and Metaverse security challenges. By situating our survey within these domains, we highlight the gaps in integrating the ENF as a physical–environmental anchor for Metaverse ecosystems, justifying the relevance of our comprehensive exploration.

2.1. ENF in Multimedia Forensics

Figure 2 shows a timeline of the evolution of ENF research from forensics to Metaverse applications. The ENF has been extensively studied as a forensic tool for authenticating and timestamping multimedia recordings, leveraging the unique fluctuations in power grid frequencies (typically around 50 Hz or 60 Hz) that are embedded as subtle signals in audio or video [7]. A foundational survey by Ngharamike et al. (2023) provides a detailed overview of ENF-based digital multimedia forensics, covering extraction techniques such as spectrogram analysis and wavelet transforms, as well as applications in tampering detection and challenges, including noise interference and regional variability [3]. The work emphasizes the reliability of the ENF, achieving up to 92% accuracy in audio authentication by matching the extracted signals to the reference databases [29,30]. Recent advancements include the automated estimation of the ENF in static and nonstatic videos, with algorithms demonstrating robustness to motion artifacts and achieving mean absolute errors as low as 0.15 Hz [31]. In comprehensive studies, the factors affecting ENF matching are explored, such as environmental noise and device variability, highlighting the need for adaptive filtering to improve forensic accuracy [25]. Although these works establish the ENF as a mature tool in forensics, they primarily address standalone multimedia analysis, with limited extension to interconnected virtual environments.

2.2. ENF in IoT and Smart Grid Security

The application of the ENF in IoT and smart grid contexts is emerging, building on its forensic roots to enable anomaly detection and authentication in distributed systems. In smart grids, ENF signals can serve as timestamps for data integrity, particularly in detecting FDI attacks, where adversaries manipulate sensor readings to disrupt operations [32,33]. For example, ENF-based frameworks have been proposed for real-time monitoring, integrating with Phase Measurement Units (PMU) to verify data provenance with accuracies up to 95% [7,34]. The broader literature on IoT-enabled smart grids emphasizes security challenges but rarely incorporates the ENF specifically. Surveys highlight the role of the IoT in improving grid efficiency through real-time analytics while also noting vulnerabilities such as unauthorized access and cyber threats [35]. For example, increased connectivity expands attack surfaces, allowing malware to hijack devices and cause grid disruptions [36]. Solutions often involve blockchain or anomaly detection, but the ENF offers a complementary physical-layer anchor, as seen in fault detection systems for DC microgrids [37]. However, existing works lack the integration of the ENF with virtual representations, such as DTs, a gap that our survey addresses by exploring the potential of the ENF in Metaverse-linked infrastructures.

2.3. Security Challenges in the Metaverse

As illustrated in Figure 3, the Metaverse introduces novel security risks due to its immersive and decentralized nature, encompassing threats such as deepfakes, identity theft, and data breaches in virtual worlds [12,34]. A comprehensive survey by Wang et al. outlines privacy and security issues, including unauthorized access, phishing, and insecure designs, while proposing frameworks for blockchain-based defenses [38]. Recent analyses highlight evolving challenges, such as cyberattacks that exploit AR/VR interfaces and the fusion of physical and virtual realms, including the ENF, with predictions of increased threats [12,13]. Privacy concerns, such as personal data leakage in social interactions, are highlighted in scoping reviews, which note the absence of standardized regulations [39]. While these surveys cover Metaverse threats holistically [34], they focus on digital solutions (e.g., zero-trust architectures) and overlook physical–environmental signals, such as the ENF, which are essential for grounding virtual actions in real-world authenticity.

2.4. Research Gaps and Contributions of This Survey

The previous literature treats the ENF predominantly as a forensic tool [7,29], with nascent extensions to smart grids [13], and Metaverse security as a separate domain, emphasizing virtual defenses [12,38]. No existing survey bridges these areas by positioning the ENF as a dynamic fingerprint for Metaverse security, addressing challenges like DT authentication, deepfake detection in grids, and quantitative trade-offs (e.g., 95–99% accuracy in our ANCHOR-Grid framework). Our contributions include the following: (1) a unified analysis of ENF reliability in deployment scenarios; (2) novel applications in Metaverse-integrated infrastructures; and (3) future directions for hybrid integrations, filling critical gaps in immersive and decentralized environments.

3. Background and Fundamentals

Integrating ENF signals into security frameworks represents a substantial advancement in securing virtual environments, such as the Metaverse, and critical infrastructures, including smart grids. The ENF, a dynamically variable signal from power grids, typically oscillates around the nominal regional frequencies of 50 Hz or 60 Hz, influenced by supply–demand variations between interconnected grids [5]. These oscillations form a unique signature that refutes real-time physical conditions [40]. This section explores advanced methods for sensing, recording, and applying the ENF, highlighting its transformative role in addressing Metaverse security needs.

3.1. Methods of Sensing and Recording ENF

ENF data acquisition methods can be categorized into two main types: direct and indirect sensing. Each uses different technologies to capture fluctuations in the frequency of the grid.

3.1.1. Direct Sensing

Direct sensing utilizes specialized equipment, such as frequency meters and spectrophotometric units (PMUs), which are directly connected to power grids [41]. PMUs, essential for wide area monitoring systems (WAMS), provide precise and synchronized frequency data across vast geographic areas, enabling real-time monitoring of grid stability [22]. These devices typically detect frequency fluctuations within tight ranges, ensuring the high precision critical to synchronizing the DTs of smart grids with actual physical states [42]. Recent advances in synchrophasor technology, including GPS-based timing, have significantly improved PMU accuracy, providing submillisecond synchronization [43]. Emerging research explores quantum sensing technologies that promise even greater sensitivity to detect frequency shifts. However, direct sensing requires significant infrastructure, presenting scalability challenges in decentralized virtual ecosystems, such as the Metaverse.

3.1.2. Indirect Sensing

Indirect sensing methods use devices not specifically designed for frequency measurement, instead capturing environmental artifacts. For example, audio recorders can detect a hum at 50/60 Hz, and cameras can detect variations in lighting [44,45]. Advanced signal processing techniques, including spectrogram analysis and wavelet transforms, enhance ENF extraction by isolating relevant signals from noisy backgrounds [46]. For example, subtle LED flicker patterns linked to the ENF can be analyzed algorithmically for the retrieval of timestamped data [47]. Such techniques have proven crucial in digital forensics, authenticating multimedia by comparing embedded ENF signatures to established reference databases [25]. These methods extend their utility to IoT and smart home systems within the Metaverse, though challenges include susceptibility to environmental noise and delays due to complex signal processing requirements. Techniques such as compressive sensing could further optimize indirect sensing methods by reducing computational demands on edge devices [48].

3.1.3. Performance Comparison of Sensing Methods

Direct sensing through PMUs achieves detection accuracies of 18–38%, with frequency deviations up to 0.012 Hz and correlation coefficients ranging from 0.68 to 0.72 [8,9]. In contrast, indirect sensing, such as audio-based methods, yields higher accuracies of 75–92%, with tighter deviations of 0.0045 Hz and correlations up to 0.88 [9]. Wavelet transforms improve indirect sensing accuracy by 8% under low signal-to-noise ratios (SNRs) of 10 dB, though direct sensing maintains subsecond latency compared to 1–2 s for indirect methods [3,7]. These trade-offs guide the selection of sensing methods for Metaverse applications. To elucidate the strengths and limitations of the ENF sensing methods,

Table 1. Comparison of ENF sensing methods.

Metric	Direct Sensing (PMUs)	Indirect Sensing (Audio)
Detection accuracy	18–38% [8,9]	75–92% [8,9]
Frequency deviation	0.012 Hz [9]	0.0045 Hz [9]
Correlation coefficient	0.68–0.72 [9]	0.88 [9]
Noise resilience (10 dB)	8% accuracy loss [9]	8% accuracy gain [9]
Latency	Subsecond [3]	1–2 s [7]

compares direct sensing (using PMUs) and indirect sensing (based on audio) in key metrics such as detection accuracy, frequency deviation, and latency. Indirect sensing achieves 75–92% accuracy, significantly outperforming direct sensing’s 18–38%, though it incurs higher latency.

3.2. ENF as a Unique Signature

ENF fluctuations form distinct signatures unique to specific power grid interconnections, driven by real-time generator responses to demand changes [3]. This distinctiveness complicates artificial replication, providing a reliable basis for authentication. ENF signatures are extensively utilized in forensic sciences to authenticate multimedia recordings, helping legal investigations by correlating embedded frequency patterns with historical grid data [49,50,51].

In real-time security applications, grid operators use the ENF to detect anomalies, such as equipment failures or cyber threats, by comparing live measurements with expected patterns [16]. Machine learning models improve these capabilities by predicting trends and identifying deviations that indicate grid stress. Integrating the ENF with blockchain technology further ensures transparency and accountability, crucial to managing virtual environments where trust is essential [52]. The stochastic nature of the ENF, analyzed through advanced signal processing techniques, supports synchronization between virtual and physical domains. Techniques such as cross-correlation facilitate the alignment of ENF signals from multiple sources, allowing geolocation capabilities within the Metaverse and further enhancing security protocols [47,53].

The practical applications of the ENF include multimedia forensics and real-time grid monitoring, which are increasingly relevant to the Metaverse. ENF analysis aids in multimedia authentication, helping to combat threats to digital manipulation, such as deepfakes [4]. Real-time monitoring leverages the ENF for predictive grid failure detection, employing AI-driven analytics to prevent disruptions proactively [54]. In virtual environments, the ENF secures DTs through frequency-based watermarking, ensuring that virtual assets accurately reflect physical realities [7]. Anomaly detection theory and time series analysis underpin these applications, utilizing supervised and unsupervised machine learning techniques for threat detection [15]. Real-time data processing frameworks facilitate immediate threat responses, potentially integrating the ENF with other signals in the physical layer to establish comprehensive multimodal security frameworks [16,55].

3.3. Security Requirements in the Metaverse

The Metaverse, a decentralized and immersive digital environment, presents unique security requirements that traditional cybersecurity methods do not fully address. Dynamic identity authentication is crucial because, unlike static environments, the Metaverse involves continuous, real-time interactions and frequent changes in identity among avatars and virtual objects [56]. The ENF tackles this challenge by enabling location-specific and time-sensitive authentication that links users and events to physical infrastructure, thereby improving identity verification. Data integrity and provenance are equally essential, particularly since DTs and virtual assets often mimic physical infrastructures in critical applications, such as smart grids and healthcare systems [38]. The ENF supports this need by functioning as a physical watermark, embedding a real-world signal that ensures the authenticity of virtual data. Furthermore, the immersive nature of these environments demands real-time threat detection and analytics to quickly address potential security breaches without disrupting user experiences [57]. The ENF’s real-time variability forms a foundation for continuous activity monitoring, flagging anomalies based on deviations from expected frequency patterns. Privacy preservation also becomes an important concern, and the ENF can be utilized with privacy-enhancing technologies to validate data without compromising user confidentiality [58]. The scalability and resilience are bolstered by the ENF’s compatibility with FL and distributed architectures, making it suitable for expansive virtual environments.

3.4. Threat Models Unique to Metaverse

Metaverse platforms are particularly vulnerable to several specialized threats. Deepfake attacks, which use sophisticated AI-generated manipulations, pose significant risks by creating realistic but fraudulent DTs and virtual objects, undermining user trust and system integrity [12]. FDI attacks further exacerbate this problem by maliciously altering data within virtual infrastructures, leading to compromised decision-making processes [59]. Replay and spoofing attacks exploit legitimate data or signals, such as the ENF, to validate fraudulent transactions, creating operational inconsistencies and temporal misalignments. Furthermore, distributed denial-of-service (DDoS) attacks threaten to overwhelm nodes or communication channels within the Metaverse, severely obstructing critical interactions [60]. Social engineering risks are significantly amplified in immersive and highly interactive environments, enabling attackers to exploit user trust and behavior patterns more effectively than on traditional platforms [61].

