Security and Privacy in the Internet of Everything (IoE): A Review on Blockchain, Edge Computing, AI, and Quantum-Resilient Solutions

Abstract

The IoE forms the foundation of the modern digital ecosystem by enabling seamless connectivity and data exchange among smart devices, sensors, and systems. However, the inherent nature of this structure, characterized by high heterogeneity, distribution, and resource constraints, renders traditional security approaches insufficient in areas such as data privacy, authentication, access control, and scalable protection. Moreover, centralized security systems face increasing fragility due to single points of failure, various AI-based attacks, including adversarial learning, model poisoning, and deepfakes, and the rising threat of quantum computers to encryption protocols. This study systematically examines the individual and integrated solution potentials of technologies such as Blockchain, Edge Computing, Artificial Intelligence, and Quantum-Resilient Cryptography within the scope of IoE security. Comparative analyses are provided based on metrics such as energy consumption, latency, computational load, and security level, while centralized and decentralized models are evaluated through a multi-layered security lens. In addition to the proposed multi-layered architecture, the study also structures solution methods and technology integrations specific to IoE environments. Classifications, architectural proposals, and the balance between performance and security are addressed from both theoretical and practical perspectives. Furthermore, a future vision is presented regarding federated learning-based privacy-preserving AI solutions, post-quantum digital signatures, and lightweight consensus algorithms. In this context, the study reveals existing vulnerabilities through an interdisciplinary approach and proposes a holistic framework for sustainable, scalable, and quantum-compatible IoE security.

1. Introduction

1.1. IoE and Security: Surge in Security and Privacy Issues Driven by the Proliferation of IoE

The IoE represents an extended paradigm of the Internet of Things (IoT), establishing large-scale interconnections among devices, people, data, and processes [1,2,3]. IoE has rapidly evolved into a critical enabler across multiple domains, including smart cities, healthcare systems, industrial infrastructures, and intelligent transportation networks. However, the expansion of IoE has also significantly intensified security and privacy concerns [4,5,6].

In IoE environments, millions of connected devices interact with each other and with decentralized infrastructures, generating critical security requirements such as authentication, data integrity, access control, and resilience against cyber-attacks [7,8,9]. Traditional security mechanisms, which typically rely on centralized systems for authentication and access management, are increasingly inadequate for large-scale, distributed IoE architectures [10,11,12].

Risks such as cyber-attacks, data breaches, identity theft, and the misuse of autonomous devices pose substantial threats within the IoE ecosystem [13,14]. Moreover, AI-driven attacks, vulnerabilities in edge computing infrastructures, and the potential of quantum computers to break classical encryption algorithms are projected to become major security challenges in the future landscape of IoE [15,16,17,18].

Considering the issues addressed above, the core security threats in IoE systems can be summarized as follows:

• Scalability Challenges: The explosive growth of connected devices generates enormous volumes of data, challenging the capacity of centralized security architectures.
• Data Privacy Risks: IoE environments collect sensitive personal, financial, and medical data, often requiring privacy-preserving AI techniques to ensure confidentiality.
• Cybersecurity Threats: IoE is inherently vulnerable to Man-in-the-Middle (MITM) attacks, Distributed Denial of Service (DDoS) attacks, and adversarial AI-driven threats.
• Quantum Computing Risks: The rise in quantum computing threatens traditional cryptographic standards, making cryptographic agility and post-quantum security essential requirements.

1.2. Emerging Security Paradigms in IoE: Blockchain, Edge Computing, AI, and Quantum-Resilient Solutions

To ensure the security of IoE systems, several modern technologies have been proposed, including Blockchain, Artificial Intelligence (AI), Edge Computing, and Quantum-Resilient Cryptography:

• Blockchain for Secure IoE Operations and Data Integrity: Blockchain offers a robust solution for ensuring data integrity and access control through its distributed ledger architecture. Particularly, Decentralized Identity (DID) and Self-Sovereign Identity (SSI) models provide decentralized security frameworks for authorization and secure data sharing in IoE environments [19,20,21].
• Edge Computing for Privacy-Preserving and Low-Latency Security: Edge Computing enables IoE devices to process large volumes of data at the network edge rather than relying on centralized cloud servers. While this approach reduces latency, it also introduces new security threats [22,23,24]. Edge AI shifts computation to local edge nodes, such as IoT gateways, 5G edge servers, and AI-enabled microcontrollers, minimizing the need to transfer data to central cloud servers. This facilitates real-time threat detection, AI-driven anomaly detection, and localized privacy-preserving processing [25,26,27,28].
• AI-Based Security Mechanisms: AI is increasingly used in security mechanisms such as threat detection, anomaly identification, and intelligent access control. However, AI systems themselves are vulnerable to attacks, including adversarial AI threats [29,30,31,32].
• Quantum-Resilient Security for Future-Proofing IoE: Quantum computing poses significant risks to classical encryption schemes, making quantum-resilient cryptographic methods essential for long-term IoE security. Approaches such as lattice-based cryptography, hash-based signatures, and multivariate polynomial cryptographic techniques offer quantum-resistant alternatives to traditional cryptographic standards [33,34,35].

Despite their individual strengths, none of these technologies can comprehensively secure IoE systems on their own. Existing solutions lack a unified and systematic approach and present different advantages and limitations depending on the specific application domain. Therefore, there is a critical need for hybrid and integrated security frameworks that combine AI, Blockchain, Edge Computing, and Quantum-Resilient Technologies for holistic IoE security.

This combination has been selected because it reflects the most influential and complementary directions currently shaping IoE security research and practice. Blockchain ensures decentralized trust and data integrity, Edge Computing enables low-latency and privacy-preserving processing, AI provides intelligent and adaptive threat detection, and Quantum-Resilient Cryptography addresses the long-term risks posed by quantum computing. Together, these domains form a holistic foundation capable of addressing diverse IoE security challenges more effectively than any single technology in isolation.

This study presents a comprehensive review that synthesizes Blockchain for IoE trust management, Edge AI for cybersecurity, and Quantum-Resilient Cryptography for cryptographic robustness into an integrated security model for IoE. The intersections of these technologies are systematically analyzed across key areas such as:

• IoE Security and Privacy Risks
• Blockchain-Based Trust and Data Integrity in IoE
• Low-Latency, Privacy-Preserving Security through Edge AI
• Quantum-Resilient Cryptographic Solutions for Future IoE Networks

Furthermore, this study classifies and evaluates existing security techniques, examines their applicability and performance trade-offs, and explores the feasibility of their implementation in IoE environments. The insights provided are intended to guide future research towards the development of next-generation security architectures that balance decentralization, efficiency, and quantum-resilient security.

1.3. Main Contributions of the Study

This study presents a comprehensive review and synthesis of state-of-the-art security paradigms for the IoE. The main contributions are as follows:

• Layered Classification of IoE Security and Privacy Risks: Security threats across all IoE architecture layers have been systematically classified, providing a holistic understanding of vulnerabilities and facilitating the development of targeted countermeasures.
• Comparative Analysis of Key Security Paradigms: Blockchain, Edge AI, and Quantum-Resilient Cryptography have been comparatively analyzed, with their strengths, weaknesses, and application scenarios identified, and IoE-specific threats mapped to the most suitable defense mechanisms.
• Evaluation of Privacy-Preserving AI and Post-Quantum Solutions: Privacy-preserving AI techniques such as Federated Learning, Differential Privacy, and Secure Multi-Party Computation, along with post-quantum cryptographic methods, have been analyzed, and their practical applicability in IoE environments has been assessed.
• Proposed Integrated and Hybrid Security Architecture for IoE: An integrated layered security architecture combining Blockchain, AI, and Quantum-Resilient Technologies has been proposed, enhancing authentication, data integrity, and threat mitigation. Its functional components and workflows have been described with attention to both theoretical and practical deployment aspects.
• Identification of Future Research Directions and Remaining Literature Gaps: Current limitations and literature gaps have been identified, and clear priorities have been outlined for advancing hybrid, scalable, and quantum-resilient IoE security solutions.

1.4. Methodology

This study is based on a multidimensional literature review focusing on emerging security solutions at the intersection of AI, IoE, Blockchain, and Quantum Computing technologies. The methodological approach is based on both the selection of technical literature and the structuring of content along specific thematic axes. The methodological framework is explained step by step below:

• Database and Publication Selection: Studies published in peer-reviewed journals indexed in leading academic databases such as IEEE Xplore, Elsevier ScienceDirect, Springer, MDPI, Wiley, ACM, and Taylor & Francis were selected. In the selection process, no publisher priority was applied; instead, the content relevance, contribution value, and recency of the studies were prioritized.
• Inclusion Criteria: The following criteria were used to include literature in the study:
  • Studies analyzing multi-layered security threats (physical, network, data, application layers) in IoE environments,
  • Architectures integrating blockchain-based decentralized security solutions into IoE,
  • AI-supported security models (particularly Edge AI, Federated Learning, Anomaly Detection),
  • Studies focusing on post-quantum cryptography (PQC) and quantum-resilient system solutions,
  • Literature examining methods such as ABAC, SMPC, and Zero-Knowledge Proof in the context of data privacy and access control,
  • Publications that include systematic proposals, architectural models, and comparative analyses at the intersection of multiple technologies.
• Keyword-Based Search: The literature review was conducted using keyword combinations. Prominent terms include:
  “Edge AI security”, “Blockchain-enabled IoT”, “quantum-resistant cryptography”, “ABAC in IoT”, “Federated Learning for edge devices”, “IoE security threats”, “IoT anomaly detection”, “zero-trust IoT architecture”, “quantum-safe identity”, “cross-domain healthcare blockchain”, and similar expressions.
• Thematic Structuring of Content: The content of the study was designed based on a predefined thematic plan (table of content). The main topics of the review, IoE security, privacy, and quantum-resilient solutions, have been systematically structured within a multi-layered content framework.

In the first stage, the security vulnerabilities and new-generation attack types in IoE environments were addressed. Threats such as identity management, data privacy, centralized and decentralized security structures, and adversarial AI-based threats were systematically classified. Subsequently, the solution methods developed against these threats were examined under three main axes: blockchain-based architectures, end-to-end security through edge AI systems, and quantum-resilient cryptographic approaches. In this context, not only theoretical models but also technical applications based on sectoral scenarios were analyzed. Additionally, within the scope of this study, a multi-layered security architecture is proposed that integrates the mentioned security technologies both horizontally by functional layers and vertically by application domains. In the subsequent sections, comparative analyses of the security technologies were conducted based on criteria such as performance, applicability, and integration level. Various solutions were thematically classified according to security layers and system architectures.

Finally, in the last section of the study, the limitations of current security systems were comprehensively evaluated. The increasing security demands and systemic bottlenecks arising from the scaling of IoE were revealed. In this context, areas requiring architectural transformation to address post-quantum threats were highlighted, and new research topics were proposed both on technological and interdisciplinary levels.

Thus, the study provides a holistic assessment that not only classifies existing solutions but also presents long-term strategic approaches to IoE security.

5.

Presentation of Analyses and Visualization: The content presented in the study is supported by figures and tables to ensure that the reader can clearly understand the relationships, advantages, and limitations between different technologies. In this context, comparative solution maps, diagrams classifying security threats by layers, and tables showing performance differences in Edge AI devices are included.

6.

Scope and Limitations: While this review provides a comprehensive overview of AI, Blockchain, Edge Computing, and Quantum-Resilient Security solutions for IoE, it does not claim to cover the entire body of IoE security literature. Instead, the focus is on works that offer layered security perspectives and multi-technology integrations relevant to emerging IoE ecosystems.

1.5. Paper Organization

The remainder of this paper is organized as follows: Section 2 analyzes the fundamental security and privacy challenges in IoE systems, including identity management and authorization, data privacy risks, and comparisons between centralized and decentralized security models, as well as next-generation threats such as AI-based cyber-attacks and quantum computing risks. Section 3 presents state-of-the-art security and privacy-preserving solutions, covering blockchain-based mechanisms, edge AI and secure computing approaches, quantum-resilient cryptographic methods, real-world technology mapping, and the integration of different security layers, culminating in the proposed integrated layered architecture. Section 4 outlines open challenges and future research directions, addressing blockchain scalability, quantum impact, AI-driven security, interoperability, and strategic opportunities. Section 5 is the Discussions section, where existing solutions are evaluated, emerging risks are explored, and the effectiveness and limitations of core technologies are discussed, along with deployment issues, practical considerations, and a proposed research roadmap. Section 6 provides the Conclusion, summarizing key findings and outlining a clear path for advancing hybrid, scalable, and quantum-resilient IoE security solutions.

