The emergence of 6G heterogeneous networks integrating unmanned aerial vehicles (UAVs), intelligent reflecting surfaces (IRSs), Internet of Things (IoT) devices, and fog/edge nodes creates new opportunities for intelligent and latency-sensitive applications while introducing significant security challenges. Traditional authentication mechanisms are inadequate for such dynamic, distributed, and heterogeneous environments that require secure collaborative communications. This paper proposes an authentication scheme based on Fog-RAN (Fog Radio Access Network) and a dual-blockchain architecture with smart contracts and elliptic curve cryptography (ECC). The proposed scheme provides secure network access, mutual authentication, traceability, auditability, and zero-trust enforcement. Formal verification using the ROR model, AVISPA and performance evaluation through smart-contract simulations indicate resilience to common network and cryptographic attacks and improved efficiency. Compared with existing schemes, the proposed approach reduces computation cost, bandwidth, and energy consumption by 64.2%, 59.6%, and 31.4%, respectively. These results support the suitability of the scheme for secure, scalable, and energy-efficient authentication in next-generation 6G networks.

This article continues the series of interviews with the Officers of the IEEE ComSoc Member and Global Activities (MGA) Council for the term 2026–2027, published every month in the Global Communications Newsletter.

During the 2024–2025 biennium, several activities of the Distinguished Lecturer Program were organized by different ComSoc chapters in Latin America. The main goal of the Distinguished Lecturer Program is to offer the benefits of world-class lectures to IEEE members around the globe. Particularly in Latin America, due to limited access to in-person high-level presentations and courses, as well as the high costs of traveling to and attending major international conferences outside the region, the Distinguished Lecturer Program is a vital initiative for attracting new student and professional members to ComSoc. It also serves as a strong retention tool by highlighting the technical and professional value of being an IEEE ComSoc member.

Federated Learning (FL) is susceptible to adversarial attacks, such as Label Flipping (LF) and Backdoor, where malicious clients manipulate the updates of their local model to reduce the global model’s performance. Traditional FL relies on a centralised aggregator, which must be trusted, creating a single point of failure. This centralization not only increases computational cost but also introduces scalability challenges. To address these issues, we propose a Blockchain (BC) based FL framework that decentralises the aggregation process and incorporates an enhanced Density-Based Spatial Clustering of Applications with Noise (DBSCAN) method to identify and remove malicious updates without ignoring minor groups. Our approach eliminates the need for a centralised aggregator by leveraging BC’s smart contracts to aggregate the global model. Simultaneously, Enhanced DBSCAN identifies malicious updates in the parameter space, effectively mitigating adversarial influence while preserving privacy. We evaluate both of our framework and the traditional FL under LF redand Backdoor attacks, experimental results demonstrate that our approach outperforms the traditional FL according to multiple metrics, including accuracy, loss, precision, recall, and F1-score. These findings emphasise the effectiveness of our BC-based decentralised aggregation combined with enhanced DBSCAN technique in improving the robustness and security of FL systems.

The rapid proliferation of the Internet of Things (IoT) has heightened exposure to diverse security threats, particularly Distributed Denial of Service (DDoS) attacks, which underscore the limitations of traditional security mechanisms in protecting resource-constrained devices. This study introduces a novel threat detection framework that exploits transport layer congestion control features to detect security breaches in real time by monitoring key indicators such as packet loss, round-trip time (RTT), and retransmission rates. The proposed lightweight solution integrates machine learning techniques to effectively distinguish between genuine network congestion and congestion induced by malicious activity. Through extensive IoT simulations, the framework demonstrated a Detection Rate (DR) of 96% with a minimal False Positive Rate (FPR) of 1.2%. It further achieved a Packet Delivery Ratio (PDR) of 95%, an average latency of 150 ms, a throughput of 5000 bytes/s, and a Response Time (RT) of 120 ms. These results confirm the framework’s ability to provide reliable, real-time IoT security while preserving high network performance.

The paper suggests a distributed cross-layer IoT architecture that combines LoRaWAN (Long Range Wide Area Network) with federated learning (FL) to improve reliability, scalability, and fault tolerance in multi-layer vertical farming systems in dense and dynamic environments. Unlike the traditional frameworks that rely on independent measures of QoS (Quality of Service), the proposed framework directly represents the inter-layer relationships, such as heterogeneity of latencies, robustness of connectivity, and propagation of faults. One of the contributions is the development of a cohesive cross-layer evaluation framework with six strictly defined metrics: MLDC (Multi-Layer Deployment Capacity), C-LCRI (Cross-Layer Connectivity Robustness Index), C-LFCI (Cross-Layer Fault Containment Index), SART (Smart Adaptive Recovery Time), and AIRSM (AI Resilience Score Metric), which allows for quantitatively characterizing latency differences, network resilience, fault containment, recovery efficiency, AI robustness, and energy-performance trade-offs. The experimental results show that the proposed Smart Distributed LoRaWAN–Federated Learning architecture operates reliably in high-density and multi-layer vertical farming environments, and is scalable to handle larger amounts of data. The proposed system guarantees a packet delivery ratio (PDR) of around 95% under a large-scale deployment with up to 1050 IoT nodes spread across seven cultivation layers, with a latency reduction of nearly 60%, less than 1.6 J/msg on average energy consumption, and a fault recovery time of less than 0.3 s in case of network disruptions. The proposed framework was validated using large-scale simulation scenarios developed based on experimentally reported LoRaWAN communication characteristics and agricultural IoT deployments, and operational conditions at the edge intelligence. This evaluation included up to 1050 sensing nodes in 7 vertical farming layers to approximate a realistic deployment of smart farming in a large-scale environment while keeping consistency with the recorded communication and reliability profile.