3.5. Insufficiencies of Traditional Authentication Methods

Traditional security methods, including static passwords, biometrics, and cryptographic keys, exhibit considerable limitations within dynamic Metaverse environments [14]. Static authentication credentials are vulnerable due to their potential to quickly be compromised or become obsolete in rapidly changing virtual contexts [62]. The ENF overcomes this by generating inherently dynamic authentication tokens based on real-time environmental data. Latency in detection and response presents another critical challenge, as conventional systems typically do not provide the near-instantaneous threat detection required for virtual interactions in real time [63]. In contrast, the ENF offers subsecond signal variability that can be continuously analyzed for anomaly detection. These traditional methods also exhibit significant weaknesses against sophisticated AI-driven techniques, such as deepfakes, which convincingly replicate authentic user interactions [64]. Physical fingerprinting from the ENF provides a layer of verification resistant to manipulation, ensuring that digital artifacts align with physical world events. Furthermore, traditional authentication approaches rarely incorporate contextual validation based on user behavior, real-time physical signals, or environmental factors. This significantly limits their effectiveness in verifying dynamic interactions, gaps that the ENF directly addresses through its physical grounding and region-specific signature.

3.6. Existing Authentication Approaches in Metaverse Security

Contemporary approaches have emerged to address these challenges and offer improved security capabilities tailored to Metaverse environments. Blockchain-based authentication methods leverage decentralized, immutable ledgers to verify transaction authenticity and maintain tamper-proof records reliably [65]. The ENF complements the blockchain as a secure and verifiable input signal, enhancing the credibility of entries recorded on the blockchain. FL facilitates decentralized anomaly detection across distributed nodes, effectively preserving privacy while improving detection accuracy without centralizing sensitive data [66]. ENF data, integrated into FL models, enhances local anomaly detection by introducing a unique environmental signal that facilitates the identification of contextually grounded threats. Multimodal authentication combines biometric verification, behavioral analytics, and context-aware authentication mechanisms to create robust systems capable of dynamically validating users [67]. In this context, the ENF is a physical-layer modality that enhances digital biometrics by incorporating real-world signal traces, thus increasing resistance to spoofing attacks. Zero-trust architectures improve security by treating every interaction as potentially malicious and continuously verifying identity and behavior to minimize trust assumptions, thus significantly reducing the potential attack surfaces [68]. The ENF’s nonreplicable nature aligns seamlessly with zero-trust principles, providing a trusted, real-time anchor that validates every transaction and interaction within the Metaverse.

3.7. Role of ENF in Enhancing Metaverse Security

The ENF offers a uniquely effective solution to the security challenges encountered in the Metaverse by connecting digital interactions to physical phenomena. The ENF can enhance Metaverse security through several essential functions. It provides a geographically specific dynamic authentication mechanism in real time, making it exceptionally difficult for attackers to replicate or spoof due to its inherent stochastic variability directly tied to physical grid conditions [69]. The authenticity of DTs and avatars is ensured by continuously verifying user identities and interactions based on these precise ENF patterns. Moreover, ENF-based watermarking techniques embed secure, timestamped signatures within digital assets, ensuring robust data integrity and enabling tamper-resistant verification [70]. This significantly reduces the threat of manipulated virtual assets, such as deepfakes. The ENF’s unique capability for proactive threat detection allows it to quickly identify anomalies or deviations from expected frequency patterns that may indicate malicious activities, facilitating immediate security responses. Furthermore, integrating ENF signals with advanced cryptographic methods such as differential privacy and homomorphic encryption safeguards sensitive user data, maintaining privacy without compromising analytical efficiency [17]. Finally, ENF-based frameworks, when paired with blockchain and FL technologies, create scalable and distributed validation systems suitable for complex and extensive Metaverse applications, significantly strengthening resilience and reliability in virtual environments.

Table 2. Synthetic summary of Section 3: ENF signal analysis in Metaverse security.

Subsection	Focus	Key Content Summary
Methods of Sensing and Recording ENF	Describes various techniques for capturing ENF signals from multimedia sources	ENF signals can be sensed via audio (e.g., microphones, smartphones) or video (e.g., CCTV, rolling shutter cameras). Precision depends on the sampling rate, hardware quality, and environmental conditions. There are two main approaches: intrusive (dedicated sensors) and non-intrusive (ambient recordings) [22,41,49].
ENF as a Unique Signature	Defines ENF as a temporal biometric feature	ENF varies randomly but consistently across a region, enabling its use as a spatio-temporal fingerprint. It enables content authentication, source verification, and timestamp validation [67,71].
Security Requirements in the Metaverse	Explores baseline security needs for immersive virtual environments	Security in the Metaverse requires identity verification, secure data transmission, environment integrity, user privacy, and non-repudiation. Traditional measures lack context-awareness needed for immersive and decentralized platforms [13,72].
Threat Models Unique to Metaverse	Outlines novel threat vectors specific to Metaverse platforms	Threats include avatar spoofing, virtual replay attacks, synchronization tampering, sensory injection, and immersive phishing. These threats require multimodal and persistent authentication mechanisms [56,73].
Insufficiencies of Traditional Authentication Methods	Examines the limitations of legacy security solutions	Passwords, biometrics, and token-based methods often fail in dynamic, immersive, or low-trust virtual environments due to latency, spoofability, and usability concerns [38,74].
Existing Authentication Approaches in Metaverse Security	Reviews current proposals to secure identity and actions in virtual worlds	Includes blockchain-based identity management, zero-knowledge proofs, and behavioral biometrics. These approaches still struggle with real-time enforcement, scalability, and integration across platforms [1,56].
Role of ENF in Enhancing Metaverse Security	Proposes ENF as an effective modality for context-aware and low-cost authentication	ENF provides a passive, environmental-based layer of authentication by correlating embedded frequency signals with grid references. It resists spoofing and supports continuous authentication without explicit user input [3,7].

summarizes the discussions in Section 3 for the convenience of our readers.

4. ENF Signal Analysis: Capabilities and Limitations

This section evaluates the performance of the ENF, incorporating updated empirical data to reveal limitations and propose mitigation strategies, aligning with the security mechanisms discussed in Section 5.1 and Section 5.3.

4.1. Reliability Across Deployment Scenarios

The reliability of the ENF application depends highly on its operational context. In centralized, infrastructure-rich environments such as power utility control rooms or large data centers, ENF signals captured through high-precision PMUs offer high temporal resolution but limited spatial variation due to strong intra-grid synchronization. These conditions create narrow frequency deviation windows (typically ±0.005 Hz); however, this homogeneity diminishes the utility of the ENF as a distinguishing fingerprint within a single grid domain [75,76].

In contrast, indirect sensing from smartphones and surveillance cameras captures the ENF through background audio or light modulation in heterogeneous edge environments, such as mobile or IoT-based Metaverse deployments. Although these sources introduce nondeterministic filtering and environmental noise, they simultaneously enhance spatial entropy and variability, supporting more robust authentication in diverse virtual contexts. Field studies show that ENF signatures derived from environmental microphones maintain a detection accuracy of 75–92% even under variable ambient conditions, outperforming direct sensing in unstructured domains by nearly 2× [77].

Additionally, cross-domain deployments, for example, virtual environments that span smart homes, critical infrastructure, and mobile devices, require harmonization across regions with different nominal frequencies (50 Hz vs. 60 Hz). Without normalization of the signal, the drift of synchronization can reach 200 ms, increasing the likelihood of false negatives in authentication algorithms by 10–15% [76].

4.2. Systematic Comparison of Sensing Modalities

A side-by-side technical analysis of ENF sensing modalities reveals critical performance differentials, as shown by

Table 3. Comparison of ENF sensing modalities.

Method	Accuracy	Dev.	Latency	Corr.	Cost	Use Case
PMU (direct)	18–38%	±0.012 Hz	<1 s	0.68–0.72	High	Critical ops
Audio (indirect)	75–92%	±0.0045 Hz	1–2 s	∼0.88	Low–med	Forensics, IoT
Video (indirect)	70–85%	±0.005 Hz	∼2 s	∼0.83	Medium	Surveillance

. Direct sensing methods, such as those using PMUs, provide deterministic precision and subsecond latency, making them ideal for grid state estimation. However, their limited scalability, high deployment costs, and sensitivity to localized grid harmonics hinder their use in lightweight or geographically distributed virtual environments [31,75].

Advanced wavelet-based feature extraction enhances the resilience of SNR in indirect sensing by 8%, especially under suboptimal conditions (e.g., environments with 10 dB). Spectrogram-based methods excel at capturing transient ENF bursts, which are beneficial for timestamping events but require greater computational resources [49,78].

4.3. Environmental and Operational Influences

The deployment of the Electric Network Frequency (ENF) as an environmental fingerprint for Metaverse security is significantly influenced by environmental and operational factors, which can impact its reliability and effectiveness across various contexts. These influences include noise interference, regional grid variations, device-specific constraints, and adversarial manipulations, each of which poses unique challenges to the ENF’s application in virtual ecosystems like smart grid digital twins (DTs) and immersive platforms.

• Acoustic environment: Reverberation and ambient noise can distort harmonics, reducing the correlation [78]. The acoustic environment significantly impacts ENF extraction from audio recordings, where background noise and interference can distort the embedded frequency signals around 50/60 Hz. Studies show that in noisy settings, such as urban areas with high ambient sound levels, ENF matching accuracy drops by up to 30% when signal-to-noise ratios (SNRs) fall below 10 dB, as the fluctuations become masked by speech or environmental hum [48]. Adaptive filtering techniques, like those using quadratic mean methods, can mitigate this by isolating ENF patterns, improving reliability in forensic applications to 85–90% even in challenging acoustic conditions [79].
• Sensor diversity: Variations in the precision of the ADC and the sampling rate between devices affect the reliability [49]. Sensor diversity introduces variability in ENF capture, as different devices, such as microphones, PMUs, or cameras, exhibit unique sensitivities to frequency signals. For instance, smartphone sensors often suffer from aliasing effects and lower resolution, resulting in mean absolute errors of 0.15 Hz in ENF estimation compared to high-precision PMUs, which achieve sub-0.05 Hz accuracy in controlled settings [31]. This diversity enhances robustness in multi-sensor fusion, but it requires calibration to handle discrepancies. Research indicates that combining direct and indirect sensors in grid monitoring can lead to an improvement of up to 20% in detection rates [80].
• Electrical load: Grid stress conditions increase frequency jitter, introducing timestamp drift [77]. Electrical load fluctuations directly influence ENF stability, causing deviations of ±0.5 Hz from nominal values due to supply–demand imbalances in power grids. In high-load scenarios, such as peak industrial usage, ENF variability increases by 15-20%, complicating timestamping in multimedia forensics and requiring longer reference recordings (e.g., over 10 min) for reliable matching [81]. Renewable energy integration exacerbates this, with studies showing up to 8% error in anomaly detection during load shifts, necessitating load-aware algorithms for accurate ENF-based security in smart grids [82].
• Lighting artifacts: Light source flicker interferes with video-based ENF extraction [31]. Lighting artifacts, such as LED flicker modulated by the ENF, provide indirect visual cues for frequency extraction in video recordings, with intensity variations capturing the ENF at harmonics up to 120 Hz. However, environmental factors like dim lighting or motion blur can reduce extraction accuracy to 70–80%, as artifacts become less discernible in low-illumination settings [83]. Advanced techniques, including super-pixel segmentation, have improved intra-grid location estimation from these artifacts, achieving 92% precision in identifying recording origins within the same power network [75].
• Infrastructure jitter: In edge clouds, latency affects the alignment of ENF frames [69]. Infrastructure jitter, arising from network delays and synchronization issues in distributed grids, affects ENF signal timing, introducing errors of 0.1–0.2 s in real-time applications. In Metaverse-linked systems, this jitter can degrade consensus-based validation by 10–15%, particularly in geo-diverse setups where latency exceeds 50 ms [84]. Mitigation through wavelet-based signature extraction has shown promise, reducing jitter-induced mismatches by 25% in power ENF localization across distribution networks [85].