2. Security and Privacy Challenges in IoE

The IoE is a broad ecosystem that establishes large-scale connections among devices, people, processes, and data. However, the expansion of these connections also introduces significant security vulnerabilities and privacy issues [1,14]. Since IoE systems typically operate on decentralized infrastructures, they face various risks related to identity management, data privacy, attack vectors, and security threats [2,7,11].

2.1. Fundamental Security Risks in IoE

Due to the continuous data exchange among billions of connected devices, IoE systems are exposed to a wide range of security vulnerabilities [3,8]. The security challenges in IoE can be examined under the following three main categories:

2.1.1. Identity Management and Authorization Problems

Authentication and access control are fundamental components of security in the IoE ecosystem [11]. However, the large-scale nature of IoE networks leads to the inadequacy of traditional identity management systems [13]. Current identity management problems include:

• Device spoofing and identity theft: Since IoE devices often utilize decentralized authentication methods, malicious actors can impersonate legitimate devices [12,36].
• Man-in-the-Middle (MitM) attacks: Due to the lack of robust authentication mechanisms during data transmission, attackers can intercept and manipulate communications between devices [37].
• Challenges of centralized identity management: Traditional authentication systems (such as OAuth, Kerberos, etc.) are not scalable for distributed environments like IoE [38,39].

As a solution, blockchain-based identity management approaches such as DID and SSI are recommended [40,41].

2.1.2. Data Privacy and Confidentiality Risks

The large datasets continuously collected by IoE devices pose serious threats to user privacy. In IoE systems, privacy breaches generally occur through data leaks, unauthorized access, and data manipulation [14,42]. The major privacy risks are as follows:

• Lack of encryption during data storage and transmission: Since IoE devices often operate with lightweight hardware, they face difficulties in supporting strong encryption algorithms [10,43].
• Lack of anonymity: When user data is collected on centralized servers, anonymity is compromised, making user behaviors traceable [44].
• Side-channel attacks: By analyzing the energy consumption or processing times of IoE devices, attackers can infer user behavior patterns [45].

To mitigate these risks, techniques such as differential privacy, homomorphic encryption, and secure multi-party computation (SMPC) are recommended [46].

2.1.3. Comparison of Centralized and Decentralized Security Models

System architectures used in IoE security are generally divided into two main categories: centralized and decentralized security approaches.

As shown in Figure 1, the Single Point of Failure (SPOF) risk inherent in centralized structures is compared with the fault-tolerant architecture provided by decentralized security models. In the centralized security model, the security of all IoE devices is managed through a single server or authority. Although this structure offers management simplicity, it poses a risk to the entire network’s security in case of a system failure.

The SPOF is the primary vulnerability of this model [11]. In contrast, decentralized security models are based on technologies such as blockchain, edge computing, and SSI [47]. Security is ensured in a distributed manner among devices without relying on a central authority. As a result, the system becomes more resilient, and there is no single attack target. As shown in the figure, in the centralized model, all devices connect to a single server, whereas in the decentralized model, security is ensured at the edge level, and secure connections are established between the devices.

However, this structure may increase computational costs and management complexity due to cryptographic processes. The advantages and disadvantages of centralized and decentralized security models vary depending on system preferences. These differences are summarized in

Table 1. Comparison of centralized and decentralized security models.

Security Model	Advantages	Disadvantages	Recommended Use Cases
Centralized Security	Easy management, Low computational cost, Simple integration	Risk of SPOF, the Central server becomes an attack target	Small-scale IoE systems, Low-threat environments
Decentralized Security (Blockchain, Edge Security, SSI)	Distributed security, High attack resilience, Better scalability	High computational cost, Management complexity	Large-scale IoE networks, Critical infrastructures

:

2.1.4. Layered Classification of IoE Security

Various technologies are utilized to ensure security and privacy in IoE environments. Blockchain enhances security by providing decentralized and immutable records, while Edge Computing offers the advantage of low-latency, real-time data processing. On the other hand, Quantum Security provides future-proof solutions by offering encryption methods resistant to post-quantum threats. FL allows data to be processed without being sent to a central server, and Zero-Knowledge Proof (ZKP) provides a verification mechanism that prevents data leakage.

However, each method has specific strengths and weaknesses. Therefore, IoE security strategies should be built upon hybrid solutions that integrate different security layers. In this context,

Table 2. IoE security and privacy classification framework.

Security Layer	Key Technologies	Security Features	Challenges and Limitations
Data Security and Privacy	Federated Learning [23,25,47], Homomorphic Encryption [46], ZKP [48,49,50]	Privacy-Preserving AI, Secure Data Sharing	Computational Overhead, Model Poisoning Attacks
Network and Communication Security	Blockchain [8,9,10,11], SMPC [46], Quantum Key Distribution (QKD) [18,51]	Data Integrity, Tamper-Proof Transactions	Latency, Scalability Issues
AI-Driven Security	Edge AI [25,44], Anomaly Detection [26], Adversarial AI Defense [29,30]	Real-Time Threat Detection, Localized Privacy	Model Training Complexity, Limited Computational Power
Quantum-Resilient Cryptography	Lattice-Based [52,53], Hash-Based [54,55], and Multivariate Polynomial Cryptography [56]	Future-Proof Security, Resistance to Quantum Attacks	Large Key Sizes, High Processing Overhead

presents a structured classification that highlights key technologies, the security features they provide, and their fundamental limitations for each security layer.

2.2. Next-Generation Attack Vectors in IoE

Today, the IoE has formed a large-scale ecosystem where devices, systems, and users are interconnected. However, this interconnectivity has also made IoE extremely vulnerable to next-generation attack vectors. In this environment, where traditional security approaches are no longer sufficient, various risk factors have emerged, as illustrated in Figure 2, including adversarial AI attacks, model poisoning, quantum computing-based threats, and edge/cloud security vulnerabilities.

These attacks target AI models, blockchain-based security systems, cloud infrastructures, IoE networks, and edge devices, posing a significant threat to the holistic security structure of IoE. In particular, adversarial AI attacks and model poisoning directly target AI models, while quantum attacks threaten blockchain security and network communication protocols by introducing vulnerabilities that could compromise encryption and consensus mechanisms.

Table 3. Threat vectors and targeted components in the IoE ecosystem.

Attack Type	Targeted Systems	Threat Mechanism	Possible Consequences	Recommended Defense Methods
Adversarial AI [29,30]	Edge Devices, AI Models, Machine Learning Systems	Manipulation of Models Through Intentionally Misleading Inputs	Incorrect Decision-Making, System Failures	Robust Adversarial Training, Secure Model Training
Model Poisoning [57]	AI Models, AI-Based IoE Devices	Manipulation of Training Data	Faulty Training, Malicious Model Behavior	Data Sanitization, Federated Learning
Quantum Attacks (Shor, Grover) [52,53,54,55,56]	Cryptographic Systems, Blockchain, IoE Security	Breaking Classical Encryption Algorithms	RSA/ECC Encryption Becomes Obsolete	PQC, Lattice-Based Encryption
Edge/Cloud Vulnerabilities [58,59,60]	Edge Devices, IoE Gateways, Cloud Infrastructures, Data Storage Systems	Physical Interventions, Network Vulnerabilities, Malware Infections, Firmware Exploitation, Unauthorized Access, Data Leakage, Multi-Tenancy Issues	Data Manipulation, Device Security Breaches, Theft of Sensitive Data	Hardware Security, Secure Boot Mechanisms, Zero-Trust Security, Homomorphic Encryption

summarizes the prominent threat types along with their targeted components, potential consequences, and recommended solutions.

2.2.1. AI-Based Cyber Attacks: Adversarial AI, Model Poisoning

Although AI systems are widely used in IoE security, AI systems themselves can also be vulnerable to attacks [61]. IoE systems utilize machine learning and AI models to analyze data, make decisions, and optimize user interactions. However, malicious actors can manipulate these systems to exploit security vulnerabilities. In particular, Adversarial AI and Model Poisoning attacks pose significant threats to the reliability of smart systems in IoE environments.

In Adversarial AI attacks, misleading data is used to cause IoE devices to make incorrect decisions [62,63]. For example, if the facial recognition algorithm of a smart security camera is subjected to an adversarial attack, it may identify an unauthorized person as an authorized one. Similarly, model poisoning attacks inject malicious data into a machine learning model, causing the model to learn incorrectly over time [64,65]. These types of attacks target IoE devices, creating serious security vulnerabilities in automation processes. In Model Poisoning, AI systems are deliberately fed with incorrect data to manipulate training, leading to the misguidance of IoE devices

2.2.2. Quantum Computing-Based Security Threats

Traditional security mechanisms rely on asymmetric cryptography and hash functions. However, quantum computing technologies pose a threat to these cryptographic methods, with the potential to break encryption algorithms. In particular, traditional cryptography methods such as RSA and ECC are vulnerable to quantum attacks [66,67]. For instance, Shor’s and Grover’s algorithms can facilitate the breaking of current IoE security protocols.

Shor’s Algorithm has the potential to render widely used cryptographic systems like RSA and ECC ineffective [66]. Authentication and data security mechanisms used in IoE systems could be compromised by such quantum attacks. On the other hand, Grover’s Algorithm weakens hash-based security measures, posing a threat to authentication processes in IoE devices [68]. These threats represent a major risk, especially for blockchain-based IoE systems.

As a solution, new-generation security approaches such as PQC and Lattice-Based Cryptography are recommended [69,70].

2.2.3. Risks in Edge and Cloud Security for IoE

In the IoE ecosystem, edge computing and cloud systems are commonly used for data processing and data sharing between the edge and the cloud. However, security vulnerabilities can arise between these two layers, making these architectures susceptible to cyber-attacks [70]. In particular, physical attacks, hardware manipulations, and security flaws are frequently observed in edge devices. Since IoE systems generally consist of small and low-power devices, their security measures are often limited. This situation can lead to unauthorized access and the takeover of devices by attackers [70].

On the cloud side, data leaks and unauthorized access issues are more prominent. When large volumes of data generated by IoE devices are uploaded to cloud platforms, ensuring the security of this data becomes critically important. However, in centralized cloud systems, the lack of encryption or the use of weak authentication mechanisms can lead to large-scale data breaches [10].

As a result, it is essential to understand next-generation attack vectors in the IoE ecosystem and develop effective solutions against these threats. AI-driven attacks, quantum threats, and edge-cloud security vulnerabilities are among the most significant risks to the sustainability of IoE. In this context, it is crucial to work on new approaches such as PQC, trusted AI models, and advanced edge-cloud security protocols [68,69,70].

3. Security and Privacy-Preserving Solutions for IoE

In the evolving IoE ecosystem, security and privacy are becoming increasingly complex. Modern technologies such as Blockchain, Edge AI, and Quantum-Resilient Security offer innovative solutions to the security threats faced by IoE systems.

3.1. Blockchain-Based Security Solutions

Blockchain enables IoE to establish a decentralized and secure data-sharing infrastructure. While traditional security solutions are managed by central authorities, blockchain secures data through its Distributed Ledger Technology (DLT) structure [71].

3.1.1. Blockchain-Based Data Storage and Security Models

IoE systems generate large amounts of data, and it is crucial to store this data securely. Traditional data storage methods rely on centralized cloud-based solutions, which are vulnerable to SPOF risks. Blockchain can be integrated with decentralized file systems such as the InterPlanetary File System (IPFS) to enhance data integrity and security [72].

Blockchain-based data storage models [73]:

• On-Chain Storage: Small-scale IoE data is stored directly on the blockchain.
• Off-Chain Storage: For large-scale IoE data, integration with IPFS and decentralized databases is applied.
• Hybrid Storage: A mixed model where frequently accessed data is stored on-chain, while large-volume data is stored off-chain.

These storage models and their respective advantages and disadvantages are summarized in

Table 4. Blockchain-based data storage models and their characteristics.

Storage Model	Advantages	Disadvantages
On-Chain Storage	Data is immutable and secure	High transaction costs and scalability issues
Off-Chain Storage	Lower cost, support for large data volumes	Dependency on external systems
Hybrid Storage	Balance between efficiency and security	Management complexity

.

3.1.2. Blockchain for Authentication in IoE: DID and SSI Methods

Identity management is a critical element in ensuring system security within the IoE ecosystem. Blockchain-based authentication solutions are generally built upon two fundamental decentralized models known as DID and SSI. These models enable the management of user and device identities without relying on centralized authorities. DID allows IoE devices to create their own digital identities on the blockchain [74], while SSI enables these identities to be managed through digital wallets controlled by individuals or devices [75].

As illustrated in Figure 3, blockchain-based identity authentication solutions allow users to verify their identities without requiring a central authority. In this structure, both the user and the validating device interact through their own SSI wallets, managing their respective digital identity information. This provides a bidirectional and secure authentication process.

These mechanisms not only ensure security during the authentication process but also provide privacy-preserving solutions.

3.1.3. ZKP Approach for Privacy Protection in IoE-Based Systems

In IoE systems, billions of devices, users, and services continuously generate data, much of which is sensitive in nature. Personal identities, health data, and location information are processed within these systems, making data protection critical for both individual privacy and system security.

At this point, ZKP mechanisms offer a solution by enabling secure authentication in authorization processes without disclosing sensitive information [76]. ZKP allows a user (prover) to prove to another party (verifier) that they possess certain information without revealing the actual content of that information [77,78]. For example, a user can prove that they have access rights to a system without disclosing personal details such as age, health status, or passwords. As shown in Figure 4, this mechanism ensures that identity remains confidential while still allowing the verifier to complete the authorization process.

ZKP protocols are based on three fundamental properties:

• Completeness: If the stated assertion is correct, the verifier always accepts the proof. Mathematically, this can be expressed as: ∀x ∈ L, if P(x) is honest ⇒ V(x) = accept, can be expressed as.
• Soundness: For an incorrect assertion, the probability of a malicious prover convincing the verifier is negligible. This is expressed as: ∀x ∉ L, Probability[V(x) = accept] < ε.
• Zero-Knowledge: The verifier is convinced of the correctness of the assertion but gains no additional knowledge beyond this fact [78].

Mathematically, it is expressed as. ∃S: S(x) ≈ View_v(P(x)), where S(x) denotes the output of a simulator, and View_v(P(x)) represents the verifier’s view during the actual proof process.

These three core properties play a vital role in constructing secure and privacy-preserving access control systems in IoE ecosystems. IoE systems are widely deployed in diverse domains such as healthcare, smart cities, industrial automation, and home automation. In these systems, the main functions of ZKP can be summarized as follows:

• User authentication: Users can authenticate themselves without revealing biometric data.
• Proof of data ownership: For example, a patient can prove possession of a medical document without sharing its content.
• Access control: In smart building systems, users can prove they have access rights to certain areas.

Different types of ZKP protocols are used in practice, varying in terms of performance, security, and integration aspects [79]. A comparative analysis of common ZKP protocols based on core performance criteria for IoE requirements is presented in

Table 5. Key technical features of common ZKP protocols and their suitability for IoE applications.

Protocol	Proof Size (Bytes)	Proving Time (ms)	Verification Time (ms)	Trusted Setup	Quantum Resilient	IoE Suitability
zk-SNARK	192	1299	1138	Required	No	High: efficient, small size
zk-STARK	6657	552	52	Not required	Yes	Medium-High: quantum secure, large size
Bulletproofs	737	6756	899	Not required	No	Medium: suitable for lightweight systems
Ligero	~1024+	Medium	Medium	Not required	Yes	Medium: balanced performance

.

zk-SNARK is suitable for embedded systems, smart cards, and mobile devices due to its small size and fast verification times. However, it requires a trusted setup and is not quantum-resistant. zk-STARK offers quantum security, transparency, and high accuracy. It is ideal for high-security IoE applications, such as smart city systems. Bulletproofs is compact, requires no setup, and is fast, making it suitable for lightweight devices, wearable technologies, and smart home systems [9,79,80].

ZKP mechanisms become even more powerful when integrated with the Confidential Transactions (CT) protocol. This structure enables the verification of transaction correctness while keeping details like amounts and parties confidential. Through the CT-ZKP integration, for example, a user can prove that they have sufficient balance in their wallet without revealing the actual amount, or prove that a condition (such as age >18) is met without disclosing the exact age [81]. In healthcare systems, ZKP protocols are particularly critical for ensuring privacy-preserving verification and sharing of patient data, enabling the secure management of sensitive health records [49].

In conclusion, ZKP mechanisms are one of the most robust methods for ensuring secure access and transaction verification in IoE environments while preserving privacy. The advantages of different protocol types should be evaluated contextually, considering hardware limitations, security requirements, and integration ease. In distributed and high-volume IoE systems, factors such as quantum resistance, verification time, and trusted setup requirements play a critical role [9,79,80].

3.1.4. Blockchain-Based Access Control Solutions

In IoE systems, access management is critical to prevent unauthorized devices from connecting to the network. Blockchain-based access control mechanisms can enhance security by integrating with Role-Based Access Control (RBAC) and Attribute-Based Access Control (ABAC) models [82,83]. Blockchain-based access management:

• RBAC: Determines access permissions based on user roles.
• ABAC: Dynamically determines access permissions based on user behavior.

The integration of blockchain-based RBAC/ABAC systems with smart contracts is illustrated in Figure 5.

A user sends an access request to an IoE system. This request is evaluated by the RBAC/ABAC policy engine operating on the blockchain. Based on the policy evaluation, a smart contract decides whether to grant or deny access. If the access criteria are met, the system responds with access granted.

3.1.5. Blockchain-Based Use Cases in IoE Security

Due to its distributed and heterogeneous structure, the IoE ecosystem is exposed to multi-layered security threats. In this context, blockchain technology offers innovative solutions to strengthen IoE security through its decentralized, immutable, and programmable features. In this subsection, the contributions of blockchain to IoE security are addressed under four main categories:

• Decentralized Identity (DID): Enables users and devices to perform identity authentication without relying on central servers.
• Smart Contracts: Allow access policies to be autonomously enforced, reducing the risk of human error.
• Data Integrity and Provenance: Through the chained record structure, IoE data becomes tamper-proof, increasing reliability.
• Consensus Mechanisms (PoS, PBFT, Directed Acyclic Graph (DAG)): Improve the scalability and transaction efficiency of blockchain, enabling real-time secure transactions in IoE systems.

The corresponding use cases for these four contribution areas are summarized in

Table 6. Blockchain security applications in IoE/IoT.

Blockchain Feature	Use Case in IoE	Security Enhancement
Decentralized Identity [74,84,85]	User authentication and IoT device identity verification	Prevents unauthorized access and identity spoofing
Smart Contracts [86,87]	Automated access control for IoE services	Reduces human errors and enhances trust
Data Integrity and Provenance [88,89]	Secure healthcare records and industrial IoT logs	Prevents data tampering and enhances regulatory compliance
Consensus Mechanisms (PoS, PBFT, DAG-based) [90,91]	Scalable blockchain for real-time IoE transactions	Lowers energy consumption and improves efficiency

.

3.1.6. Performance of Blockchain Systems

Blockchain platforms used in IoE systems must be evaluated not only in terms of security but also regarding performance and energy efficiency [92,93]. This is because massive simultaneous data transmission, latency sensitivity, and limited resources (CPU, battery, bandwidth) are critical in IoE networks. Therefore, different blockchain infrastructures should be compared using key metrics such as transaction speed (TPS), latency, energy consumption, data privacy, scalability, and access control, and the most appropriate system should be selected according to specific IoE requirements [92,93,94].

In this context,

Table 7. Comparison of blockchain platforms in terms of latency, transaction speed, and energy consumption.

Criteria and Features	Ethereum (PoW → PoS)	Hyperledger Fabric	IOTA (Tangle-Based)
Governance	Community-driven, decentralized	Consortium-based (e.g., Linux Foundation)	Foundation-based (IOTA Foundation), protocol roadmap
Participation Type	Permissionless	Permissioned	Permissionless (but Coordinator for security)
Use Cases	DeFi, NFTs, dApps, public registry	Supply chain, healthcare, enterprise apps	IoT, industrial networks, M2M payments
Consensus Mechanism	PoS (formerly PoW), Nakamoto-style	PBFT, Raft, Kafka (modular)	DAG-based Tangle + Tip Selection (no mining, no blocks)
Smart Contract Language	Solidity/EVM	Chaincode (Go, Java, JS)	Not native (external logic), Smart Contract via ISCP
TPS	~15 (PoW), ~30–100 (PoS)	~2000	~1000+ (asynchronous, parallel structure)
Latency (ms)	~6500	~250	~300–400
Energy (kWh/1000 txns)	~70 (PoW), ~5–10 (PoS)	~0.5	~0.1–0.3 (very low, no mining)
Data Privacy	Low (transparent ledger, pseudonymity)	High (private channels, access control)	Medium (public DAG, off-chain encryption possible)
Security Risk	High (e.g., 51% attacks in PoW)	Low (controlled node participation)	Medium (central Coordinator used for finality)
Scalability	Limited	High (modular and parallelizable architecture)	High (parallel tip selection, no block bottleneck)
Energy Efficiency	Low (PoW), improving with PoS	High (lightweight consensus)	Very high (no mining, lightweight protocol)

presents a comparative analysis to clearly demonstrate how different blockchain platforms are positioned in terms of performance and security for IoE integration.

IoE systems consist of billions of interconnected devices, generating massive amounts of data where low-latency communication plays a critical role. Therefore, blockchain platforms used in IoE must be evaluated not only for their security features but also for their performance criteria such as transaction speed, energy efficiency, data privacy, and flexibility [93,95]. Table 7 provides a comparative presentation of the technical capabilities of various blockchain infrastructures against these multidimensional requirements.

Permissionless blockchains such as Ethereum offer high levels of decentralization and cryptographic security. However, during the Proof-of-Work (PoW) phase, they demonstrated high energy consumption and latency, making them less suitable for real-time and resource-constrained IoE scenarios. Although improvements in transaction capacity and energy efficiency have been achieved with the transition to Proof-of-Stake (PoS), challenges still remain in critical areas such as data privacy and access control due to the open-access nature of permissionless systems [95].

On the other hand, Hyperledger Fabric, with its permissioned structure, offers a controlled participant network and modular architecture, providing high transaction throughput, low latency, and advanced access control mechanisms. These features make it an extremely suitable alternative for enterprise-level IoE applications. The platform’s channel-based data isolation, flexible smart contract development, and low energy consumption support operational sustainability, making Hyperledger Fabric a prominent choice in IoE contexts [92,96,97]. DAG-based platforms such as IOTA differ from traditional blockchain architectures by offering micro-transaction capability, zero transaction fees, and low latency, directly addressing the hardware limitations of IoT/IoE devices. However, in terms of access control, governance models, and enterprise-grade privacy requirements, current implementations of DAG systems are still limited. Additionally, the maturity level of decentralized security mechanisms in these systems lags behind permissioned platforms [94,95,98].

Overall, permissioned blockchain infrastructures provide more holistic and balanced solutions to the key requirements of IoE systems, such as security, transaction capacity, and energy efficiency. The DAG-based structure of IOTA eliminates the need for blocks and miners, enabling a parallel transaction confirmation architecture where transactions validate each other directly. This offers significant advantages such as low latency, high transaction throughput, and near-zero transaction costs. However, it also shows limitations in access control, data privacy, and enterprise governance. In contrast, Hyperledger Fabric, with its permissioned structure, offers strong access management, a customizable modular architecture, and high scalability, making it particularly suitable for real-time IoE applications where security and privacy are critical.

Therefore, building the system architecture on permissioned platforms such as Hyperledger Fabric emerges as a more rational choice in terms of both technical capability and operational sustainability [92,93,94,95,96,97].

3.2. Edge AI and Secure Computing in IoE

In traditional IoE systems, data is typically processed and analyzed in the cloud environment [99]. However, this approach involves disadvantages such as high latency, network bandwidth requirements, and privacy risks [100,101,102,103]. Edge AI solutions overcome these issues by processing data closer to the devices, providing significant improvements in security, privacy, and real-time decision-making processes [104,105].

3.2.1. The Role of Edge AI in Ensuring Security in IoE

Edge AI offers various advantages, especially in critical applications such as healthcare, smart cities, and industrial systems [106]:

• Reduces latency and enables rapid response [100,101].
• Enhances data privacy, since data is processed locally without being sent to a central server [102,103].
• Detects anomalies in real-time, allowing immediate responses to attacks or unexpected behaviors [106].

As illustrated in Figure 6, the Edge AI model analyzes data locally from IoE devices, performs anomaly detection, and can instantly make security decisions against potential threats [107,108].