4.4. Quantitative Performance Trade-Offs

A comprehensive understanding of ENF trade-offs is essential for optimizing security deployments:

• Latency vs. robustness: PMUs offer subsecond response but lower detection accuracy [75]. In ENF-based systems, latency versus robustness presents a key trade-off, where direct sensing via PMUs delivers subsecond response times but only 18–38% detection accuracy due to grid variability. At the same time, indirect audio extraction offers 79–92% robustness at the cost of 5–10 s delays from noise processing [48]. Forensic applications highlight that shorter recordings (under 5 min) exacerbate this, reducing matching reliability by 20–30% in low-SNR environments [86].
• Entropy vs. privacy: ENF entropy supports nonce/key generation but may risk location leakage [76]. ENF signals provide high entropy for dynamic key generation in security systems, with fluctuations offering up to 0.9 bits per sample for cryptographic purposes, but this raises privacy concerns as location data can be inferred from grid-specific patterns, potentially leaking user geolocation with 85% accuracy [87]. Balancing this, privacy-preserving techniques like differential privacy reduce entropy by 15–20% while mitigating leakage in biometric or Metaverse authentication scenarios [88].
• Authentication vs. energy cost: Continuous ENF monitoring requires optimization for low-power systems [49]. Authentication accuracy in ENF systems reaches 95–99% with ML-enhanced models, but this incurs high energy costs, consuming 1.7 MAC/cycle in edge devices compared to baseline filters at 92% accuracy with 50% less power [76]. In resource-constrained IoT forensics, this trade-off limits deployment, with studies showing a 20% energy overhead for real-time ENF validation in battery-powered sensors [3].
• Security precision vs. coverage: Threshold tuning (e.g., Pearson > 0.8) balances detection vs. false positives [89]. Security precision in ENF detection achieves 92% in controlled environments but drops to 70–80% when expanding coverage across large-scale grids due to regional variability and interference [31]. Multi-node setups improve coverage by 40% in FDI attack mitigation, yet precision trades off with increased false positives (up to 15%) in diverse infrastructures like Metaverse-linked smart grids [90].

Hybrid methods that combine the ENF with voltage harmonics or transient signals improve detection accuracy by 10–15% [49,78].

4.5. Toward Resilient and Scalable ENF Integration

To ensure ENF robustness in the Metaverse, the following strategies are recommended:

• Multisignal fusion: This combines the ENF with harmonics, transients, and voltage signatures [91]. Multisignal fusion enhances ENF resilience by combining it with other environmental signatures, such as GPS timestamps or acoustic data, achieving 95% anomaly detection accuracy in smart grids while reducing false positives by 25% through wavelet-based integration [7]. In Metaverse applications, this fusion supports DT authentication, with simulations showing a 40% improvement in robustness against deepfakes in virtual–physical hybrid systems.
• Edge-compatible AI models: These use quantized or pruned neural nets for embedded ENF analysis [49]. Edge-compatible AI models for ENF processing use lightweight neural networks like QNNs to enable subsecond responses on resource-limited devices, cutting computational overhead by 15–20% while maintaining 90% accuracy in grid monitoring [92]. These models facilitate scalable integration in Metaverse edge environments, supporting adaptive training that boosts VR simulation reliability by 10–15% [93].
• FL: This normalizes the ENF models across nodes without data centralization [76]. FL normalizes ENF models across distributed nodes in smart grids without centralizing sensitive data, achieving 92% convergence in anomaly detection while preserving privacy through local training [94]. This approach scales to Metaverse healthcare, reducing model divergence by 20% in multi-device setups and enabling secure aggregation with 40% lower communication overhead [95].
• Geo-adaptive calibration: This compensates for regional grid variability using trained correction models [76]. Geo-adaptive calibration compensates for regional grid variability (e.g., 50 Hz vs. 60 Hz) using trained ML correction models, improving ENF matching accuracy by 10–15% in cross-border Metaverse applications [7]. These models, based on probabilistic filtering, reduce errors from load imbalances by 8%, enabling seamless integration in global smart grids with 95% reliability in location estimation [75].
• Consensus-based validation: This uses distributed voting (e.g., PBFT) for anomaly confirmation across nodes [53]. Consensus-based validation leverages protocols like PBFT to verify ENF signals across nodes, tolerating up to 30% faulty participants while achieving over 95% agreement in smart grid DTs [7]. In Metaverse ecosystems, this scales to multi-user environments, cutting FDI attack success by 40% and enhancing trust with blockchain fusion for tamper-evident logging [96].

By addressing these technical and operational considerations, ENF signal analysis can evolve into a foundational security primitive for immersive and decentralized cyber–physical environments. Its integration with edge computing, AI, and blockchain frameworks provides a scalable path forward for next-generation Metaverse infrastructures.

5. ENF-Based Security for Metaverse Systems

Building on the capabilities and limitations outlined in Section 4, developing an ENF-focused security framework must consider both the variability of signal fidelity and the specific requirements of the deployment environment. While ENF signatures offer significant potential for device and media authentication, their effectiveness across different operational settings necessitates adaptive solutions. This section presents a multi-layered security architecture designed to systematically address these challenges through signal validation, multifactor decision models, and environmental calibration, thereby bridging theoretical knowledge with practical mitigation strategies. This framework lays the groundwork for evaluating real-world applicability in Section 6 and Section 7.

5.1. The CIA Triad: Confidentiality, Integrity, and Availability

The CIA triad—Confidentiality, Integrity, and Availability—is fundamental in information security, protecting data and systems against unauthorized access, manipulation, and disruptions [97]. The ENF presents a dynamic approach to reinforce these security aspects within the context of the Metaverse and smart grid DTs. The natural variability of the ENF, directly related to fluctuations in the physical grid, introduces an innovative layer of protection that complements conventional cryptographic methods [29]. This section examines how ENF enhances the CIA triad by incorporating cryptographic principles, advanced signal processing, and anomaly detection methodologies designed explicitly for virtual ecosystems.

5.1.1. Confidentiality with ENF-Based Encryption Key Generation

Confidentiality ensures that sensitive information remains secure from unauthorized access [98]. Traditional encryption uses static keys that become vulnerable over time due to interception or brute-force attacks [30]. ENF-based encryption offers a dynamic alternative, employing real-time fluctuations in grid frequencies as unique, temporal, and geographically specific seeds to generate cryptographic keys [3]. The strength of ENF security lies in its high level of unpredictability, resulting from the inherent stochastic nature of the dynamics of the power grid. Using ENF values captured in specific instances and applying cryptographic hashing techniques, such as SHA-384 or HMAC, produces transient keys that significantly reduce the risk of attacks, including replay or man-in-the-middle attacks [7]. This approach is in line with modern cryptographic standards, ensuring high entropy, which is essential for secure key generation [99]. Combining the ENF with elliptic curve cryptography (ECC), known for strong security with shorter keys, further optimizes encryption processes for resource-limited devices prevalent in Metaverse applications [100]. In IoT scenarios, this facilitates continuous key refreshing, enhancing security by minimizing exposure to potential threats. Future research incorporating quantum-resistant algorithms with ENF-based methods could fortify confidentiality against emerging cybersecurity threats [98]. ENF-based keys generated using SHA-384 achieve 128-bit entropy, reducing replay attack success rates by 95% in IoT scenarios [7]. This high entropy ensures robust confidentiality for resource-constrained devices in Metaverse applications.

5.1.2. Integrity with ENF-Based Watermarking

Integrity guarantees that data remain unmodified and accurate during storage or transmission [101]. ENF-based watermarking incorporates frequency signatures in digital content, such as audio, video, or virtual grid logs, as verifiable markers of authenticity [29]. During creation, specific ENF patterns are embedded in content using least significant bit (LSB) embedding or frequency domain modulation. Subsequent extraction and comparison to known ENF databases validate the authenticity, with discrepancies signaling potential tampering [7]. Using the ENF’s region-specific characteristics provides robust protection against content manipulation in the Metaverse [102].

This technique utilizes advanced signal processing approaches, such as discrete wavelet transforms (DWTs), to embed watermarks that are resilient to data compression and noise interference [103]. In Metaverse environments, watermarking secures critical digital assets, including virtual property records or blockchain transaction ledgers, by creating an immutable record verified through real-time grid data [104]. This approach effectively counters threats such as deepfakes, providing scientifically supported mechanisms for digital authenticity [105]. Ongoing research into perceptual hashing combined with the ENF could further scale these verification techniques, supporting massive datasets and effectively combating misinformation [103]. The watermarking of the ENF through DWT achieves 90% detection accuracy against tampering in noisy environments with an SNR as low as 12 dB [1]. This resilience makes it effective for securing digital assets in Metaverse environments.

5.1.3. Availability with ENF-Based Anomaly Detection

Availability ensures uninterrupted access to systems and data despite attacks or system failures [106]. ENF-based anomaly detection leverages grid frequency monitoring to identify disruptive threats, including DDoS attacks or unauthorized system intrusions [107]. By establishing standard ENF behavior patterns from historical data, machine learning models can detect deviations promptly, initiating proactive responses such as node isolation to maintain service continuity [108]. In virtual grids and Metaverse environments, such uninterrupted access is critical [109].

ENF anomaly detection employs machine learning techniques, including supervised algorithms such as Support Vector Machines (SVMs), which identify known threat patterns, and unsupervised models such as autoencoders, which discover novel anomalies [15]. Sudden changes in the ENF can indicate cyber–physical attacks, triggering immediate mitigation strategies [16]. Real-time analytics frameworks, such as Apache Flink, process ENF data continuously, enabling the rapid response necessary to maintain virtual control operations [55]. ENF analysis also predicts equipment failures in smart grids, improving preemptive maintenance for virtual DTs [54]. Future research into adaptive detection models employing reinforcement learning could further reduce false positives in complex Metaverse ecosystems [110]. ENF-based anomaly detection using SVMs reduces false positives by 15% compared to traditional intrusion detection systems (IDSs), enhancing availability in virtual grids [41].

Integrating the ENF with network security protocols, such as Transport-Layer Security (TLS), improves the authentication of the data stream, ensuring secure virtual interactions [100]. Utility companies deploying ENF-based IDSs can effectively protect virtual critical infrastructure from global cyber threats [107]. The physical nature of ENF signals significantly reduces spoofing risks, providing resilience against attacks without direct access to the network [109]. Combining the ENF with other physical-layer signals, such as voltage harmonics, could enhance multimodal detection frameworks, improving accuracy through sensor fusion techniques [16].

The integration of the ENF into the CIA framework offers a comprehensive security approach. Pairing the ENF with advanced cryptographic methods, such as zero-knowledge proofs, enhances both confidentiality and integrity simultaneously [101]. Statistical process control (SPC) methodologies further refine anomaly detection by modeling ENF fluctuations as stochastic processes, thus clearly distinguishing between attacks and natural variations [48]. As virtual and physical worlds converge within the Metaverse, the ENF supports cyber–physical system (CPS) security principles, effectively bridging tangible grid dynamics with virtual operations [97].

Real-world examples, such as the FNET/GridEye monitoring network, demonstrate the ability of the ENF to track grid stability effectively [40]. Forensic applications validated in IEEE standards provide precedents for ensuring the authenticity of digital content within the Metaverse [53]. These practical applications underscore the adaptability of the ENF, supported by interdisciplinary research across engineering, computing, and analytics, poised to ensure robust protection for evolving virtual ecosystems [111].

Table 4. ENF-based CIA triad applications.

CIA Aspect	ENF Application	Key Metric	Reference
Confidentiality	Dynamic key generation	128-bit entropy, 95% attack reduction	[7]
Integrity	Watermarking (DWT)	90% detection accuracy (12 dB SNR)	[1]
Availability	Anomaly detection (SVM)	15% false positive reduction	[41]

summarizes ENF applications in improving CIA within Metaverse security frameworks. For example, ENF-based watermarking achieves 90% detection accuracy in noisy environments, reinforcing data integrity against tampering.

5.2. Cyber–Physical Threats and Resilience Strategies in Metaverse-Integrated Smart Grids

Integrating smart grids with the Metaverse significantly advances energy infrastructure management through the use of DTs, immersive virtual environments, and sophisticated data analytics. However, this integration exacerbates existing vulnerabilities and introduces new cyber–physical threats. This section examines the primary threats to smart grids integrated with the Metaverse, their underlying technical mechanisms, and advanced strategies to build resilience, drawing on recent research and technological advancements.

5.2.1. Data Manipulation Threats: Exploiting Virtual and Physical Interfaces

Data manipulation threats compromise the integrity of information flows within smart grids, exploiting enhanced connectivity and virtual representations facilitated by the Metaverse. Key threats include deepfake attacks targeting DTs, FDI attacks, and spoofing ENF signals, each uniquely undermining grid reliability.

DTs are detailed, virtual replicas of physical grid components, enabling real-time monitoring and operational simulations. Their extensive use of sensor and audiovisual data makes them susceptible to deepfake attacks, where attackers use generative AI techniques to produce realistic but false data representations [112]. For example, attackers could manipulate virtual readings to show normal operating conditions despite imminent failures of physical components, delay response actions, and risk severe damage [113]. Detecting such manipulations in dynamic data streams is a challenging task. Multimodal data fusion approaches—combining thermal imaging, ENF signals, and acoustic data—offer effective detection methods by identifying inconsistencies between data modalities [114]. Furthermore, blockchain technology can securely log digital twin interactions, creating tamper-proof records crucial for forensic analysis [115].

FDI attacks involve inserting malicious data into grid control systems to disrupt operational decisions, such as frequency regulation or load balancing. In the Metaverse, attackers can simulate and refine FDI strategies using virtual grid environments, significantly enhancing their threat capabilities [110]. Effective countermeasures involve real-time validation of distributed sensor ENF data, detecting discrepancies between virtual representations and actual physical states [116]. Machine learning models trained on historical grid data, such as deep neural networks (DNNs) and SVMs, further improve the detection of subtle data deviations indicative of FDI attacks [117].

ENF signals, naturally derived from frequency fluctuations in the grid, provide essential physical-layer authentication for data integrity. Attackers spoof these signals by creating artificial frequency patterns that mimic legitimate behaviors, undermine monitoring systems, and conceal anomalies [118]. Advanced techniques, such as wavelet transforms, are employed to dissect ENF signals, uncovering irregularities that are hidden in simpler analyses [119]. Combining the ENF with additional physical-layer attributes, such as voltage harmonics, provides a robust multifactor authentication approach, significantly complicating spoofing efforts [120].

5.2.2. Disruption Threats: Compromising Grid Availability

Disruption threats focus on impairing essential grid functionalities by exploiting real-time communication and data dependencies within the Metaverse. Key disruption threats include replay attacks and denial-of-service (DoS) attacks, each presenting distinct operational risks.

Replay attacks involve capturing legitimate ENF data and retransmitting it to bypass security protocols or disrupt timely operations. Attackers exploiting cached virtual grid data can introduce outdated or malicious information, thereby destabilizing operations [121]. Effective defenses utilize ENF-based dynamic tokens, leveraging real-time frequency data to ensure the freshness and authenticity of communications. Blockchain technologies support these defenses by securely timestamping ENF data, preventing its unauthorized reuse [122].

DoS attacks overwhelm communication networks, obstructing the critical data flow required for grid operations. Attackers target virtual control interfaces in the Metaverse with excessive traffic, limiting operator responses to real-time events [39]. Defense strategies incorporate multi-layered security measures, such as AI-driven anomaly detection to identify abnormal traffic patterns and decentralized network architectures to reduce vulnerability points [123,124]. These strategies collectively maintain operational integrity during disruptive conditions.

5.2.3. Advanced Resilience Strategies: Utilizing the Metaverse for Enhanced Defense

Despite increasing threats, the Metaverse also provides opportunities to strengthen the resilience of the grid. Integrating virtual simulations, blockchain, AI-driven analytics, and decentralized network structures enables smart grids to counter vulnerabilities and adapt to emerging threats proactively.

Immersive Metaverse environments facilitate detailed virtual simulations of grid operations, providing safe platforms for threat modeling and resilience testing. Realistic simulations can replicate scenarios such as deepfake or DoS attacks, evaluating the resilience of the grid without physical risks [125]. Game theory and adversarial machine learning optimize the allocation of cybersecurity resources, ensuring preparedness against real-world threats [126,127].

Blockchain establishes a decentralized and immutable record system that ensures data integrity and secure interactions in smart grids. It supports secure identity management and authenticates virtual devices in the Metaverse, preventing identity spoofing and unauthorized access [128,129]. This decentralization aligns perfectly with the resilience required for grid operations, eliminating single points of failure.

Artificial intelligence (AI) models, such as recurrent neural networks (RNNs), effectively analyze ENF signals and other sensor data to identify operational anomalies and potential cyberattacks [130,131]. These AI systems integrated with DTs deliver real-time threat alerts, enabling rapid responses. Predictive maintenance strategies further utilize AI to anticipate equipment failures through subtle changes in ENF signals, thus enhancing grid reliability [132].

Self-healing networks autonomously detect and mitigate threats, significantly enhancing grid resilience [133]. In the Metaverse, these systems dynamically reroute data and isolate compromised nodes, maintaining operational continuity [134]. Decentralized approaches, including edge computing and FL, distribute computational loads and security functions, reducing vulnerabilities, and enhancing scalability [135].

5.2.4. Emerging Threats and Future Considerations

New risks emerging from Metaverse integration, such as quantum computing threats to cryptographic protections and social engineering within immersive interfaces, require ongoing research into quantum-resistant cryptography and human-focused security protocols [72,136]. Hybrid cyber–physical attack models, which integrate digital and physical threats, necessitate comprehensive security frameworks to anticipate and counteract complex attack vectors [137]. Using advanced technologies and deep scientific insights, future strategies can robustly protect Metaverse-integrated smart grids from sophisticated threats, ensuring sustainable resilience. Quantum computing poses a significant threat, potentially reducing ECC key strength by 50% [6]. The stochastic nature of the ENF can enhance the generation of quantum-resistant keys, increasing entropy by 10% [122]. Additionally, social engineering attacks in immersive Metaverse interfaces have a 30% higher success rate due to their engaging nature, necessitating human-focused security protocols [48].

The deployment of the ENF in the Metaverse presents ethical challenges, particularly with respect to privacy and the potential misuse of geolocation data. ENF signals, due to their ability to provide geolocation-specific signatures, can inadvertently reveal user locations within virtual environments, raising significant privacy concerns [3]. For instance, in multimedia forensics, the ENF has been used to track the geographical origin of recordings, a capability that could be exploited in the Metaverse to monitor user movements without consent [1]. Such tracking could lead to ethical dilemmas, especially in applications involving sensitive data, such as virtual healthcare or financial transactions, where user anonymity is paramount. To mitigate these risks, privacy preservation techniques, such as differential privacy, can be applied to anonymize ENF data, reducing privacy leakage by an estimated $15\%$ while maintaining its utility for authentication [122]. Furthermore, ethical guidelines must be established to govern the use of the ENF in the Metaverse, ensuring that its deployment aligns with the principles of user consent and data protection. Future research should focus on developing frameworks that balance the security benefits of the ENF with ethical considerations, potentially drawing from forensic standards to ensure responsible use [1].

Integrating the ENF into Metaverse security frameworks highlights significant regulatory gaps that complicate compliance with international data protection laws, such as the GDPR [122]. Currently, there are no standardized regulations governing the use of ENF data in virtual environments, which poses challenges to ensure consistent privacy and security practices in jurisdictions. For example, in multimedia forensics, the lack of standardized ENF logging protocols has led to legal disputes over the admissibility of evidence. This challenge could extend to Metaverse applications where the ENF is used for authentication [138]. This regulatory ambiguity may lead to compliance failures, particularly for global Metaverse platforms that operate across different legal frameworks. International regulatory bodies, such as the IEEE, could develop ENF-specific standards to address this, building on existing forensic protocols to ensure uniformity and legal compliance [1]. Such standardization efforts could streamline compliance processes, potentially reducing legal risks by 20–25% by providing clear guidelines for the handling of ENF data. Future research should prioritize collaboration with policymakers to establish these standards, ensuring that ENF deployment in the Metaverse is secure and legally sound.

Implementing ENF-based security mechanisms across multiple jurisdictions presents technical and operational challenges due to variations in power grid standards and regulatory environments. For instance, the nominal ENF frequency differs between regions, with $50\mathrm{Hz}$ in Europe and $60\mathrm{Hz}$ in the US, affecting the interoperability of ENF-based systems in a globally interconnected Metaverse [139]. These variations can lead to synchronization issues, as a Metaverse platform spanning multiple regions may struggle to align ENF signatures from different grids, potentially introducing errors in authentication processes. Advanced signal processing techniques, such as wavelet transforms, can mitigate these issues by normalizing ENF signals across regions, but they require additional computational resources [140]. FL offers a promising solution by enabling distributed ENF analysis, allowing regional nodes to collaboratively learn and normalize ENF patterns without centralized data sharing, potentially reducing synchronization errors by 10–15% [141]. Moreover, cross-jurisdictional deployment must navigate different privacy laws, as regions such as the EU impose stricter data protection requirements than others, complicating ENF data logging and sharing [142]. Future research should focus on developing adaptive ENF frameworks that account for regional grid differences and legal requirements, ensuring seamless integration in a global Metaverse context.

5.3. ENF-Driven Security Mechanisms for Metaverse Smart Grids

Integrating smart grids with the Metaverse creates a cyber–physical ecosystem where DTs replicate physical power infrastructure. This integration requires advanced security mechanisms that leverage the ENF to ensure data integrity, system reliability, and user authentication.

5.3.1. Cyberattack Detection and Mitigation

The dynamic nature of the ENF allows for effective detection of cyberattacks within Metaverse-based smart grids. Operators can compare frequency data from virtual sensors with real-world baselines to spot irregularities, including FDI attacks designed to destabilize grid functions [143]. Virtual dashboards can display ENF anomalies in immersive environments, enabling timely responses [144]. Advanced IDSs incorporate the ENF as a real-time parameter to simulate attack conditions, such as load spoofing, in virtual environments to improve operator readiness [35]. ENF signals typically reflect a feedback loop controlled by generator systems and are disrupted by cyber intrusions. Techniques such as long short-term memory (LSTM) networks analyze ENF time series data to detect anomalies outside normal ranges [145]. Other enhancements, such as Kalman filters, estimate real-time grid conditions by combining ENF with voltage data, improving detection accuracy [146]. Regulatory systems, such as NERC, already utilize frequency data to prevent blackouts, a strategy that can be extended to virtual grids [147]. Future efforts may include applying stability analysis methods to better understand the impact of attacks and developing more robust countermeasures. LSTM models for ENF-based cyberattack detection achieve 92% accuracy in identifying FDI attacks, with a mean absolute error of 0.015 Hz in time series analysis [148].

5.3.2. Data Provenance and Integrity Assurance

In Metaverse-integrated smart grids, ensuring data provenance is critical. The ENF provides a timestamped identifier that verifies the authenticity of transmitted data, aligning virtual actions with physical events [149]. For example, a power report from a virtual substation can include embedded ENF data validated against live grid signals to prevent tampering [150]. This method is grounded in statistical signal processing, using the natural randomness of the ENF as a unique identifier [151]. Techniques such as spread spectrum watermarking embed ENF patterns into data packets while preserving their quality, allowing integrity to be verified using cross-correlation methods [152]. In Metaverse environments, ENF-enhanced blockchain records provide immutable logs of activity [73]. From audio forensic methods, signal-to-noise optimization improves watermark retrieval, even in noisy or altered content [153]. Supervisory control and data acquisition (SCADA) systems already use similar time synchronization techniques, offering a foundation to expand this approach to virtual environments [154]. Further resilience may be achieved by integrating error correction algorithms to restore partially altered ENF data. ENF-based blockchain logging reduces validation latency, ensuring efficient data provenance in Metaverse-integrated smart grids [155].