3.2.2. Privacy-Preserving Data Processing in IoE with Edge Computing and AI

Edge computing eliminates the need for centralized data processing in IoE systems by enabling data to be processed close to the source [15,16]. This approach not only reduces latency but also allows local data processing before sending data to the cloud, thereby enhancing privacy [103,106].

Edge-based data processing is especially critical in scenarios requiring high levels of privacy, such as the management of personal health data or sensitive information from smart home devices [36,42,87,88].

Key techniques for privacy-preserving data processing include:

• Federated Learning (FL): Model weights are shared, but data is not. User data remains on the edge device [23,25,65].
• Differential Privacy (DP): At the end of the modeling process, data is statistically obfuscated to prevent the re-identification of individual users [109,110].
• Secure Multi-Party Computation (SMPC) and Homomorphic Encryption (HE): Multiple devices collaborate in joint analysis without accessing each other’s data [46,56,104].

Through these approaches, AI processing on the edge becomes both secure and efficient [107,108]. Especially in IoE systems, distributed AI algorithms with a privacy-preserving focus offer great potential for applications in healthcare, transportation, and smart cities [26,31,87].

3.2.3. Federated Learning-Based Edge AI Solutions

FL allows edge devices such as IoE sensors, mobile devices, and medical equipment to perform local model training on their own data [23,25,65]. This method enables a collaborative learning process by sharing only model updates rather than raw data, thus eliminating the need to send personal data to central servers. As a result, data privacy is preserved, network traffic is reduced, and security risks are minimized [16,104].

FL is particularly preferred in domains where privacy is critical, such as healthcare, finance, and industrial IoT [23,25,65,109]. Each device updates its own model locally with its private data and sends the model updates to the central server. However, raw data is never shared. This ensures compliance with data protection regulations such as the General Data Protection Regulation (GDPR) and supports a secure learning process [46,56,104,109].

As illustrated in Figure 7, edge devices train their local models and send only the model updates to the FL server. This central server then provides a global model to the IoE network, ensuring that an up-to-date and secure AI model is deployed system-wide.

In the FL architecture, each IoE device performs local model training on its own data [25,65]. During this process, the data never leaves the device; only model parameters, such as weights or gradients, are sent to a centralized or distributed aggregator. This ensures data privacy while simultaneously enabling collective and collaborative learning at a global level [109].

Table 8. Comparison between traditional AI and federated learning.

Criterion	Traditional AI	Federated Learning
Data Privacy	Centralized	Local Processing
Latency	High	Low
Security Risk	High	Low

illustrates how FL differs from traditional AI models in terms of key criteria such as privacy, latency, and security risk [23,109,110].

Providing high-accuracy learning without accessing centralized data is one of the fundamental pillars of this study. However, the success of FL should not be evaluated solely based on privacy; metrics such as communication overhead, convergence speed, and latency must also be considered. In this context,

Table 9. Comparative FL model performance and communication latency.

Model	Accuracy (%)	Comm. Overhead (MB)	Convergence Rounds	Latency (ms)
FedAvg	88.7	3.2	120	850
FedProx	90.1	3.4	110	910
Quantum FL (QFL)	91.8	2.1	100	760

presents a quantitative comparison of different FL approaches.

Table 9 clearly demonstrates the performance differences between the traditional FedAvg method and more advanced FL techniques [109,110]. For instance, the FedProx model achieves convergence in fewer rounds due to its resilience to network heterogeneity. Notably, the Quantum-assisted FL (QFL) model achieves higher accuracy with lower communication overhead and reduced latency. This comparison highlights that algorithmic choices in FL architectures directly impact both efficiency and energy consumption.

3.2.4. Edge AI Security Mechanisms for IoE

In IoE environments, various AI-based security mechanisms are directly implemented at the edge layer to enhance security and privacy [15,16,108]. These methods enable real-time decision-making, significantly reducing cloud dependency and minimizing latency in data processing and threat detection [100,104,105].

Table 10. To enhance security, Edge AI employs the following mechanisms.

Mechanism	Function	Security Benefits	References
AI-Driven Intrusion Detection (Edge IDS/IPS)	Detects and mitigates real-time cyber threats at IoE edge nodes	Faster threat response, reduces cloud dependency	[25,108]
FL on Edge Devices	Local AI model training on edge devices without sending data to centralized servers	Prevents data exposure, supports privacy compliance	[23,65,109]
Adversarial AI Defense at the Edge	AI models that detect and counter adversarial attacks on IoT networks	Enhances robustness against AI-based cyber threats	[29,30,57,64]
Threat Intelligence and Anomaly Detection	AI continuously monitors edge devices for abnormal behavior	Prevents zero-day attacks, malware propagation	[26,31,59]

summarizes the core Edge AI security mechanisms that contribute to building trusted and resilient IoE ecosystems.

The mechanisms listed in Table 10 demonstrate how AI is leveraged at the edge level to counter modern cybersecurity threats in IoE. Each method serves a unique function: AI-based IDS/IPS systems provide real-time threat detection and rapid response, while FL protects data privacy by performing model training locally. Adversarial AI defense mechanisms increase model robustness against manipulated inputs, and anomaly detection enables continuous monitoring to identify unknown or zero-day attacks. When these approaches are evaluated collectively, they provide a scalable foundation for creating secure and intelligent edge computing frameworks, reducing reliance on centralized infrastructures.

3.2.5. AI Performance Metrics of Edge Computing Devices

The hardware capacity of devices used in edge computing architectures plays a critical role in executing AI operations effectively [15,106,108]. In this context, factors such as real-time data processing capabilities, memory capacity, the presence of AI accelerators, and energy efficiency must be considered [100,104,105].

Table 11. AI processing capacity and performance metrics of edge and cloud devices [111,112,113,114,115,116,117,118].

Device Name	AI Accelerator        (TIPU/TPU/GPU)	RAM	AI Performance (TOPS)	Energy        Consumption	Use Case
Raspberry Pi 4	None/External (USB NPU)	4–8 GB	<0.1 TOPS	Low (~5 W)	Education, Lightweight IoT Projects
Arduino Portenta H7	None/Microcontroller-based	0.5 MB	None	Very Low	Sensor Data Collection, Prototyping
Google Coral Dev Board	Edge TPU	1 GB	4 TOPS	Low (~2.5 W)	Image Processing, Smart Sensors
Intel Neural Compute Stick 2	Intel Movidius Myriad X	External USB	1 TOPS	Very Low	Prototyping, Video Analysis
BeagleBone AI-64	Embedded DSP + Vision Engine	4 GB	~1 TOPS	Medium	Robotics, Automation
Jetson Xavier NX	NVIDIA Volta GPU + 384 CUDA Cores	8–16 GB	21 TOPS	Medium (~15 W)	Edge AI, Autonomous Systems
Jetson AGX Orin	Ampere GPU + 2048 CUDA + 64 Tensor Cores	32–64 GB	275 TOPS	High (~50 W)	Autonomous Vehicles, Drones
Huawei Atlas 200 DK	Ascend 310 NPU	8 GB	16 TOPS	Medium	Industrial Visual Analytics
Cloud (AWS EC2 p4d)	NVIDIA A100 GPU	1 TB+	1000+ TOPS	Very High	Large-Scale Model Training, Analytics

presents a comparative overview of commonly used edge devices, evaluating them based on AI support, use cases, and performance metrics. For the effective implementation of AI-based security and processing mechanisms, the hardware capacity and processing performance of the devices used in edge and cloud infrastructures are of paramount importance [16,106,108]. Due to the requirements of real-time data processing, low latency, and on-device decision-making, Edge AI architectures are particularly sensitive to hardware selection [100,105,107]. To technically evaluate the feasibility of the proposed system, Table 11 provides a comparative analysis of the AI-enabled performance metrics of commonly used edge and cloud devices.

Table 11 presents the key hardware features and performance metrics of various edge and cloud devices used in AI applications. Devices are evaluated based on their AI accelerator types (GPU, TPU, NPU), memory capacities (RAM), AI processing power (TOPS), energy consumption levels, and use case scenarios. These metrics are particularly critical in edge architectures, where applicability, energy efficiency, and task compatibility must be carefully assessed. For example, high-performance devices like the NVIDIA Jetson AGX Orin are suitable for advanced visual analytics and autonomous decision systems, while lower-power devices like the Raspberry Pi 4 or Arduino Portenta H7 are more appropriate for education, prototyping, or lightweight IoT scenarios. On the other hand, cloud-based solutions such as AWS EC2 offer very high AI processing capacity, but may present limitations in real-time performance and energy consumption compared to edge solutions.

3.3. Quantum-Resilient Security Approaches

With the advancement of quantum computing, existing encryption methods are under threat, making the development of quantum-resilient security solutions mandatory for IoE systems.

3.3.1. Post-Quantum Cryptographic Methods

PQC methods encompass a new generation of algorithms designed to be resistant to quantum algorithms that threaten classical cryptography [33,34,35]. Notable examples include lattice-based encryption, hash-based signature algorithms, multivariate polynomial cryptography, and code-based encryption [52,53,54,55,56].

Figure 8 illustrates how PQC is integrated into IoE systems through a layered architecture. The objective is to provide protection against quantum attacks across all components, from the hardware level to the application layer.

In the IoE, post-quantum security must be ensured through a multi-layered architecture, spanning from hardware to the application layer [18,20,39,70]. The following structure explains how quantum-resistant cryptographic components are integrated into IoE systems across four fundamental layers:

• Hardware Layer: This bottom layer involves physical components such as cryptographic accelerators and secure hardware modules. These devices enable post-quantum algorithms in the upper layers to operate securely and efficiently [70].
• PQC Algorithms Layer: Running on hardware, this layer includes lattice-based, hash-based, and code-based algorithms. These are designed to be resilient against quantum computers and form the cryptographic foundation of the system [33,34,35,52,53].
• Protocols Layer: At this level, the aforementioned algorithms are integrated into communication and security protocols. Key exchange, public key encryption, and digital signature protocols are made quantum-resistant at this stage [33,39,49].
• Applications Layer: At the top, this security structure is deployed within IoE applications. Smart cities, cyber-physical systems, autonomous vehicles, and large-scale IoT solutions benefit from the security provided by this layered approach [20,49,69,88].

Each of the methods used in these layers offers specific advantages for different IoE scenarios, but they vary in terms of performance, key size, processing time, and hardware requirements. Therefore, it is crucial for system designers to select the most suitable algorithm according to the required security level and resource constraints.

Table 12. Feature and application analysis of prominent PQC methods.

Method	Strengths	Weaknesses	Application Areas	References
Lattice-Based Encryption	High security level, post-quantum resistance, versatile usage	Large key sizes, high computational load	Blockchain data encryption, IoT security, digital signatures	[33,34,35,53]
Hash-Based Signatures	Low computational cost, long-term security, and quantum resistance	Limited use per key, large signature sizes	Blockchain protocols, digital signatures, and authentication	[43,49,54,55]
Code-Based Cryptography	Long-term security, low computational cost, and quantum resistance	Large key sizes, impractical in some scenarios	Authentication, IoT data security, email encryption	[119,120]
Multivariate Polynomial Cryptography	Small key sizes, fast processing time, and quantum resistance	Limited research and applications, potential weaknesses	Mobile device security, digital signatures, IoT devices	[56,121]
Quantum Key Distribution	Theoretically, absolute security, full protection against quantum attacks	High cost, requires special hardware, scalability issues	High-security data transmission, finance and healthcare, military apps	[18,51]

presents a comparative analysis of prominent quantum-resistant cryptographic methods, outlining their technical features and evaluating them based on application areas.

3.3.2. Quantum-Resilient Blockchain Protocols: Post-Quantum Digital Signatures

With the advancement of quantum computing, classical digital signatures such as RSA and ECDSA are becoming vulnerable to attacks. This development threatens the core security layers of blockchain systems [18,54,66,67,68,69,70]. Therefore, quantum-resistant signature algorithms are critically important, especially for blockchain-based healthcare systems and IoE infrastructures [19,49,69,88].

Prominent Post-Quantum Digital Signature methods include:

• Lattice-Based Signatures (e.g., CRYSTALS-Dilithium): Proposed in the NIST PQC standardization process, these structures are resilient against quantum attacks [33,52,122].
• Hash-Based Signatures (e.g., XMSS, SPHINCS+): Based on Merkle trees, these methods provide forward security [123,124].
• Multivariate Polynomial Signatures (e.g., Rainbow): Considered among alternative structures offering quantum resilience [125].
• Code-Based Signatures: Built on classical error-correcting codes, providing high resistance to quantum attacks [126].

When these protocols are integrated into the blockchain layer, both data integrity and digital identity authentication are secured against quantum threats. This is particularly significant in medical data storage systems using blockchain, where the protection of sensitive information is paramount [127,128].

3.3.3. Security Optimization in IoE with Quantum AI

Quantum AI offers significantly greater computational capabilities compared to traditional AI methods, enabling more precise security analyses [17,62,66]. In IoE systems, Quantum AI facilitates anomaly detection, attack vector analysis, and dynamic optimization of security policies [26,59,64,110]. Figure 9 presents the general structure of a Quantum AI-supported security framework. Data collected from IoE devices is encrypted using quantum security measures, such as PQC and QKD, and transmitted to the Quantum AI Engine.

This engine operates through a three-stage process chain: Threat Prediction, Anomaly Clustering, and Decision Recommendation Development. This structure enables end-to-end security enhancement and increases the effectiveness of real-time defense mechanisms in IoE systems [129,130,131].

3.3.4. Security and Performance Comparison of Post-Quantum Encryption Algorithms

The advancement of quantum computing demonstrates that traditional encryption methods will gradually become inadequate [18]. In this context, evaluating quantum-resistant algorithms is essential to ensure the sustainability of system security [50,54].

Table 13. Security and performance comparison of post-quantum encryption algorithms.

Algorithm	Key Size (Bytes)	Ciphertext (Bytes)	Encryption Time	Decryption Time	Security Level	References
Kyber1024	1568	1568	1.3 ms	1.1 ms	Level 5	[34,50,132]
Dilithium III	1952	3293	2.4 ms	2.2 ms	Level 3	[54,122,132]
BIKE-3	3136	3100	3.0 ms	2.9 ms	Level 5	[33,34,35,132]

presents a comparative analysis of different post-quantum algorithms.

The algorithms in Table 12 represent advanced post-quantum algorithms recommended by NIST. Kyber1024 stands out with its high security level while maintaining low encryption/decryption times. Dilithium III is suitable for digital signatures, whereas BIKE-3, despite its large key size, offers balanced performance. This comparison provides a concrete reference for achieving balance and efficiency criteria in the algorithm selection process for the proposed system.

3.4. Real-World Problems and Technology Mapping

While emerging technologies such as AI, IoE, Blockchain, Edge Computing, and Quantum-Resilient Cryptography promise transformative impacts, their integration is often driven by pressing real-world challenges [20,85,106,132]. These challenges span across sectors including healthcare, smart cities, supply chains, and personal data ecosystems [9,127,133,134]. A set of concrete real-world problems is presented in

Table 14. Mapping of real-world problems to suitable enabling technologies for effective and sustainable solutions.

Real-World Problem	Sector	Risk/Impact	Suitable Technology Stack	References
Unauthorized access to EHRs in hospital networks	Healthcare	Data breach, misdiagnosis	Blockchain + ABAC + AI	[49,69,88,127,128,129]
Fake IoE devices injecting false data in smart homes	Smart Home	Security breach, manipulation	IoE + Blockchain + DID	[74,84,85]
Counterfeit medicines in pharmaceutical supply chains	Supply Chain	Health risk, loss of trust	Blockchain + RFID + Traceability	[9,48,97]
Sensitive data leakage due to cloud-only processing	Healthcare/IoT	Privacy loss, regulatory violation	Edge AI + FL + Differential Privacy	[23,87,133,134]
Delayed anomaly detection in traffic systems	Smart Cities	Accidents, congestion	Edge AI + IoE + Real-time ML	[31,59,108,135,136]
Incompatibility in data sharing across cross-domain systems	Healthcare	Fragmented data, incomplete medical records	Interoperable Blockchain + Semantic AI	[82,83,84,137,138]
Vulnerability of digital signatures to quantum attacks in long-term records	Legal/Medical	Future forgery risk	Quantum-Resilient Signatures (Lattice, Hash-Based)	[33,35,54,122,127]
Impersonation attacks in IoE infrastructure	Critical Infrastructure	Unauthorized command execution	Blockchain + AI + IoE Trust Layer	[7,20,130,139]
Data tampering in emergency disaster response coordination	Disaster Management	Wrong routing, loss of aid	Blockchain + Edge + Secure Access Control	[86,97,102,140,141]

, each mapped to the corresponding enabling technologies that are most suitable, either individually or in combination, for delivering effective and sustainable solutions.

As shown in Table 14, there is no single technology that addresses all aspects of real-world complexity. Instead, hybrid frameworks that combine Blockchain’s immutability, AI’s adaptability, Edge’s low-latency processing, and post-quantum security measures offer the most promising architecture. These mappings form the basis for future implementations in secure, scalable, and ethically aligned digital infrastructures.

3.5. Integration of Different Security Layers

In the IoE ecosystem, security has become too multidimensional to be ensured by individual solutions alone. Therefore, the recent literature shows an increasing trend toward solutions that employ multiple security layers simultaneously. Technologies such as AI, Blockchain, Edge Computing, and Quantum-Resilient techniques are increasingly positioned as complementary tools across various application domains. These technologies provide more comprehensive and effective security solutions not only individually but also when used in combination [20,85,87,105].

3.5.1. Security Layer Combinations According to IoE Use Cases

Table 15. Security layer combinations according to IoE use cases.

Application Area/Sector	Security Layers Used	Example Technologies	Problems Addressed	Justification	References
Healthcare Systems	Privacy-Preserving Computation, Blockchain, AI, Quantum Resilient	FL, DP, Smart Contracts, FL, PQC	Data privacy, traceability, anomaly detection, future-proof security	Comprehensive security is required; patient data is both private and critical	[49,87,88,127,128,134]
Smart City Traffic Management	Privacy-Preserving Computation, Blockchain, AI	Edge AI, FL, Smart Contracts, IDS, SSI	Real-time decision-making, attack detection, traffic optimization	Real-time intervention and data security are both necessary	[31,142,143,144]
Smart Home and IoT-Based Security	Privacy-Preserving Computation, AI	DP, IDS/IPS, FL	Preventing unauthorized access, privacy protection, local threat detection	Resource-constrained devices require fast and local security	[145,146]
Supply Chain and Logistics	Blockchain, AI, Quantum Resilient	Blockchain, DID, QKD, Quantum Signatures	Counterfeit prevention, traceability, long-term integrity	Secure verification across the entire chain is mandatory	[9,48,80,97,122,147]
Data Archiving and Legal Records	Blockchain, Quantum Resilient	Blockchain, Hash-based PQC, SSI	Document integrity, quantum-resistant signatures	Long-term data storage requires quantum security	[148,149]
Disaster Management, Emergency Response	Blockchain, AI	Blockchain Ledger, IDS, Semantic AI	Secure access during critical moments, prevention of data manipulation	Multi-stakeholder systems require auditability and security	[140,141,150,151]

presents selected examples where these technologies are used together across various application domains, showing which security layers target which problems in each scenario and which technological tools support them. This table serves as a structural foundation that justifies the integrated layered security architecture proposed in the subsequent sections of this study.

3.5.2. Multi-Layered Security Architectures

In the evolving IoE ecosystem, singular security solutions are no longer sufficient, and there is a growing need for holistic and integrated architectures to address complex and multi-layered security threats [1,2,3,10,14]. In this context, when blockchain-based solutions, Edge AI-based anomaly detection systems, and quantum-resilient cryptographic approaches, which were discussed separately in the previous sections, are considered together, they present more robust and sustainable security strategies [25,33,54,108].

Table 16. Contributions of edge AI, blockchain, and quantum-resilient methods to IoE security.

Approach	Features/Definition	Advantages	Disadvantages	Application Areas
Blockchain-Based Security [11,38,47,83,84,85,88,92,93]	Access control via smart contracts, Decentralized identity authentication, Immutable data storage	No single point of failure, Transparency and auditability, Increased trust	High computational cost, Scalability issues, Regulatory barriers	IoE identity management, Data integrity, Supply chain security
Edge AI for Secure Computing [15,16,25,26,27,28,59,106,112,113]	Real-time anomaly detection, Privacy-preserving AI model training, Low-latency security decisions	Fast decision-making, Reduces network congestion, Less dependency on central servers	Limited hardware capacity, AI model security risks, Adversarial attacks	Smart cities, IoT security monitoring, Cyber-attack detection
Quantum-Resilient Security [33,34,35,52,53,54,55,66,67,68,69,70,120,127,132]	PQC, hash-based signatures, ZKP, and Quantum AI for long-term data security	Future-proof security, Resistance to quantum attacks, Stronger cryptographic structure	Lack of widespread adoption, Complex implementation processes, High computational cost	Healthcare data security, Financial transactions, IoE data transmission

compares the three main approaches proposed to address the security and privacy challenges of IoE: Blockchain-Based Security, Edge AI-Enabled Security, and Quantum-Resilient Security Solutions.

• Blockchain-Based Security provides security for IoE in critical areas such as identity authentication, access control, and data storage through its decentralized structure [9,11,38,83]. However, it faces challenges such as scalability issues and high transaction costs [88,92,93].
• Edge AI for Secure Computing aims to perform real-time attack detection and security decisions using local AI models while preserving data privacy [16,25,59]. Nevertheless, it has limitations including adversarial attacks and restricted hardware capacity [28,137,146].
• Quantum-Resilient Security offers more robust solutions against future threats through PQC and quantum-enhanced AI methods [33,34,35,54,127]. However, it also presents challenges such as complex implementation processes and high computational costs [52,53,130].

In conclusion, Blockchain, Edge AI, and Quantum Security solutions provide different advantages and limitations across IoE security layers. Blockchain creates a robust security framework with its decentralized structure, while Edge AI enhances security through real-time data processing capabilities. Quantum Security plays a critical complementary role in ensuring long-term resilience. It is clear that IoE security cannot be achieved through a single solution alone. Therefore, adopting hybrid approaches based on layered architectures is essential for establishing sustainable security strategies.

3.5.3. Proposed Integrated Layered Architecture

In modern systems, the proliferation of IoE devices has made it possible to monitor and process patient data in real time. However, this situation also brings multi-layered threats, including data security, privacy, access control, and system reliability. In such a complex environment, ensuring security cannot be accomplished through a single technology. Instead, a task-oriented integration of various approaches, including AI, Edge Computing, Blockchain, and Quantum-Resilient Cryptography, is required. In this context, Figure 10 illustrates a layered and integrated security architecture that covers the entire process in IoE-based systems, from data generation to threat analysis. This structure not only represents an academic model but also offers a multi-component solution framework for real-world applications.

This architecture offers a multi-layered solution to the security and privacy requirements emerging in IoE-based systems. From the moment data is generated, the processes of semantic interpretation, privacy-preserving processing, immutable storage, and future-oriented analysis are performed through sequential but complementary modules. Each layer not only fulfills its own function but also enables secure and meaningful data transfer to the upper layer, thereby building the holistic security of the system. As shown in Figure 10, the modules operate in an integrated manner, covering AI-based data analysis, edge-level privacy preservation, access-controlled blockchain records, and post-quantum-ready analytics processes. Below, the function of each layer, the methods used, and their role in the system are explained in detail.

IoE Devices: This layer includes sensors, ECG devices, blood pressure monitors, wearable devices, and mobile health applications located in hospital environments. Raw patient data is generated at this point. This data is not directly suitable for processing and is vulnerable to security threats [13,36,42,88].

AI/ML-Based Detection Layer: This is the first intelligent layer that analyzes raw data from IoE devices. Machine learning algorithms such as CNN, LSTM, Decision Trees, and Autoencoders are used for anomaly detection and adversarial attack detection. The system tags risky situations with risk labels and transmits the filtered data to the next layer [26,31,32,59].

Edge Privacy-Preserving Computation: This layer processes the labeled data using privacy-preserving methods. The techniques used include:

• FL (Federated Learning): Model training is performed locally on edge devices, and data is not transferred to a central server [116,137,145,146].
• HE (Homomorphic Encryption): Encrypted data is processed directly [46,104,105].
• DP (Differential Privacy): Individual data protection is added to cumulative data patterns [109,110].

The resulting outputs are transmitted as encrypted/private outputs to the Blockchain layer. If the system operates with FL, model updates are also sent from this layer.

The Quantum AI layer in the proposed architecture is conceptualized as a forward-looking module. While real-world quantum AI systems are still in experimental phases, this layer represents a visionary approach for future threat intelligence and security pattern analysis in IoE environments.