5.3.3. Device Authentication and Tamper Resistance

Authentication of devices in Metaverse smart grids can be achieved through ENF-based fingerprinting, as each physical sensor emits a unique frequency signature [7]. By comparing a device’s ENF output to expected values across the grid, anomalies suggesting tampering or unauthorized activity can be detected [4]. In virtual control systems, discrepancies in ENF readings are visualized alongside device performance data, enabling the rapid identification and isolation of compromised components [156]. This technique relies on principles from graph theory, where devices and their interconnections exhibit signal synchrony due to shared infrastructure [157]. Tampering disrupts these connections, which can be identified using spectral analysis. Bayesian methods can refine the likelihood estimates of tampering by comparing statistical differences in ENF signals [158]. These approaches align with broader IoT security practices, such as biometric access control [159], and can be extended through distributed validation protocols inspired by swarm robotics [160]. Predictive frameworks, such as graph neural networks, could help identify emerging tampering patterns. Bayesian methods improve the authentication of ENF-based devices, improving the accuracy of detection of tampering by 12% through statistical analysis of frequency deviations [74].

5.3.4. Forensic Analysis and Attack Attribution

The ENF provides a powerful tool for forensic analysis and attack attribution in virtual smart grid environments. Investigators can simulate attacks within virtual replicas to reconstruct incident timelines using ENF signatures, which vary by geographical region [116,161]. This specificity helps distinguish between virtual and physical disruptions. Digital forensics frameworks utilize the ENF’s millisecond resolution to sequence events accurately [162]. Integrating the ENF with location metadata enhances the tracking, much like epidemiological models that track the spread of infections [47]. Law enforcement already uses the ENF for audio authentication, and similar practices can be applied to Metaverse-based investigations [138]. The ENF also strengthens deep packet inspection, enabling the detection of combined cyber–physical threats [163]. Tools such as DeFakePro illustrate how ENF-based verification can counterdeep fake content, with applications in both media and critical infrastructure security [164]. Techniques such as Markov modeling could help predict attack sequences and improve attribution accuracy.

5.3.5. Real-World Anchoring for Event Validation

ENF signals serve as anchors in the real world to validate virtual grid events. When a virtual control center reports an incident, an embedded ENF signature is compared with live data to ensure authenticity [7]. If inconsistencies are found, the data may be flagged as potentially tampered with. This strategy effectively counters off-grid data injection attacks [165]. Using the ENF as a natural timestamp aligns with real-time control systems, which often employ synchronization techniques such as phase-locked loops [147]. Comparison of virtual and physical signals through spectral analysis helps detect anomalies [166]. Power utilities routinely log the ENF to verify incidents, a practice that can be scaled to virtual spaces [167]. Noise filtering methods, such as least mean squares algorithms, can enhance signal clarity for accurate validation [168]. Organizations like Bonneville Power Administration already use frequency data for event auditing, suggesting its applicability to future Metaverse applications [169]. Additional research may explore the synchronization of distributed ledgers with ENF timestamps to enhance trust and transparency.

5.4. Secure Communication with ENF-Based Nonces

The ENF enhances secure communication in smart grids by enabling the generation of time-sensitive nonces to thwart replay attacks. Embedded in communication packets, these nonces use the continuous fluctuations of the ENF to ensure data freshness [148]. Mismatched or outdated nonces are invalidated, preventing packet reuse and enhancing IoT security. This concept is grounded in cryptographic theory, where the natural entropy of the ENF makes each nonce difficult to predict [170]. Incorporating the ENF into established protocols, such as TLS, strengthens session validation [171]. Smart cities already use time-based tokens for secure device communications, offering a blueprint for Metaverse grid systems [172]. From chaos theory, unpredictable patterns of ENFs can also generate complex and secure nonces [173]. Standards such as IEEE 802.15.4 utilize similar mechanisms in IoT environments, adaptable to virtual grids [174]. Future advancements may involve blending the ENF with hardware identifiers to create hybrid nonces for multifactor authentication, enhancing robustness and temporal validation.

Pearson Correlation for ENF Authentication

Pearson’s correlation serves as a robust metric for authenticating ENF signals by quantifying the similarity between client and server ENF streams in Metaverse smart grid environments. This method leverages the stochastic nature of ENF fluctuations to distinguish genuine signals from manipulated ones, such as those generated by deepfake or replay attacks [7]. Empirically, a Pearson’s correlation cutoff of 0.8 was selected as the threshold for validating ENF pairs based on observations that genuine pairs consistently exceed 0.9 under conditions with up to 5% noise and 200 ms latency, while manipulated segments (e.g., replayed or deepfake ENF) never exceed 0.7 [7]. This threshold, positioned at the elbow of the correlation coefficient curve, achieves over 97% attack detection accuracy with false positives below 1.5% across deepfake and replay scenarios. A higher threshold risks rejecting legitimate readings affected by minor jitter, while a lower threshold may admit falsified signals; therefore, 0.8 is a practical balance for real-world applications.

Figure 4 illustrates this approach, showing a comparison of reference and sensor ENF signals over 1200 s (top) and their corresponding Pearson’s correlation coefficients across 250 windows (bottom). The red dashed lines in the top graph at 500 and 800 s highlight periods of significant deviation, while the bottom graph correlation dips below the 0.8 threshold, indicating potential manipulation. This visual analysis underscores the effectiveness of Pearson’s correlation in identifying anomalies, improving the security of virtual grid operations by ensuring the authenticity of the ENF data streams [7]. Future research could explore adaptive thresholds that dynamically adjust according to network conditions, further improving detection reliability.

5.5. Mitigation Strategies for ENF Security in the Metaverse

As smart grids transition into the Metaverse, ensuring the security of ENF data becomes increasingly critical. The ENF is central to grid synchronization, anomaly detection, and system stability. However, extending these functions into virtual environments brings unique challenges that require innovative mitigation strategies.

5.5.1. Multipoint Measurement and Distributed Consensus for ENF Validation

Using multipoint measurements significantly improves ENF data security within Metaverse-based smart grids. By deploying a network of virtual nodes that emulate real-world PMUs and IoT devices, frequency data is collected from diverse locations, capturing fluctuations around 60 Hz typical of US grids [175]. These spatially distributed measurements enable cross-verification to identify anomalies, such as spoofed signals, with distributed consensus mechanisms ensuring consistent validation across nodes. Scientific research confirms that ENF signals propagate synchronously throughout interconnected power systems, exhibiting strong correlations (Pearson coefficient) [48]. This synchronicity enables the reliable comparison of ENF patterns from different regions, facilitating the detection of inconsistencies that indicate tampering [41]. Studies show that even minor deviations (Hz) can be detected within subsecond intervals [176].

Anomalies are highlighted when the ENF reading of a node deviates from the group’s average by more than a predefined threshold, typically twice the standard deviation [177]. For example, if five nodes report 60.05 Hz, while the system average is 59.99 Hz, the deviation is significant enough to warrant investigation. Consensus protocols, such as Practical Byzantine Fault Tolerance (PBFT), verify this information despite the presence of potential fault nodes [19]. With 10 nodes, PBFT can tolerate 3 faulty participants while maintaining a high probability (over 99.9%) of correct consensus [14]. Blockchain technology improves this system by recording each frequency sample in an immutable ledger, allowing verifiable and tamper-resistant historical records [7,19]. With nodes sampling at 100 Hz, alerts can be issued almost immediately when consensus detects discrepancies, significantly minimizing the risk of false readings corrupting grid operations. PBFT consensus maintains 99.9% reliability with 10 nodes, scaling to 1000 nodes with only a 10% increase in latency, ensuring robust ENF validation in large-scale Metaverse environments [122].

5.5.2. Redundant Sensing and Fault-Tolerant ENF Monitoring

Redundant sensing improves reliability by integrating virtual and physical sensors into the ENF monitoring process. DTs of PMUs and environmental sensors, including those that capture the ENF from ambient 60 Hz hums, provide overlapping data streams that increase fault tolerance [178]. The accuracy in indirect detection remains high when SNRs exceed 20 dB [3]. By cross-referencing these data with high-precision PMU readings, discrepancies can be quickly identified and addressed. For example, if two sensors differ by more than 0.02 Hz, a threshold is triggered beyond the typical PMU resolution [41]. Historical ENF data also supports predictive maintenance. LSTM models trained on past frequency patterns anticipate sensor degradation with mean absolute errors as low as 0.015 Hz over 24 h windows [149]. In high-stress scenarios, such as a 50% node failure during a simulated DoS attack, the system’s reliability remains intact with over 99.9% operational coverage when three or more redundant sensors are in place [41]. Platforms like Apache Spark enable real-time analytics with sub-50 millisecond latency [44], validating sensor data and minimizing disruptions.

5.5.3. Threshold Voting for Anomaly Detection in Virtual Grids

Threshold voting provides a robust mechanism for anomaly detection by aggregating ENF readings from multiple nodes and applying weighted decision rules based on node reliability. Under normal conditions, the ENF values cluster within 59.97 to 60.03 Hz, forming a Gaussian distribution [179]. The reading of a node that deviates beyond a confidence interval (typically Hz) from the weighted average triggers an alert. The weights are assigned based on factors such as past accuracy and geographic stability [48]. For example, suppose that a high-reliability node reports a reading that diverges significantly from the group consensus. In that case, an equally divergent report from a lower-reliability node is scrutinized. Random Forest classifiers enhance this process by assigning confidence scores, identifying low-score readings as potential anomalies [29]. Dynamic thresholds that adjust based on contextual factors, such as load fluctuations, improve accuracy. For instance, allowable variation increases slightly in high-demand periods to avoid false positives. This approach has demonstrated a 15% reduction in false alarms while maintaining over 92% detection accuracy for spoofed signals [54,180].

5.5.4. Dynamic Node Selection for Enhanced Security

Dynamic node selection adds unpredictability to ENF validation by rotating the subset of nodes involved in monitoring. By randomly selecting, for example, 5 out of 20 nodes every 10 min using cryptographic randomization, attackers are less likely to target the correct subset [100]. With more than 15,000 potential combinations, randomness makes attack coordination difficult. Studies have shown that even with high correlation among ENF signals across the grid, rotating subsets maintain accurate representation [48]. Monte Carlo simulations validate this strategy, showing a 40% decrease in successful FDI attacks [109]. Reinforcement learning can optimize node selection by prioritizing those with a history of anomalies [181]. Cryptographically secure random selection using SHA-256 ensures tamper resistance [7], while geographically diverse nodes expand the detection coverage. Dynamic node selection reduces the success rates of FDI attacks by 40% in networks with 20 or more nodes, using cryptographic randomization to improve security [41].

5.5.5. Geographic Diversity for Robust ENF Analysis

Distributing ENF sensors across different geographical regions enhances resilience by capturing localized anomalies. The US grid, divided into western and eastern interconnections, displays region-specific ENF variance due to differences in load types and renewable integration [43]. By comparing regional ENF averages, operators can detect outliers that single-location monitoring might miss. For example, a deviation of 0.06 Hz from the grid average indicates potential tampering, particularly when regional fluctuations typically stay within 0.01 to 0.03 Hz [41,149]. Integrating weather data, such as wind speed, refines accuracy by compensating for natural fluctuations [54]. Geographic diversity also improves system resilience. If one region experiences node failure, others can continue monitoring with minimal disruption. Visualization tools render this data in real time, helping operators quickly identify and respond to regional threats. When implemented together, these strategies form a comprehensive framework for mitigating ENF data security within the Metaverse. By combining statistical validation, cryptographic protocols, AI, and geographic diversity, smart grids can achieve robust and scalable protection against emerging cyber–physical threats.

The proposed ENF-based framework addresses the challenges of signal integrity and threat classification highlighted in earlier sections, establishing a precedent for the deployment of these mechanisms in realistic and diverse environments. This transition from conceptual architecture to practical implementation is examined in the following sections. Section 6 presents a case study that exemplifies the application of the framework in a real-time forensic scenario. Meanwhile, Section 7 extends the findings to broader application domains, including surveillance, smart grid authentication, and immersive Metaverse platforms. These empirical discussions reinforce the adaptability of the framework and demonstrate how ENF-based security can be integrated into complex digital ecosystems.

Table 5. Synthetic summary of Section 5: ENF-based security for Metaverse systems.

Subsection	Focus	Key Content Summary
The CIA Triad	Fundamentals of cybersecurity: Confidentiality, Integrity, Availability	ENF-based security frameworks support the CIA triad by leveraging frequency-based environmental signatures. Confidentiality is preserved through randomized ENF encoding, integrity through verifiable ENF traces, and availability via distributed ENF-based verification mechanisms [98,101,106].
Cyber–Physical Threats and Resilience	Risk landscape at the intersection of Metaverse and smart grids	Highlights the vulnerabilities of cyber–physical infrastructures, including synchronization attacks, identity spoofing, and grid instability. Proposes resilience strategies such as dynamic ENF verification layers and real-time anomaly detection to maintain operational security [146,154,182].
ENF-Driven Security Mechanisms	Integration of ENF-based authentication in smart grid-enabled Metaverse environments	ENF supports passive authentication by embedding natural frequency variations into data streams. Used for verifying content authenticity, timestamp integrity, and spatial origin in distributed immersive systems [7,8].
Secure Communication with ENF-Based Nonces	Use of ENF variations to generate secure session keys and nonces	Temporal unpredictability of ENF provides a low-cost entropy source for generating dynamic nonces and session keys, reducing susceptibility to replay and man-in-the-middle attacks [98,100].
Mitigation Strategies	Defense techniques against known and potential ENF threats	Strategies include multi-source ENF correlation, threshold-based anomaly filters, and hybrid fusion with other biometric/environmental signals. Aims to enhance robustness in decentralized and latency-sensitive virtual systems [48,124,159].

summarizes the discussions of Section 5.

6. Case Study: ANCHOR-Grid—A Real-World Application of ENF in Smart Grid Security

Building upon the threat models, security mechanisms, and mitigation strategies outlined in the previous sections, we present a concrete implementation to demonstrate the practical viability and effectiveness of ENF-driven security frameworks in a real-world cyber–physical context. The following case study, ANCHOR-Grid, demonstrates how ENF-based authentication can address sophisticated cyber threats within Metaverse-integrated smart grid environments, such as deepfake attacks. This case study bridges the theoretical framework and its operationalization in complex, distributed infrastructures.

The ANCHOR-Grid framework exemplifies the practical deployment of ENF signals to secure the CPS, particularly smart grid DTs within the Internet of Smart Grid Things (IoSGT) ecosystem [7]. By leveraging the ENF’s unique temporal and spatial signatures, as discussed in Section 3 and Section 5.1, ANCHOR-Grid counters sophisticated deepfake attacks, ensuring the CIA triad of virtual grid representations in Metaverse-integrated environments. This case study expands on the implementation of the framework, incorporating recent empirical data to highlight its scalability, resilience, and potential for broader applications in cyber–physical systems.

6.1. ANCHOR-Grid Framework Overview

ANCHOR-Grid addresses the vulnerability of smart grid DTs to deepfake attacks, where adversaries inject fraudulent virtual models to mislead operators, risking operational disruptions or cascading failures [7]. As described in Section 5.2, such threats exploit the virtual–physical nexus, necessitating robust authentication mechanisms. ANCHOR-Grid embeds ENF signals—typically fluctuating within $\pm 0.02$ Hz in the US grid (Section 3.1)—into DT data streams, creating a grid-specific environmental fingerprint. Recent tests demonstrate that the synchronicity of the ENF throughout the network, with a Pearson correlation coefficient greater than 0.96, ensures reliable DT authentication across distributed networks [3].

The architecture of the framework integrates the components of the physical grid (for example, substations, renewable energy sources) with their DTs through a cloud layer and a blockchain-based validation system [7]. ENF signals are captured in real time using PMUs and embedded into Distribution Test (DT) communications. The blockchain ledger records ENF signatures, allowing decentralized cross-validation among DTs, preventing tampering and ensuring data provenance (Section 5.3.2). Recent enhancements, including optimized blockchain consensus algorithms, have reduced validation latency by 15%, improving scalability for large-scale grids with more than 1000 nodes, which aligns with the distributed security frameworks presented in Section 8.4.

6.2. ENF-Based Signature for Authentication

The stochastic fluctuations of the ENF, driven by supply–demand dynamics, form a temporal fingerprint specific to the grid that is difficult to replicate artificially (Section 3.2). In ANCHOR-Grid, ENF signals are captured from physical grid components using PMUs, processed via short-time Fourier transform (STFT) to isolate frequency deviations with subsecond resolution, and compared to a reference ENF database. Recent optimizations in STFT algorithms have improved extraction accuracy by 10%, achieving a Pearson correlation coefficient of 0.96 for authentic DTs [7]. A correlation threshold of 0.95 validates the authenticity; otherwise, the DT is flagged as fraudulent. Valid signatures are recorded on a blockchain, ensuring tamper resistance (Section 5.3.3).

To evaluate the practical effectiveness of ANCHOR-Grid, experiments were conducted under various network conditions, assessing its performance in authenticating DTs against deepfake and replay attacks [7]. Under low-latency conditions (<5 ms), ANCHOR-Grid achieved 99.9% precision and recall, demonstrating near-perfect detection of fraudulent DTs. Even under high latency (200 ms), the framework maintained 95.4% precision and 95% recall, showcasing its robustness to network delays. In scenarios with 5% packet loss, precision and recall were 96.7% and 90%, respectively, while high-jitter conditions resulted in 95.6% precision and 88% recall. These results highlight ANCHOR-Grid’s ability to reliably authenticate DTs in real-world Metaverse-integrated smart grids, even under challenging network conditions. The authentication process, depicted in Figure 5, plays a critical role in achieving this performance by generating ENF signatures from received data, comparing them with the server-side ENF based on client timestamps, and applying a correlation threshold of 0.8 to validate authenticity. This process ensures that only DTs with valid real-time ENF signatures are accepted, effectively mitigating the risk of deepfake attacks and ensuring operational integrity.

The authentication process, illustrated in Figure 5, is lightweight, with recent updates that reduce computational overhead by 20% through vectorized signal processing, making it suitable for resource-constrained IoT devices. The temporal variability of the ENF prevents replay attacks, as historical data misaligns with fluctuations in real time, achieving a 99% detection rate for unauthorized DTs [7]. Integration with edge computing has further reduced processing latency to under 50 ms, enhancing the real-time applicability in dynamic Metaverse environments.

6.3. Countering Deepfake Attacks

ANCHOR-Grid was evaluated in a Microverse-based virtual smart grid environment, simulating components such as substations and solar panels using Unreal Engine 5 [7]. Four attack scenarios were tested: injecting fake DTs, replaying historical ENF data, generating AI-driven synthetic DTs, and introducing noise to disrupt ENF signatures. In the first scenario, fraudulent DTs lacked real-time ENF signatures, failing authentication with frequency deviations outside the $\pm 0.02$ Hz range. The second scenario detected replay attacks via time-sensitive correlation analysis, achieving 99.8% accuracy (Section 5.4). In the third, deepfake DTs did not replicate the stochastic patterns of the ENF, and the detection accuracy reached 99.9% under low-latency conditions (<4 ms) [7]. The fourth scenario tested the resilience to noise, maintaining 90% precision in SNR as low as 12 dB, a 5% improvement over prior results.

These results highlight the efficacy of the ENF as a tamper-resistant authenticator, complementing the anomaly detection strategies in Section 5.3.1. ANCHOR-Grid’s optimized signal processing reduces computational overhead by 25%, enabling deployment on resource-constrained IoT devices (Section 8.3). Its robustness against adaptive deepfake threats surpasses traditional cryptographic methods, which falter under sophisticated AI-driven attacks (Section 5.1.1). Compared to conventional methods, ANCHOR-Grid achieves 99.9% detection accuracy, surpassing ECC-based methods at 95% and TLS-based methods at 92% [7]. However, it requires 20% higher computational resources. The ENF’s noise resilience (90% accuracy at 12 dB SNR) outperforms TLS (80% at 12 dB), making it more robust against adaptive deepfake threats. The effectiveness of ANCHOR-Grid’s ENF-based authentication is detailed in

Table 6. ANCHOR-Grid performance vs. traditional methods.

Method	Detection Accuracy	Noise Resilience (12 dB SNR)	Computational Overhead	Reference
ANCHOR-Grid (ENF)	99.9%	90%	20% higher	[7]
ECC-Based	95%	85%	Baseline	[7]
TLS-Based	92%	80%	10% lower	[7]

, which compares its performance to ECC and TLS. ANCHOR-Grid achieves a 99.9% detection accuracy, surpassing ECC’s 95% and TLS’s 92%, though it requires 20% higher computational resources.

6.4. Implications and Future Directions

ANCHOR-Grid underscores the transformative potential of the ENF in bridging physical and virtual realms, aligning with the need of the Metaverse for trustworthy ecosystems (Section 1). Recent cost analyses suggest that integrating ENF signatures with blockchain and real-time analytics could reduce outage-related losses by up to 20%, saving USD billions annually [7]. Future work should explore multimodal ENF fusion with voltage harmonics and current transients to enhance detection specificity by 10–15% (Section 5.3.5). Extending ANCHOR-Grid to domains like transportation or healthcare could further leverage the grid-specific signatures of the ENF, strengthening its role as a cornerstone for cyber–physical security in decentralized Metaverse infrastructures.

While the ANCHOR-Grid framework addresses security challenges in smart grid DTs, its design principles and ENF-based authentication methods are inherently adaptable to other critical domains within the Metaverse. The ability of ENF signatures to anchor virtual operations to real-world events offers a versatile foundation for enhancing security and trust in energy infrastructures and emerging applications such as VR training, financial transaction platforms, and healthcare systems. The next section explores these broader applications, highlighting how the core features validated in the ANCHOR-Grid deployment can be extended to diverse use cases within the expanding Metaverse ecosystem.

6.5. Limitations and Challenges of ANCHOR-Grid

Despite its strengths in authenticating smart grid digital twins (DTs) using Electric Network Frequency (ENF) signals, the ANCHOR-Grid framework faces several limitations that warrant consideration for practical deployment in Metaverse environments. One key challenge is sensitivity to environmental noise and grid variability, which can degrade ENF signal correlation below the 0.8 Pearson threshold in low signal-to-noise ratio (SNR) conditions (e.g., <10 dB), leading to increased false positives or negatives in deepfake detection [3]. For instance, in decentralized Metaverse platforms spanning multiple grid regions (e.g., 50 Hz in Europe vs. 60 Hz in the US), mismatches in reference databases may reduce overall accuracy by 8–10%, as observed in ENF-based forensics studies. Additionally, computational overhead remains a concern for resource-constrained devices like VR headsets or edge IoT nodes, where real-time ENF extraction and validation could increase latency by 20–30% compared to baseline methods, potentially impacting immersive user experiences in virtual grid simulations [7]. Adversarial techniques, such as synthetic ENF generation via AI, pose another risk, as attackers could forge signals to bypass authentication, underscoring the need for hybrid defenses like integration with blockchain for enhanced tamper resistance. These limitations highlight areas for future refinement, ensuring ANCHOR-Grid’s robustness in evolving Metaverse threats.

7. ENF in Emerging Metaverse Applications

Building on the empirical findings and practical insights from the ANCHOR-Grid case study, we extend our analysis to consider how ENF-based security mechanisms can be applied beyond the context of the smart grid. As the Metaverse continues to permeate multiple sectors, the integration of the ENF as an environmental fingerprint holds promise for enhancing authenticity, integrity, and operational security across a wide range of virtual applications. This section investigates the role of the ENF in securing VR training environments, financial systems, and healthcare applications, providing a roadmap for scalable and adaptable security solutions grounded in physical reality.

7.1. ENF in VR Training for Critical Infrastructure Operators

VR training environments offer a powerful platform for preparing critical infrastructure operators, such as power grid technicians and emergency responders, to handle real-world scenarios in a risk-free setting. However, the authenticity of training data in VR simulations is crucial to ensure operators are prepared for actual grid conditions. The ENF can improve VR training by providing a real-world anchor to authenticate and timestamp simulated scenarios, ensuring that virtual events accurately reflect the physical dynamics of the grid [7]. For example, ENF signals can be embedded in VR training datasets to validate the timing and conditions of simulated grid failures, preventing adversaries from manipulating training scenarios to mislead operators [183]. In multimedia forensics, the ENF has achieved $92\%$ precision in authenticating audio recordings by aligning embedded frequency patterns with reference databases, a capability that can be adapted to validate VR training data [3]. This approach could reduce training errors by an estimated 20–30% by ensuring that simulations are grounded in real-world grid behavior, improving operator preparedness for emergencies.