Blockchain (Smart Contracts + ABAC): The privacy-protected outputs are logged in this layer. Smart Contracts evaluate access requests using the ABAC model. Each transaction is immutably recorded, access is validated, and access control is enforced. Thus, only authorized individuals can access the data, and all transactions are securely stored in accordance with the principle of auditability [82,83,92,97].

Quantum-Resilient Cryptography: This layer operates as if beneath the blockchain layer. It provides long-term data security against quantum attacks through lattice-based cryptography [33,34,35], hash-based digital signatures [52,53,54,55], and ZKP [48,49,77,78,79]. This layer does not actively produce data but cryptographically protects incoming data [69,70,120,127].

Quantum AI Analyzer: This is the top analysis layer of the system. It takes Immutable Logs and Access Records from the blockchain layer and analyzes them using Quantum AI algorithms (e.g., Quantum-Enhanced ML, Quantum Support Vector Machine, QAOA) [127,129,130,135,136].

As a result of these analyses, the system produces Threat Predictions, Proactive Alerts, and Pattern Classifications. These outputs enhance the system’s defense capacity and help develop proactive security strategies against future attacks.

The concretization and functionality of the proposed architecture depend on defining algorithmic processes that explain how each technology in each layer works and how they interact with each other. In this context, the algorithms presented below detail the operational logic of the layers defined in the figure and provide a fundamental reference for potential real-world applications of the proposed system. These algorithms are conceptualized as a prototype-level approach, each representing the data flows, decision mechanisms, and security processes in the corresponding layer. By systematically structuring the input-output data and processing steps, these algorithms offer a critical roadmap for the practical implementation of the architecture.

Algorithm 1. AI-Based Pre-Processing of IoE Data: The purpose of this algorithm is to classify the raw data (S_i) coming from IoE devices and to make it meaningful for the next layer.

Algorithm 1.  AI-Based risk labeling for IoE streams

Input:  S_i          //Raw data from the IoE device             M_AI     //Trained AI model Output:  Risk_Label: {Low, Medium, High}, Filtered_Data 1:      Receive data stream S_i from IoE device 2:      X_i ← preprocess(S_i) 3:      threat_features ← ExtractFeatures(X_i) 4:      threat_score ← M_AI(threat_features) 5:      if threat_score ≥ 0.8 then 6:            Risk_Label ← “High” 7:      else if Threat_Score ≥ 0.5 then 8:            Risk_Label ← “Medium” 9:      else 10:          Risk_Label ← “Low” 11:    endif 12:    Filtered_Data ← summarize(X_i, threat_features) 13:    return Risk_Label, Filtered_Data

This algorithm processes the raw data received from IoE devices in real-time, converting it into an analyzable form. Initially, data from sensors is collected and passed through a preprocessing phase. In this phase, operations such as noise filtering, missing value completion, and normalization may be performed. Subsequently, AI-based analysis is conducted. This analysis is typically performed using pre-trained models (M_AI) or edge AI components. Data samples are classified into categories such as risky, sensitive, or ordinary. These labels contain information that shapes the system’s security policy, including data sensitivity levels or potential anomaly risks.

Finally, only the labeled and meaningful data is forwarded to the next layer (edge privacy and FL) for deeper analysis. In this way, the system eliminates unnecessary load while prioritizing security.

Algorithm 2. Federated Learning Based Privacy-Preserving Training: The purpose of this algorithm is to perform local model training on distributed edge devices and update the global model without sharing user data.

This algorithm enables AI models trained locally on edge devices or IoE nodes to learn without sending data to a central server. Initially, edge nodes receive the labeled data. Each device conducts model training on its own data. At the end of the training process, model weights (e.g., weight matrices, learning coefficients) are extracted.

Before these model updates are sent to the blockchain, they pass through a privacy check. If the data label indicates “high confidentiality”, the corresponding model contribution is not shared, ensuring compliance with the privacy policy. The allowed updates are transmitted via the blockchain to be integrated into the global model. During this process, aggregation (such as federated averaging) is performed to construct a new global model.

Each edge node performs FL training on its processed data. Updates are sent not directly to the central model but to the aggregator. If the aggregator is decentralized, the model is updated via consensus on the blockchain. Additionally, quantum-resilient encryption is applied during this transmission.

Algorithm 2.  Federated learning phase on edge devices

Input:  Filtered_Data_i      //Processed data from the i-th edge device             M_local_i            //Local model on the edge node             E                         //Epoch             η                         //Learning rate Output:  Model_Update_i        //Weight update of the i-th device 1:    Initialize M_local_i ← Global_Model 2:    for epoch = 1 to E do 3:            Mini_Batch ← CreateBatches(Filtered_Data_i) 4:            foreach (x, y) ∈ Mini_Batch do 5:                    y_pred ← M_local_i(x) 6:                    loss ← ComputeLoss(y_pred, y)                        //Update of model parameters (weights and biases) 7:                    M_local_i ← M_local_i − η ∇loss 8:            endforeach 9:    endfor 10:  Model_Update_i ← M_local_i − Global_Model 11:  return Model_Update_i    //to aggregator on blockchain or                                                //secure aggregator node

Following Algorithm 2,

Table 17. Layer–Technology function mapping of the proposed FL and security architecture.

Layer	Associated Technology	Function
Edge Node	FL (FedAvg/FedSGD)	Local model training
Blockchain Layer	Smart Contract + Logging	Model updates are hashed and logged
Quantum Resilient Crypto	Lattice/Hash-based Encryption	Secure transmission of FL model updates

summarizes the layered structure of the proposed system by mapping each security layer to the specific technology used and its corresponding function. This representation clarifies how FL, blockchain, and quantum-resilient cryptography interact to ensure secure, decentralized, and privacy-preserving model updates in IoE environments.

Algorithm 3. Blockchain Logging with Quantum-Resilient Encryption: The purpose of this algorithm is to utilize a quantum-resistant blockchain for the secure storage and sharing of FL model updates. This algorithm is designed as a chaincode (smart contract) running on a blockchain infrastructure. It ensures that system events are timestamped, immutable, and that authorization processes are enforced. The algorithm securely integrates the model updates generated during the FL process, which are authorized for sharing, into the blockchain network. In the first step, data such as model parameters and training metrics to be transmitted are encrypted using quantum-resistant encryption algorithms. Unlike classical systems like RSA/ECC, these algorithms provide protection against quantum computer attacks. The encryption schemes are generally based on lattice-based or hash-based cryptography. Subsequently, these encrypted updates are written to the blockchain as a transaction. Each transaction serves both as a historical log and as reference data. Moreover, through the use of hash functions, the integrity and authenticity of the model updates can always be verified. This process ensures not only security but also auditability and traceability, forming a secure and verifiable chain of model development.

Algorithm 3.  Secure global aggregation and blockchain logging

Input:  LocalModelUpdates[]      //Updated model parameters from each IoE device group             PublicKey_QR                //Public key for Quantum-Resilient Encryption           SmartContract_SC          //Access control and logging smart contract                                                      //running on the blockchain Output:  BlockchainRecord        //Secure and immutable record of the updated model               AggregatedModel        //Securely aggregated (global) model 1:    Initialize AggregatedModel ← ZeroModel 2:    foreach LocalModel in LocalModelUpdates[] do 3:            AggregatedModel ← AggregatedModel + Weighted(LocalModel) 4:    endforeach 5:    EncryptedModel ← Encrypt_QR(AggregatedModel, PublicKey_QR) 6:    TX ← FormatTransaction(EncryptedModel, Metadata) 7:    if ValidateSmartContract(SC, TX) == True then 8:            BlockchainRecord ← CommitToBlockchain(TX) 9:    else 10:          Raise Error(“Smart Contract Rejected the Aggregation Record”) 11:  endif 12:  return BlockchainRecord, AggregatedModel

Each local model trained by the devices of each IoE group is aggregated using a weighted average. Then, the aggregated model is encrypted with Quantum-Resilient Encryption (e.g., lattice-based cryptography). After encryption, the data is written to the blockchain in a format compatible with the smart contract. If the input is not verified (e.g., access control fails), the system returns an error. In the final step, both the log record and the new global model are returned as outputs.

Algorithm 4. Quantum AI-based Threat Detection and Alerting: The purpose of this algorithm is to analyze the data received from the blockchain to detect potential threats and trigger an early warning system. This algorithm represents the final analysis layer of the system and is executed by advanced Quantum AI models. The model updates and event data obtained from the blockchain are first decrypted and transformed into meaningful input. Then, this data is compared with historical threat patterns, anomaly scores, and system behavior logs. The Quantum AI module has superior capabilities compared to classical AI algorithms, especially in recognizing high-dimensional and complex data patterns. Using methods such as Quantum Annealing or Quantum Support Vector Machines (Q-SVM), risky behaviors and vulnerabilities are detected much faster. Based on this analysis, the algorithm produces real-time threat predictions and sends proactive alerts to system administrators. These alerts allow security teams to take action minutes or even hours before a potential attack.

Algorithm 4. Quantum AI-Based threat analysis and proactive alerting

Input: BlockchainLog[]                //Encrypted model updates and event logs stored on the blockchain           QuantumModel_QAI      //Pre-trained Quantum AI threat analysis model           Threshold_Threat            //Risk score threshold Output: ThreatPrediction          //List of potential threats with confidence scores               ProactiveAlerts[]        //Early warnings and in-system defense triggers 1:    DecryptedData ← Decrypt_QR(BlockchainLog[], PrivateKey_QR) 2:    StructuredInput ← Preprocess(DecryptedData) 3:    ThreatScores ← QuantumModel_QAI.Predict(StructuredInput) 4:    ThreatPrediction ← [] 5:    ProactiveAlerts ← [] 6:    foreach score in ThreatScores do 7:            if score > Threshold_Threat then 8:                    Append(ThreatPrediction, (EventID, score)) 9:                    alert ← GenerateAlert(EventID, score, “HIGH-RISK”) 10:                  Append(ProactiveAlerts, alert) 11:          else 12:                  continue 13:  endforeach 14:  return ThreatPrediction, ProactiveAlerts

Quantum-resilient encryption is first decrypted (Decrypt_QR), and access is granted only to the authorized analysis engine. The data is then preprocessed for analysis, including log normalization and vectorization. After that, the Quantum AI model scores potential threats, and for threats exceeding the defined threshold, the system generates alerts. These alerts help initiate defensive actions proactively.

4. Open Challenges and Future Research Directions

The increasing complexity and multi-component nature of IoE-based security systems bring forth significant challenges, causing many current solutions to fall short in real-world applications. This section addresses the key difficulties encountered in the application of advanced technologies such as Blockchain, Edge AI, and Quantum-Resilient Security within the IoE context and evaluates future-oriented research directions to address these challenges. Additionally, technical and structural issues emerging from multi-layered security architectures are discussed through interdisciplinary strategies.

4.1. Challenges of Blockchain-Based Security Models in IoE

While blockchain-based security solutions offer benefits such as decentralized identity management, data integrity, and access control in IoE systems, there are critical challenges that hinder their widespread adoption.

Scalability, Security, and Performance Trade-Offs: Blockchain struggles with high computational costs and slow transaction times when processing large-scale IoE datasets, making real-time applications difficult. Current Proof of Work (PoW) and Proof of Stake (PoS) systems are not fully optimized for IoE environments. Therefore, more lightweight and efficient consensus mechanisms need to be developed for IoE applications [11,47,88].

Integration with Decentralized Identity (DID and SSI): As the number of IoE devices grows, identity authentication and authorization processes need to be conducted through decentralized methods. Although DID and SSI solutions are evolving, their practical applicability at the IoE scale remains an open research topic [38,93].

Security of Smart Contracts in IoE: Smart contracts are utilized in IoE to enable automated device-to-device transactions. However, software vulnerabilities, cyberattacks, and auditability issues pose significant security risks. There is a critical need to strengthen smart contracts through formal verification and advanced security auditing mechanisms [84,92].

Open Research Areas:

• Development of lightweight and scalable blockchain consensus mechanisms
• Investigation of decentralized identity management systems for IoE scalability
• Protection of smart contracts against cyberattacks

Blockchain-based solutions provide crucial advantages such as secure data recording and identity verification in IoE systems, but also introduce high latency and scalability limitations. The large volume of data generated by IoE devices can create bottlenecks in classical blockchain architectures. This leads to a growing need to balance security and processing speed.