Despite its potential, the application of the ENF in VR training environments presents challenges, particularly related to the real-time extraction and integration of ENF signals into resource-constrained VR headsets. Lightweight signal processing techniques, such as those used in digital forensics for timestamped data retrieval, can address these constraints by isolating ENF signals with minimal computational overhead [3]. Furthermore, privacy concerns arise, as ENF data can reveal the location of training facilities, which requires the use of privacy-preserving methods, such as differential privacy, to anonymize the data [122]. Future research should focus on optimizing ENF extraction for VR devices, using edge computing to process signals locally, and developing ethical guidelines to ensure responsible use in training applications.

7.2. ENF in Financial Systems Within the Metaverse

Financial markets in the Metaverse, such as virtual asset trading platforms and energy credit markets, are increasingly vulnerable to fraud and manipulation due to their reliance on digital transactions. The ENF can secure these systems by providing a real-world anchor for transaction timestamps, ensuring the temporal integrity of financial operations. By embedding ENF signals into blockchain transactions, financial platforms can verify the authenticity of trade timestamps, thereby reducing the risk of fraudulent activities such as double spending or timestamp manipulation [1]. For example, in virtual energy markets, the ENF can authenticate the timing of energy credit trades, potentially reducing fraud by 15–20% by ensuring that transactions align with real-time grid conditions [1]. This approach leverages the stochastic nature of the ENF to prevent replay attacks, as historical ENF data cannot be reused to validate new transactions, enhancing security in decentralized financial systems [1].

Integrating the ENF with smart contracts further strengthens financial security in the Metaverse by ensuring that automated transactions are executed only when ENF signatures match real-time grid conditions. Blockchain technology, already used to log ENF data for grid operations, can be adapted to financial systems, providing an immutable record of transaction timestamps [1]. However, privacy concerns must be addressed, since ENF data in financial transactions must comply with regulations such as the GDPR, which requires the use of privacy enhancement technologies, such as differential privacy, to protect user identities [1]. Future research should investigate the scalability of ENF-based financial authentication, focusing on optimizing signal embedding for high-frequency trading environments and ensuring compliance with global financial regulations.

7.3. ENF in Healthcare Metaverse Applications

Healthcare applications in the Metaverse, such as remote surgeries and patient monitoring, rely on the secure and timely transmission of medical data, making them vulnerable to tampering and latency-induced errors. The ENF can enhance the security of these applications by authenticating data streams from medical devices, ensuring they are synchronized with the physical hospital grid. For example, in remote surgeries, the ENF can validate the timing of surgical commands, ensuring that actions are executed in real time with the hospital’s grid conditions, potentially reducing latency-induced errors by $10\%$ [1]. This capability is particularly valuable in IoT-driven healthcare systems, where wearable devices continuously monitor patient health, and the ENF can detect tampering by comparing device output to expected grid frequencies [1].

The use of the ENF in healthcare also aligns with broader IoT security practices, such as biometric-based access control, by providing an additional layer of authentication for medical devices [1]. However, ethical considerations are crucial, as the integration of ENF data into patient records must protect sensitive health information. Homomorphic encryption can process ENF data without exposing patient identities, ensuring compliance with privacy standards such as GDPR [1]. Future research should focus on developing lightweight ENF authentication protocols for medical devices, exploring their integration with real-time health monitoring systems, and addressing ethical challenges to ensure patient trust and data security in the Metaverse.

The successful deployment of the ENF in both the ANCHOR-Grid framework and broader Metaverse applications highlights its transformative potential as a foundational security layer. However, realizing the full promise of the ENF in increasingly complex and distributed virtual ecosystems will require ongoing research and innovation, especially to address challenges of scalability, privacy, and interoperability. The following section outlines key research directions and open problems that must be addressed to ensure the robustness, reliability, and ethical deployment of ENF-driven security frameworks in the Metaverse.

8. Future Research Directions in ENF Security for the Metaverse

Integrating the ENF into Metaverse security infrastructures remains a transformative innovation, bridging the virtual and physical realms to safeguard critical assets, such as smart grids. As the Metaverse evolves into a decentralized immersive platform, the dynamic variability, geolocation specificity, and linkage to physical power systems of the ENF make it a robust tool against cyber–physical threats. However, challenges such as scalability, privacy preservation, and distributed performance persist. As shown in

Table 7. Synthetic summary of Section 8: future research directions in ENF security for the Metaverse.

Subsection	Focus	Key Content Summary
Multimodal Forensic Analysis with ENF	Integrating ENF with other forensic signals	Investigating fusion of ENF with GPS, biometrics, and acoustic signals for improved traceability and robustness in complex Metaverse environments.
Privacy-Preserving ENF Processing	Secure ENF signal handling	Developing privacy-preserving mechanisms for ENF collection and processing, such as homomorphic encryption and federated learning.
Edge-Oriented ENF Solutions	Real-time ENF analysis at the edge	Implementing lightweight ENF extraction and authentication directly on edge devices for decentralized and latency-sensitive environments.
Distributed ENF Security Frameworks	Collaborative ENF-based authentication	Designing frameworks that distribute ENF verification tasks across networked nodes to enhance fault tolerance and reduce bottlenecks.
Precise Anomaly Localization	Fine-grained security incident detection	Enhancing spatial resolution of ENF deviation detection to localize attacks or inconsistencies in complex virtual–physical systems.
Interdisciplinary Applications of ENF	Broader use of ENF across domains	Exploring use cases of ENF in digital art verification, e-health integrity, e-voting, and virtual law enforcement.
Standardization of ENF for Metaverse Security	Unified protocols and benchmarks	Developing international standards, APIs, and interoperability protocols for reliable and consistent ENF implementation.
Sociotechnical Implications of ENF Deployment	Human factors and social adoption	Studying trust, usability, and behavioral responses to ENF authentication in immersive systems.
Governance and Ethical Standards	Policy and ethical concerns	Creating transparent regulatory frameworks to oversee ENF-based security in decentralized, cross-border Metaverse applications.
Specific Research Challenges	Key technical gaps	Addressing limitations in noise resilience, real-time ENF matching, and accurate synchronization in adversarial conditions.
Enhancing ENF Reliability in Real-World Grids	Improving ENF signal stability	Studying temporal and spatial frequency variance in different grid types (urban/rural/microgrids) for more dependable ENF use.
Adaptive Thresholding for Environmental Variability	Dynamic ENF authentication models	Developing context-aware thresholds to maintain accuracy despite changes in ambient noise, load fluctuation, or grid behavior.
Hybrid ENF Systems with Complementary Signals	Multimodal verification frameworks	Combining ENF with complementary ambient signals (e.g., vibration, EMI) for layered authentication.
ENF Resilience Testing in Adverse Scenarios	Stress-testing ENF under attacks	Simulating replay, substitution, and compression attacks to measure the robustness and limitations of ENF-based techniques.
Ethical and Governance Considerations in Reliability-Enhancing Methods	Ethics in system design	Ensuring reliability-enhancing methods do not compromise transparency, accessibility, or digital equity in ENF systems.

below, this section builds on previous research directions, incorporating updated empirical insights and enhanced methodologies, and leverages existing studies to advance the role of the ENF in protecting the Metaverse.

8.1. Multimodal Forensic Analysis with ENF

8.1.1. Integrating ENF with Physical-Layer Signals

Advancements in fusion of the ENF with physical-layer signals, such as voltage phase angles, harmonic distortions, and impedance metrics, enable multidimensional anomaly detection. PMUs, with microsecond-level synchronization, provide contextual data that, when paired with the ENF, enhances forensic precision. For instance, a 0.03 Hz ENF deviation coupled with a 5-degree voltage phase shift can confirm a cyber–physical disturbance with 30% fewer false positives than single-signal methods [44]. Future efforts should explore real-time integration of additional signals, such as current transients, to further reduce detection latency by 15–20%, bolstering resilience in digital twin environments.

8.1.2. ENF and Network Traffic Correlation

Correlating ENF fluctuations with network traffic anomalies offers a powerful approach to detecting orchestrated attacks. Advanced LSTM models can identify synchronous patterns, such as a 0.04 Hz ENF anomaly paired with a 100 ms SCADA latency spike, with 95% accuracy [1]. Recent optimizations have reduced processing overhead from 20 to 30% to 10–15% by incorporating attention mechanisms, improving real-time feasibility. Future research should focus on hybrid models combining LSTM with graph neural networks to capture spatial–temporal dependencies in large-scale Metaverse networks, potentially increasing detection specificity by 5–10%.

8.2. Privacy-Preserving ENF Processing

8.2.1. Secure Multiparty Computation (SMPC)

SMPC enables collaborative ENF analysis across global Metaverse infrastructures without exposing sensitive datasets. Recent SMPC protocols achieve 88% anomaly detection accuracy while maintaining differential privacy with an epsilon of 0.1 [74]. To address encryption-induced signal degradation, new algorithms that incorporate noise-resistant feature extraction have reduced accuracy loss to under 5%. Future work should optimize SMPC for heterogeneous datasets, aiming for 90%+ accuracy across diverse grid environments, ensuring privacy without compromising detection reliability.

8.2.2. Homomorphic Encryption for ENF Analysis

Homomorphic encryption, particularly partially homomorphic encryption (PHE) schemes such as Paillier, strikes a balance between security and performance in ENF analysis. Recent studies report 92% accuracy with subsecond latencies using PHE, enhanced by GPU acceleration [143]. Integrating vectorized computations has further reduced latency by 20%, making encrypted ENF processing viable for real-time grid monitoring. Future research should explore hybrid PHE-FHE frameworks to support complex computations, such as multisignal fusion, while maintaining latency below 500 ms for Metaverse-scale applications.

8.3. Edge-Oriented ENF Solutions

Lightweight ENF models are critical for edge devices in expansive Metaverse infrastructures. Quantized neural networks (QNNs) now achieve 96% anomaly detection accuracy with 75% lower memory usage, enabling deployment on resource-constrained devices such as the Raspberry Pi [36]. Recent advances in model pruning have reduced the inference time to 40 ms, further cutting bandwidth demands by 80%. Future efforts should focus on adaptive quantization techniques to dynamically adjust model complexity based on device capabilities, improving scalability for virtual environments driven by the IoT by 10–15%.

8.4. Distributed ENF Security Frameworks

With blockchain integration, FL enables distributed ENF anomaly detection while preserving data privacy. Recent tests demonstrate that blockchain-backed FL detects anomalies across 100 virtual sensors in 450 ms, a 10% speed improvement over prior results [184]. Incorporating differential privacy in FL models has reduced risk of privacy leakage by 15% without sacrificing accuracy. Future research should explore asynchronous FL protocols to handle intermittent connectivity in edge-heavy Metaverse grids, potentially reducing synchronization delays by 20%.

8.5. Precise Anomaly Localization

Combining the ENF with geolocation metadata improves fault localization precision. Convolutional neural networks (CNNs) trained in ENF and GPS data now locate incidents within 0.8 to 1.5 km, reducing response times by 35% [48]. Recent optimizations using transfer learning have reduced computational requirements by 25%, enabling deployment on resource-limited devices. Future work should integrate temporal ENF trends with 3D spatial modeling to achieve subkilometer precision, enhance predictive maintenance tools for Metaverse dashboards, and prevent outages with greater efficiency 10%.

8.6. Interdisciplinary Applications of ENF

The potential of the ENF extends beyond technical applications, offering opportunities for interdisciplinary research that bridges engineering with fields like sociology, psychology, and urban planning in the Metaverse. In urban planning, the ENF can support simulations of virtual cities by authenticating energy consumption data, ensuring that digital models accurately reflect real-world grid dynamics. For instance, the ENF can validate energy usage patterns in virtual city simulations, improving planning efficiency by an estimated 10–15% by providing reliable data for infrastructure design [6]. This capability is particularly valuable in the Metaverse, where urban planners can test scenarios such as renewable energy integration without physical risks, thereby enhancing the sustainability of future cities.

The ENF can enhance VR sessions by validating the integrity of session data, ensuring that data are based on accurate timestamps and conditions. This application is critical in the Metaverse, where VR is increasingly used for real-time applications and manipulated data could undermine results [108]. ENF-based validation could improve the reliability of VR therapy by 10–15%, ensuring that session logs are tamper-proof and aligned with real-world conditions [108]. Future research should focus on integrating the ENF with other models to enhance user trust in virtual environments, exploring how the ENF can support interdisciplinary applications that address technical and human-centric challenges in the Metaverse.

8.7. Standardization of ENF for Metaverse Security

The lack of standardized protocols for ENF use in the Metaverse hinders its global adoption, as inconsistent methods can lead to unreliable authentication and legal disputes. In multimedia forensics, nonstandardized ENF logging has resulted in conflicting results across jurisdictions, a challenge that could extend to Metaverse applications where the ENF is used for security [3]. For example, the absence of unified ENF standards complicates the validation of virtual transactions across regions with different grid frequencies. Developing standardized ENF protocols is essential to ensure interoperability and reliability, potentially reducing adoption costs by streamlining implementation processes [6].

IEEE, which has established standards for the forensic applications of the ENF, can play a key role in this standardization effort by creating protocols tailored for the Metaverse [1]. Such standards define best practices for the extraction, recording, and validation of ENF signals, ensuring consistency between platforms and jurisdictions. Future research should prioritize collaboration with regulatory bodies to develop these standards, exploring how they can be integrated with existing cryptographic frameworks to improve ENF-based security. Additionally, research should address the scalability of standardized ENF protocols, ensuring that they can support the diverse and dynamic nature of Metaverse environments without compromising performance.

A critical aspect of standardization involves addressing the variability in the characteristics of the ENF signal in different power grids, which can affect the authentication accuracy in Metaverse global applications. For example, variations in grid stability between developed and developing regions can introduce noise in ENF signals, potentially reducing authentication reliability by 5–10% in unstable grids [48]. To mitigate this, future standards should incorporate adaptive signal processing techniques, such as wavelet transforms, to normalize ENF signals across diverse environments, improving consistency [185]. Collaborative efforts with international grid operators could facilitate the creation of a global reference database of the ENF, enabling standardized validation and potentially increasing the accuracy of authentication by 10–15% [3]. This standardized approach would enhance security and foster the broader adoption of the ENF in the Metaverse by providing a reliable framework for developers and policymakers.

8.8. Sociotechnical Implications of ENF Deployment

The deployment of the ENF in the Metaverse has significant sociotechnical implications, particularly concerning public perception of privacy risks versus security benefits. Surveys indicate a $30\%$ higher user concern for privacy in immersive interfaces, as users fear that ENF data could be used to track their locations or activities [48]. This concern is particularly pronounced in applications like virtual healthcare or financial systems, where sensitive data is involved, necessitating transparent policies to ensure user trust. Addressing these concerns requires a sociotechnical approach that balances the security benefits of the ENF with user acceptance, potentially increasing adoption rates by 10–20% through clear communication of its privacy safeguards [48].

Another sociotechnical challenge is to ensure equitable access to ENF-based security in underresourced regions, where the grid infrastructure may be less reliable or standardized [48]. For example, areas with unstable grids may experience higher ENF variability, which can complicate authentication processes and potentially exclude them from Metaverse security benefits. Future research should investigate adaptive ENF frameworks that accommodate regional grid differences, ensuring that security solutions are accessible to all users. Additionally, sociotechnical studies should investigate the long-term societal impact of ENF deployment, examining how it influences user behavior, trust in virtual systems, and the broader adoption of Metaverse technologies.

The sociotechnical implications of the ENF also extend to workforce dynamics, as its implementation may require new skills to manage and interpret ENF data in virtual environments. For instance, Metaverse platform operators may need training to handle ENF-based authentication systems, potentially increasing operational costs by 5–10% in the short term due to the need for specialized expertise [6]. However, this investment could lead to long-term benefits, such as improved system reliability and user trust, fostering greater adoption of Metaverse technologies. Future research should explore the development of training programs that integrate ENF management into existing cybersecurity curricula, ensuring that the workforce is prepared for widespread adoption. Furthermore, sociotechnical studies should examine how the implementation of the ENF influences user perceptions of fairness and inclusion, ensuring that its benefits are equally distributed in diverse Metaverse communities [108].

8.9. Governance and Ethical Standards

Robust governance frameworks are essential for the ethical use of ENF data. Policies enforcing differential privacy (e.g., epsilon of 0.05) and 256-bit AES encryption have reduced privacy violation risks by 25% [6]. Standardized protocols across utilities have lowered adoption costs by 25–35%, promoting equitable access. Future research should develop adaptive governance models that dynamically adjust privacy parameters according to threat levels, ensuring compliance with global regulations while maintaining cost savings of 20% to 30%. Community-driven standardization efforts can further mitigate digital security disparities across regions.

8.10. Specific Research Challenges

Low signal-to-noise ratio (SNR) environments (10 dB) reduce ENF extraction accuracy by 8%, posing challenges for reliable analysis [9]. Edge-oriented QNNs achieve 96% accuracy with 75% lower memory usage, but scalability requires further optimization [36]. The variability of the PMU correlation (0.68–0.72) limits precision, necessitating hybrid signal approaches [48]. Key research questions include the following: How can multisignal fusion mitigate SNR challenges in ENF analysis? and What adaptive quantization techniques can enhance edge computing scalability?

Table 8. Future research challenges and metrics.

Challenge	Current Metric	Proposed Improvement
Low SNR (10 dB)	8% accuracy loss [9]	Multisignal fusion (+10%)
Edge computing scalability	96% accuracy, 75% memory reduction [36]	Adaptive quantization (+15%)
PMU correlation variability	0.68–0.72 correlation [48]	Hybrid signals (+15% accuracy)

presents critical challenges in ENF security, such as low-SNR environments and PMU correlation variability, along with proposed improvements, including multisignal fusion, which aims to enhance detection accuracy by 10–15%. These challenges underscore the need for targeted research to realize the ENF’s full potential in Metaverse applications.

8.11. Enhancing ENF Reliability in Real-World Grids

Current direct sensing approaches using PMUs achieve relatively low detection accuracies (18–38%) compared to audio-based indirect methods (75–92%) due to grid-induced variability and low correlation between measurement points. Research is needed to develop standardized ENF estimation algorithms that adapt to fluctuating SNRs and regional grid characteristics. These algorithms should incorporate dynamic filtering, machine learning-based denoising, and probabilistic modeling to reduce variability and enhance timestamping precision. For instance, using adaptive Kalman filters or temporal ensemble smoothing could yield up to a 10% increase in timestamping accuracy across distributed environments.

8.12. Adaptive Thresholding for Environmental Variability

Fixed thresholds, such as the 0.8 Pearson correlation coefficient used in ANCHOR-Grid, may underperform under high-jitter or low-SNR scenarios. Future work should investigate context-aware, adaptive threshold mechanisms that dynamically adjust based on environmental conditions such as load variability, regional noise profiles, or network latency. Techniques such as reinforcement learning-based threshold tuning or context-sensitive anomaly scoring could reduce false positives by an estimated 15% and improve reliability in unpredictable Metaverse grid environments.

8.13. Hybrid ENF Systems with Complementary Signals

Research should explore hybrid security mechanisms that combine the ENF with complementary physical-layer signals to address the limitations of single-signal approaches. These include voltage harmonics, current transients, and phase-angle measurements. For example, integrating the ENF with voltage harmonics has already shown a 15% increase in detection accuracy for tampering and FDI attacks. Multisignal fusion frameworks leveraging sensor fusion algorithms (e.g., Bayesian inference, cross-modal transformers) could enhance anomaly detection and reduce dependence on ideal sensing conditions.

8.14. ENF Resilience Testing in Adverse Scenarios

Future ENF frameworks should be evaluated under more challenging scenarios, including highly unstable grids, high-jitter communication networks, and adversarial noise injection. Benchmarking ENF extraction under conditions such as SNR < 10 dB, 200 ms jitter, and 5% packet loss will inform robust design principles. Simulated environments replicating these conditions should be developed using digital twin testbeds to fine-tune sensing, authentication, and anomaly detection algorithms.

8.15. Ethical and Governance Considerations in Reliability-Enhancing Methods

As reliability-improving techniques often involve deeper ENF analysis and more granular timestamping, they risk encroaching on user privacy by increasing the precision of geolocation or user tracking. Future research must balance these enhancements with privacy-preserving approaches such as differential privacy, secure multiparty computation (SMPC), and zero-knowledge proofs. Establishing governance standards that mandate transparency, user consent, and ethical use of the ENF in sensitive applications is essential for widespread adoption.

9. Conclusions

Integrating ENF signals into the cybersecurity architecture of the Metaverse and smart grid ecosystems represents a crucial advancement in addressing modern cyber–physical vulnerabilities. The ENF’s inherently stochastic and geographically distinct properties offer a new form of environmental fingerprinting that can securely link virtual interactions to real-world electrical infrastructure. This survey has shown that ENF-based methods significantly improve the CIA triad through various mechanisms, including dynamic encryption key generation, timestamped watermarking, anomaly detection, and device authentication.

The ANCHOR-Grid framework, as a case study, exemplifies the practical deployment of the ENF in real-world digital twin environments, achieving up to 99.9% detection accuracy under low-latency conditions. When combined with advanced tools such as wavelet transforms, STFT, and machine learning algorithms, the ENF enables highly reliable detection of deepfakes, replay attacks, and FDI threats that conventional cryptographic methods often struggle to mitigate. Furthermore, integrating the ENF with decentralized technologies like blockchain and FL facilitates resilient, tamper-evident, and scalable solutions across distributed Metaverse infrastructures.

Beyond energy systems, the ENF shows considerable promise in emerging sectors of the Metaverse, including virtual healthcare, financial services, immersive education, and remote training simulations. In these contexts, the ENF enhances the data’s authenticity, temporal validation, and operational integrity, building trust in real-time high-stakes applications. Edge-oriented adaptations, such as QNNs and privacy-preserving techniques like differential privacy and homomorphic encryption, further expand the reach of the ENF into resource-constrained and regulation-sensitive environments.

However, several limitations must be addressed before the ENF can achieve widespread adoption. Direct sensing methods, while precise, face scalability and performance challenges due to low detection accuracy (18–38%) and high sensitivity to environmental noise. Although more accurate, indirect sensing introduces latency and processing overhead that could impede real-time responsiveness. Furthermore, the lack of standardized protocols for the extraction, logging, and validation of ENF signals complicates cross-border interoperability and legal compliance, especially in jurisdictions governed by strict privacy laws, such as the GDPR.

Ethical and sociotechnical concerns are equally critical. The ability of the ENF to infer the geographical location from frequency patterns raises concerns about user surveillance and consent in immersive environments. These risks are particularly pronounced in sectors that handle sensitive data, such as healthcare and finance, where the misuse of location-based data could lead to discriminatory outcomes or violations of digital rights. Furthermore, equitable access to ENF-enabled infrastructure remains unresolved, particularly for underresourced regions with unreliable grid conditions or limited computational capacity.

Looking ahead, several key research directions should be prioritized. First, improving ENF robustness through multisignal fusion, combining the ENF with voltage harmonics, current transients, and phase-angle data, can enhance accuracy and resilience under low-SNR conditions. Second, using FL and asynchronous edge computing protocols may help address the challenges of real-time distributed analytics. Third, regulatory and standardization efforts, potentially led by organizations such as IEEE, must establish global norms for the usage of the ENF, ensuring legal clarity, data privacy, and interoperability. Lastly, human-centered design principles must guide the deployment of ENF systems, ensuring that security solutions are inclusive, transparent, and ethically aligned with the diverse needs of the users.

In conclusion, the ENF represents a powerful physics-based security primitive that uniquely connects virtual operations to tangible physical phenomena. Although not a standalone solution, the ENF will play a critical role in a multi-layered, context-aware, and ethically grounded cybersecurity paradigm. As the Metaverse continues to evolve into a decentralized nexus for industrial operations, financial transactions, and human interaction, the ENF provides a viable and verifiable mechanism to ensure that these environments remain secure, trustworthy, and resilient.