Research Direction: Future studies should focus on more scalable infrastructures such as Layer-2 solutions [152,153,154,155] and DAG-based blockchain systems [153,156]. Additionally, energy-efficient consensus algorithms need to be adapted for IoE environments [157,158]. As the number of IoE devices continues to rise, centralized authentication is becoming insufficient. Enhancing the applicability of DID and SSI-based decentralized identity management systems in IoE devices will provide a substantial contribution to the field [159,160].

4.2. Future Research on the Impact of Quantum Computing on IoE Security

Quantum computing has the potential to break traditional cryptographic methods by using mathematical approaches such as Shor’s Algorithm and Grover’s Algorithm. Therefore, IoE-based security systems must be reinforced with post-quantum cryptographic solutions to withstand these threats [33,49,54,66].

Resistance of Current Cryptographic Systems to Quantum Attacks: Classical cryptographic algorithms like RSA and ECC can be rapidly compromised in the face of quantum attacks. As a result, research is focusing on lattice-based cryptography, hash-based signatures, and code-based cryptography as quantum-resistant alternatives [52,53,68,132].

Quantum-Enhanced Security Solutions: Quantum technology is not only a threat but also an opportunity to enhance security. For example, QKD can provide secure key exchange in IoE networks, and Quantum Random Number Generators (QRNG) can generate stronger encryption systems [53,54,66,70].

Quantum Computing-Based Security Protocols (Hybrid Security Models): To minimize the impact of quantum computing on IoE systems, existing blockchain and AI-based solutions must be made quantum-resistant. Research should focus on Quantum-Resilient Blockchain Protocols and Quantum AI-supported IoE security mechanisms [52,66,68,127].

Open Research Areas:

• Development of quantum-resistant blockchain models [52,58].
• Implementation of QKD for key management in IoE systems [53,54,66].
• Investigation of quantum computing–AI-supported security mechanisms for IoE applications [127,130].

The potential impacts of quantum computers pose a serious threat to classical encryption algorithms. Since IoE systems are expected to be long-lived infrastructures, they must be prepared for the post-quantum era. However, lattice-based cryptographic systems, while promising, still face challenges related to computational overhead and standardization.

Research Direction: Future studies should focus on integrating hybrid encryption systems, including post-quantum digital signatures and ZKP, into IoE architectures. Additionally, it is crucial to investigate how quantum AI techniques can be employed for advanced cyber threat analysis [132,161,162,163,164].

4.3. AI-Driven Security: Federated Learning and Edge AI Challenges

AI is increasingly employed in IoE systems for tasks such as threat detection, anomaly recognition, and proactive defense mechanisms. However, AI-based security solutions themselves are becoming vulnerable to attacks.

Federated Learning for Privacy-Preserving AI Security: FL enables AI models to be trained without transmitting user data to a central server. The use of FL in IoE devices provides significant advantages, especially in data-sensitive domains such as healthcare, finance, and smart cities [15,25,27,59,106]. Nevertheless, several critical challenges remain:

• Data synchronization and model update frequencies must be optimized for IoE environments.
• Blockchain-based verification mechanisms should be incorporated to ensure the reliability of model sharing.

Real-Time Threat Detection with Edge AI: Edge AI allows IoE devices to analyze data locally without the need for cloud processing. This enables instant responses to security events such as DDoS attacks, data forgery, and malware detection. However, edge devices face limitations in computational power, requiring optimized AI models. Additionally, secure model updates are necessary, where blockchain and FL can play supportive roles [16,28].

AI-Based Attacks and Model Security: AI models in IoE systems are susceptible to adversarial attacks, model poisoning, and data manipulation, which can lead to incorrect decisions or system vulnerabilities. To enhance the security of AI models used in IoE, advanced FL and Edge AI solutions are under development [108,112,113].

Open Research Areas:

• Efficient implementation of FL in IoE devices and Edge AI
• Optimization of Edge AI for real-time IoE security
• Protection of AI models against Adversarial Attacks and Model Poisoning
• Secure sharing of AI model updates using blockchain-supported mechanisms

Model Complexity and Edge Constraints: Edge AI applications enable fast and local analysis in IoE devices. However, these devices often have limited computational power and memory, making it challenging to deploy complex models such as deep learning. Furthermore, continuously updating these models imposes communication and resource burdens on the network and hardware.

Research Direction: Technologies such as TinyML, model quantization, knowledge distillation, and adaptive update strategies are at the forefront of developing AI systems optimized for resource-constrained edge devices [165,166]. Additionally, FL allows data to be processed locally, enhancing privacy protection while reducing dependence on centralized servers [167,168,169,170]. Future research should focus on combining TinyML and FL to develop low-resource yet secure Edge AI solutions. Moreover, blockchain-supported FL frameworks and privacy-preserving update mechanisms should be further investigated to strengthen security in IoE environments.

4.4. Interoperability and Trust Management

IoE systems consist of highly heterogeneous environments where numerous devices from different manufacturers operate together. These systems involve various protocols, standards, and security policies, which raise significant compatibility and trust challenges. In particular, there is a pressing need to develop cross-domain security models that ensure seamless and secure operation across multiple ecosystems [47,88,120,127].

Open Research Areas:

• Development of secure and standardized data sharing protocols between heterogeneous IoE platforms [93,120].
• Integration of blockchain-based provenance systems with federated identity management solutions [84,93,130].
• Implementation of behavior-based dynamic trust scoring mechanisms for cross-domain trust management [93,127].
• Design of lightweight and context-aware access control mechanisms to enhance interoperability [171,172,173,174].

Research Direction: The integration of blockchain-based provenance systems with federated identity management and context-aware trust scoring mechanisms could significantly improve inter-platform trust management. In particular, cross-chain data sharing scenarios can benefit from the combination of identity federation and behavior-driven access policies, enabling the establishment of a robust, interoperable trust chain across diverse IoE systems [171,172,173,174].

4.5. Holistic Future Challenges and Strategic Opportunities

IoE systems operate within complex, multi-layered infrastructures that integrate a wide array of heterogeneous devices, platforms, and protocols. As IoE ecosystems continue to evolve, several critical challenges remain unresolved, particularly in areas such as the integration of decentralized technologies [33,93], quantum-resilient security solutions [38,52,53], and scalable AI models [49,59,108,113]. Addressing these challenges is not only essential for building secure, sustainable, and future-proof systems, but also requires interdisciplinary approaches to develop ethical, inclusive, and trustworthy IoE applications.

Table 18. Future research challenges and potential solutions.

Challenge	Problem Description	Potential Solutions	References
Scalability of Blockchain and AI	High latency in blockchain transactions and FL model training in large-scale IoE networks	Optimized consensus mechanisms, lightweight privacy-preserving AI techniques, DAG and Layer-2 solutions for high-throughput IoE security	[33,49,66,84,92,108]
Quantum Threats to Security	Shor’s and Grover’s algorithms threaten classical encryption and blockchain integrity	Quantum-safe cryptography (lattice-based, hash-based signatures); hybrid quantum-secure AI integration	[38,52,53,54,108]
Privacy-Preserving AI Optimization	FL, Differential Privacy (DP), and Secure Multi-Party Computation (SMPC) require high computational overhead	Hybrid privacy models combining DP, Edge AI, and Homomorphic Encryption	[59,113]
Decentralized Identity in IoE	Lack of standard security protocols across heterogeneous IoE domains	DID and SSI-based authentication models supported by blockchain trust layers	[93,127,130]
Cross-Chain IoE Data Security	Data interoperability and security challenges across different IoE platforms	Cross-chain privacy protocols and SMPC-based inter-blockchain security mechanisms	[175,176,177]

presents a structured overview of these fundamental challenges alongside potential strategic solutions.

The challenges summarized in Table 18 highlight the need for not only technological advances but also holistic strategies that address scalability [49,84], privacy [59,113], interoperability [93,127], and post-quantum security [38,53]. Given the inherent complexity of IoE ecosystems, comprising diverse platforms, device types, and varying levels of data sensitivity, solutions must be hybrid and adaptive rather than singular or static [70,108,120]. In this context, the proposed solutions are designed to both strengthen current systems and enhance resilience against future threats. The mappings provided in the table offer a concrete perspective on aligning technological priorities with research strategies, enabling the development of next-generation IoE security frameworks.

5. Discussions

5.1. Evaluation of Existing Solutions for IoE Security

As the IoE continues to expand, security and privacy threats are becoming increasingly complex. Emerging technologies such as Blockchain, Edge AI [16,25], Quantum-Resilient Cryptography [18,35,54], and FL [23,144,170] have been proposed to address these challenges. However, these solutions are still under development and face various limitations [10,14,17,59]. For example:

• Blockchain offers a decentralized security model but must be adapted to handle IoE’s high-volume data traffic due to scalability constraints and high transaction costs [20,73,152,178,179].
• Edge AI enhances IoE device security through AI-powered threat detection but remains vulnerable to adversarial AI attacks and model poisoning threats [25,27,61,64].
• Quantum-Resilient Cryptography provides protection against quantum computing threats, yet the high computational overhead of PQC algorithms makes them difficult to implement on resource-constrained IoE devices [52,53,54,70,180,181].
• FL allows decentralized AI training while preserving privacy, but poses security risks in model updates and incurs communication overhead [57,144,169,170].

A systematic analysis of these technologies’ advantages and limitations is essential when evaluating current solutions.

Table 19. Advantages and challenges of IoE security technologies.

Security Technology	Advantages	Challenges	References
Blockchain-Based Security	Decentralized security, immutability, transparency	High transaction costs, scalability limitations	[73,152,155]
Edge AI for Security	Real-time threat detection, local data processing	Vulnerable to adversarial AI and model poisoning	[16,29,30,64]
Quantum-Resilient Cryptography	Long-term security, resistant to quantum attacks	High computational demand, limited IoE deployment	[52,70,127,132]
FL for Privacy	Decentralized AI training, preserves user privacy	Security risks in model sharing, communication cost	[144,169,170]

provides a summary of this evaluation.

5.2. Emerging Security Risks with the Expansion of IoE

The rapid scaling of IoE systems has revealed that traditional security approaches are no longer sufficient [14,38,172]. In particular:

• Centralized security models create a SPOF for large-scale IoE networks, significantly weakening overall security [10,140,161].
• Digital identity and authorization issues compromise the reliability of device authentication and access control in IoE environments [40,82,83,84,160].
• Quantum computing-based attacks, while currently theoretical, pose a serious long-term threat to the integrity of IoE security infrastructures [66,67,139].

5.3. Effectiveness and Limitations of Blockchain, AI, and Quantum-Resilient Solutions

An essential research area in IoE security is the integration of Blockchain, AI, and Quantum-Resilient Security to create a more effective defense architecture [20,144,164]. However, each technology presents distinct challenges:

• Blockchain offers decentralized security but remains insufficiently optimized for large-scale IoE systems due to scalability and latency issues [73,91].
• Edge AI provides efficient real-time threat detection and anomaly analysis, but remains vulnerable to adversarial attacks [25,30,64].
• Quantum-Resilient Security ensures data protection against quantum threats, but its practical implementation on current IoE devices remains debatable due to hardware and resource constraints [54,180,181].

Table 20. Comparison of blockchain, AI, and quantum security integration in IoE systems.

Technology	Security Contribution	Challenges	References
Blockchain	Decentralized security, data immutability	High transaction cost, latency issues	[73,90,178]
AI for Security	Autonomous threat detection, fast anomaly recognition	Vulnerable to adversarial AI attacks	[29,30,60,64]
Quantum Security	Long-term protection against quantum attacks	High computational requirements, hardware compatibility in IoE	[120,132,180,181]

provides a comparative overview of how Blockchain, AI, and Quantum-Resilient Security contribute to IoE protection, highlighting their respective strengths and current limitations in terms of scalability, attack resilience, and hardware compatibility.

In addition to the limitations shown in Table 20, energy efficiency is a critical consideration for IoE security solutions, given the resource constraints of many edge devices. Technologies such as blockchain and AI-driven analytics, while powerful, can be computationally intensive. Future designs should incorporate lightweight consensus algorithms, hardware accelerators, and model compression techniques to minimize power consumption. Studies have shown that optimizing task allocation between edge, cloud, and possibly Non-Terrestrial Networks (NTNs) nodes can significantly improve energy profiles without sacrificing security performance. This remains an open area for innovation, especially in battery-powered or energy-harvesting IoE systems.

5.4. Critical Open and Deployment Issues in IoE Security

While the proposed security architectures offer strong theoretical guarantees, their real-world deployment involves challenges such as scalability, energy constraints, protocol interoperability, and integration with legacy IoE infrastructures. The heterogeneous nature of IoE devices, ranging from high-performance edge servers to low-power sensors, requires adaptive frameworks that balance computational cost with security requirements. Insights from practical case studies assessing how distributed learning approaches perform in real-world 6G-based IoT scenarios [182] suggest that deployment success depends on modular, configurable architectures, cross-layer optimization, and the careful selection of cryptographic primitives according to hardware capabilities. In light of these findings, several fundamental security challenges remain unresolved and must be addressed in the coming years:

• How can Blockchain scalability be achieved in IoE environments?
• How can Edge AI models be made more resilient to attacks?
• How can the impact of quantum computing on IoE security be minimized?
• How can autonomous self-protecting security systems be developed for IoE devices?

5.5. Emerging Directions and Practical Considerations for IoE Security

5.5.1. Split Learning (SL) and Federated Learning (FL) Synergies

Recent advances in Split Learning (SL) have introduced new possibilities for privacy-preserving machine learning in IoE environments. Unlike FL, where only model updates are shared, SL partitions the model between client and server, transmitting intermediate activations instead of raw data. This approach can mitigate certain privacy risks, especially in environments with constrained devices or intermittent connectivity. Furthermore, combining SL with FL can exploit their complementary strengths, FL’s decentralized training and SL’s communication efficiency, to create hybrid frameworks for large-scale IoE deployments. For example, recent research on distributed multi-layer learning frameworks for intelligent transportation systems in next-generation aerial-ground integrated networks has demonstrated how such synergies can enhance both privacy and performance [183]. Future research should explore these hybrid designs in diverse IoE domains, particularly healthcare and smart city infrastructures.

Beyond SL and FL, Transfer Learning (TL) can further enhance learning efficiency in IoE security frameworks by reusing knowledge from pre-trained models. Integrating TL with FL or SL can accelerate convergence, reduce communication overhead, and improve model generalization across heterogeneous IoE nodes. Such integration is particularly promising for dynamic environments, where rapid adaptation to new attack vectors or operational contexts is essential.

5.5.2. Non-Terrestrial Networks (NTNs) for Secure IoE Connectivity

Non-Terrestrial Networks (NTNs), including satellite and aerial platforms, are emerging as critical enablers for IoE connectivity, particularly in remote or infrastructure-limited regions. Their integration into IoE security architectures can ensure seamless coverage, improve resilience against local network failures, and facilitate secure data exchange across heterogeneous domains. Recent research focusing on the use of distributed learning techniques for processing Earth observation data within complex, multi-layered non-terrestrial network architectures has demonstrated that NTNs can be combined with blockchain-based trust mechanisms and federated intelligence to support secure, low-latency communication in geographically dispersed IoE deployments [184]. As IoE adoption expands into rural, maritime, and disaster-response scenarios, NTN-enabled security frameworks will likely play a decisive role.

5.5.3. Regulatory and Interdisciplinary Perspectives

IoE security inherently spans multiple sectors, healthcare, transportation, manufacturing, and beyond, making regulatory compliance and interdisciplinary collaboration vital. Compliance with data protection regulations (e.g., GDPR) must be embedded into the design of security frameworks from the outset, ensuring lawful processing, user consent management, and auditability. Moreover, effective IoE security requires collaboration between cybersecurity specialists, AI researchers, communication engineers, and policymakers. Establishing such cross-disciplinary ecosystems will accelerate the translation of theoretical frameworks into deployable, sector-specific solutions.

5.6. Proposed Research Roadmap and Future Directions

Ensuring security, privacy, and system integrity in the expanding IoE ecosystem requires multi-layered and integrated solutions [3,10,20]. The following research roadmap is structured around two key dimensions:

• The individual technological maturation timelines,
• The evolutionary steps of integrated security architectures.

5.6.1. Timeline of Core Technology Maturity

This section presents the evolutionary timelines of key technologies that play a critical role in IoE security, including Blockchain [8,73,178], AI [25,59,164], Quantum-Resilient Cryptography [18,54,180], and IoE infrastructures [107].

Figure 11 presents a strategic timeline projecting how advanced technologies will evolve over the years to address the security challenges in IoE systems.

For the 2025–2028 period, it is assumed that the systems will still be in a relatively immature phase, characterized by resource-constrained devices. Therefore, security solutions will focus on lightweight and practical technologies such as optimized blockchain algorithms [178,179], PQC-compatible digital signatures [35,52,180], and FL methods that enable local model training without compromising data privacy [23,44,170]. During this phase, reactive and on-site security mechanisms like Edge AI-based anomaly detection [25,59,145] will also become prominent.

In the 2029–2032 period, infrastructures are expected to mature, and interactions between devices will intensify. At this stage, new requirements such as interoperability and resilience against quantum threats will emerge. Consequently, more complex, secure, and flexible solutions will be implemented, including QKD [18], post-quantum blockchain architectures [54], context-aware access control [84,160,171], and interoperable AI-based security systems [144,164]. Additionally, technologies like cross-chain privacy protocols and secure multi-party computation (SMPC) will gain importance during this period [47,175,176].

By the 2033–2036+ period, it is assumed that the systems will evolve into self-protecting and self-learning structures rather than merely reactive ones. Security will transition towards architectures that are based on zero-trust principles [161,164], capable of autonomous threat detection and mitigation [144,145], resistant to quantum attacks [54,81], and supported by semantic AI for proactive and self-managing security frameworks [136,166].

Thus, the figure represents not only a technological sequence but also the structural transformation of security in parallel with the maturation of system requirements.

Table 21. Strategic timeline and technology roadmap for security in IoE ecosystems.

Time Frame	Emerging Technologies and Challenges	Strategic Focus Areas
2025–2028	Early-stage FL and Edge AI deployments [25,144] Pilot PQC integration [35,180] Lightweight blockchain optimization [178,179]	FL infrastructure and data privacy [23,144,170] Lightweight consensus for IoT [73,91] PQ-safe digital signatures [52,180]
2029–2032	QKD and PQ Blockchain applications [18,54] Interoperable AI security systems [144,164] Context-aware access and decentralized identity models [84,160,171]	Cross-chain privacy and SMPC integration [47,175,176] DID and SSI-based access control [84,85] Adaptive AI for anomaly detection [59,145]
2033–2036+	Self-healing and autonomous security systems [144,164] Quantum Blockchain [54,181] Semantic AI + Zero Trust architecture [161,164]	Autonomous threat mitigation [145] Quantum-Resistant interoperable trust layers [54,132] Self-aware zero-trust agents [136,161]

illustrates the strategic roadmap for advancing security technologies in IoE ecosystems.

This roadmap outlines the evolving priorities in IoE security technologies between 2025 and 2036, focusing on scalable, interoperable, and quantum-resilient solutions aligned with FL, zero-trust architectures, and semantic AI.

5.6.2. Strategic Roadmap for Integrated Cyber-Physical Systems

This section presents the development stages of holistic systems where the mentioned technologies are used in an integrated manner. The reflections of this integration on cyber-physical systems are classified by years with a focus on functional gains. The main emphasis is on when multi-technology collaborative systems will acquire specific competencies [3,10,144].

Strategic Roadmap and Future Perspective: The integration of IoE, Blockchain, AI, and Quantum-Resilient Security Protocols is a strategic necessity in the evolution of cyber-physical systems [8,64]. The development of these technologies can be evaluated through a multi-stage roadmap.

Between 2025 and 2028, the groundwork phase focuses on early integration and awareness. Blockchain-based ABAC mechanisms are piloted in small-scale IoE environments [11,82,83]. Edge AI-based anomaly detection is experimentally deployed at the device level [25,59,145]. Hybrid architectures incorporating post-quantum cryptographic primitives begin initial testing [54,180,181]. The main challenges of this period include the lack of standardization, regulatory uncertainties, and protocol incompatibilities [10,14,38]. Additionally, FL [23,144], lightweight consensus algorithms [73,178,179], and homomorphic encryption [46,177] are explored to balance privacy and computational requirements.

Between 2029 and 2032, the focus shifts to widespread adoption and interoperability, marking a period of maturity and sectoral expansion. Privacy-preserving AI systems based on FL are integrated into corporate IoE platforms [23,144,170]. DID and SSI solutions are implemented across various sectors [84,85,160]. Blockchain-AI integrations find concrete use cases in healthcare [88], smart cities [142], and energy infrastructures [8,158]. Although post-quantum digital signatures do not become fully widespread, they begin to deliver stable test results in regulated environments [35,52,120]. The increasing need for interoperability directs attention toward context-aware access policies [171] and semantic AI solutions [136,164].

From 2033 to 2036 and beyond, the focus moves to adaptation and autonomy, achieving full integration and intelligent systems. Fully PQC-enabled blockchain protocols start to be established [54,67,180]. Edge AI becomes the fundamental security layer for real-time autonomous systems [25,144,145]. Cross-domain data sharing is facilitated through privacy-preserving computation techniques [47,176] and semantic blockchain networks [164,167]. Quantum-assisted behavioral modeling enables advanced threat prediction systems [130,164].

Beyond 2037, the proactive and predictive security phase begins. Cyber-physical systems evolve into self-healing architectures that are resistant to quantum attacks [144,161]. Locally trained quantum-assisted AI models are used in defense systems at the edge [130,136]. IoE devices transform into self-defending systems through autonomous security policies based on identity, behavior, and context [161,164].

6. Conclusions

This study presents a holistic analysis of the integrability of blockchain, edge AI, and quantum-resilient approaches within the context of IoE security. The global proliferation of IoE devices has revealed the inadequacy of traditional security architectures, increasing the need for next-generation, multi-layered security solutions. From this comprehensive review, five key conceptual conclusions have been drawn:

1.

IoE Security Requires Multi-Layered and Context-Aware Architectures: Traditional security methods based on centralized structures are not suitable for the distributed nature of IoE. Blockchain-based identity and access control systems such as DID and SSI can play a critical role in addressing this gap.

2.

Edge AI Can Reduce Privacy and Latency Issues in IoE: Processing data at the edge device, without relying on centralized cloud systems, not only reduces privacy risks but also improves decision-making processes with lower latency. Models like FL enable local machine learning without sharing data with a central server.

3.

Quantum Computing Threatens IoE Security and Existing Infrastructures, Requiring New Cryptographic Solutions: Algorithms like Shor’s and Grover’s undermine the foundations of classical cryptography. PQC, including lattice-based encryption, hash-based methods, and quantum-safe blockchain protocols, is critical to the future evolution of security.

4.

Security Solutions Vary by Application Context: Each security layer (blockchain, Edge AI, quantum security) has different advantages and trade-offs in terms of scalability, energy consumption, and computational load. Therefore, hybrid solutions and context-aware strategies must be developed.

5.

Open Research Areas Remain:

• How can threat prediction systems be developed using Quantum AI to optimize IoE security?
• How can blockchain-based security protocols become more scalable for IoE environments?
• How can energy-efficient, low-computation, privacy-preserving solutions be built for edge devices?

Looking forward, research in IoE security must be planned in stages according to the maturity levels of related technologies. In the short term, priorities include developing FL infrastructures, testing the practical deployment of Edge AI systems, and conducting early trials of post-quantum digital signatures. Concurrently, research should focus on energy-efficient and lightweight blockchain consensus mechanisms, as well as identity and access control systems that can be seamlessly integrated into IoE devices.

In the medium term, integration of QKD and post-quantum blockchain protocols into IoE systems should be pursued. Moreover, hybrid AI-blockchain solutions and context-aware access control models should be widely adopted to enable interoperable security frameworks.

In the long term, the focus is expected to shift towards autonomous, self-healing security systems, fully quantum-resistant protocols, and semantic AI solutions operating in synergy with zero-trust architectures. In this scenario, IoE devices are envisioned not only to protect themselves from threats but also to predict potential risks and autonomously manage their own security, setting the vision for the future of cyber defense.

In conclusion, the rapidly expanding IoE universe mandates security strategies that go beyond conventional approaches, requiring multi-layered, context-aware, and future-proof solutions. Blockchain, Edge AI, and Quantum-Resilient Security solutions will become foundational pillars of the IoE ecosystem, playing a critical role in ensuring data security and system integrity.

This study aims to provide a comprehensive framework for IoE security, offering guidance for researchers, industry professionals, and policymakers in terms of both technological evaluation and future research directions. Future studies should not only focus on security but also consider scalability, energy efficiency, and sustainability as integral research priorities.