Vehicular ad-hoc networks (VANET) are the key to advancing intelligent transportation and reducing traffic congestion. Nevertheless, the rapid increase in vehicle diversity introduces numerous challenges to maintaining secure and efficient vehicle communications. To address these challenges, we propose a new identity-based encryption scheme (IBES) based on lattice-based cryptography for secure communication in VANETs. Our scheme ensures timely message receipt, preserves the confidentiality of vehicle identity, and enables rapid authentication. Our scheme protects VANET devices against both active and passive attacks in their resource-constrained environment. The formal analysis verifies that adaptively chosen message attacks cannot break the encryption scheme using the random oracle model. The informal analysis shows that the scheme meets the security goals and is strong enough to protect against side-channel, man-in-the-middle (MITM), and replay attacks. Additionally, the SCYTHER tool is used for formal security validation, and the security analysis shows that the protocol is strong against chosen-plaintext attacks. We evaluate the effectiveness of the proposed scheme against existing approaches, focusing on the security attributes offered, computation, and communication cost. Our scheme shows lower storage overhead and higher authentication efficiency than similar existing schemes. Therefore, this combination establishes the proposed scheme as an efficient and practical solution for VANET.

Multi-hop communications are known to be more energy efficient than direct or single-hop communications in wireless networks. However, in multi-hop communication networks, intermediate relaying nodes introduce a new degree of vulnerability to security threats. In this paper, we propose a secure energy-aware multi-hop communication protocol for M2M (Machine to Machine) wireless area networks. A solution based on a two-layer blockchain with a sidechain integration, located at edge level, is adopted to ensure authentication in a preliminary phase. Depending on a specific application, the communication phase from machine to machine is achieved via base stations or directly using D2D (Device to device) communications in multi-hop mode. Simulation results confirm that the proposal, compared to other protocols, provides a secure M2M communication with relatively low level of energy consumption, limited computation and storage overheads.

This paper studies an efficient DT training problem in network intelligence and focuses on a cloud-edge-end network scenario. We first propose a hierarchical grouped federated learning framework with knowledge distillation (FedHG-KD) for training a DT model for the network. Then, we consider a sub-global model type deployment (MTD) and hyper-parameters setting (HPS) problem for training a DT model based on the framework. We formulate the problem as a mixed optimization problem with an objective to minimize DT training delay by optimizing the deployment of sub-global model types, the frequencies of local and sub-global model aggregations, and the number of global model updates while satisfying the loss requirement of the DT model. To solve the formulated problem, we first decompose the problem into an MTD subproblem and an HPS subproblem, and then propose an MTD-HPS algorithm, which incorporates an MTD algorithm and an HPS algorithm to solve the two subproblems. Simulation results show that the proposed FedHG-KD framework and MTD-HPS algorithm can significantly improve the DT training performance in terms of model accuracy and DT training delay as compared to several benchmark algorithms.

In today’s interconnected environments, the rapid growth of Internet of Things (IoT) devices has led to a dramatic rise in data traffic, placing increasing demands on energy-constrained networks. Traditional routing protocols are often unable to meet the challenges posed by battery-powered IoT devices, which must balance efficient data transmission with limited energy availability. To address these challenges, we introduce ARROW (Additive Reinforcement-driven Routing for Optimal Wireless networks), a new paradigm in IoT routing that combines reinforcement learning with additive encoding embedded directly in packet headers. This integration enables lightweight, decentralized learning, allowing each node to dynamically adapt its routing behavior based on real-time network feedback and in-packet energy metrics. In this paper, we define ARROW-IoT as a specific implementation of the ARROW paradigm, applied to IoT networks. It leverages Q-learning and additive header encoding to enable energy-aware, decentralized routing decisions in dynamic, resource-constrained environments. ARROW highlights how in-packet intelligence can support autonomous, energy-aware decision-making across dynamic networks with minimal overhead.

Within a NOMA-based cognitive wireless network (CRN) framework (e.g. in B5G), video users must collaborate to enhance their perceived QoE and energy efficiency while adhering to interference limitations. Traditional approaches could hardly solve the continuous apace non-linear programming problems raised in CRNs well. They also cannot handle issues such as imperfect channel state information and scalability when the network dimensions surpass a certain limit. Therefore, this paper employs Deep reinforcement learning (DRL) methodology to achieve energy-efficient non-linear quality control for each secondary/cognitive user (SU). Accordingly, this paper employs deep deterministic policy gradient (DDPG) as a model-free, off-policy actor-critic, sample-efficient and continuous actin space DRL sub-type to facilitate scalable energy-efficient quality control for each cognitive user while guaranteeing the total imposed interference to PUs. Two centralized and cluster-based implementations of DDPG are presented. The centralized deep reinforcement learning quality control (DRLQC) approach can efficiently handle a large number of video users in dense CRN scenarios. As the centralized approach is not efficient in ultra-dense scenarios, we have proposed another distributed multi-agent version of DRLQC (named MADRLQC). Numerical results indicate that these methods surpass similar approaches in maximizing the overall perceived QoEs of cognitive users while adhering to energy efficiency constraints and minimizing computational complexity. It can be verified that the proposed DRLQC methodology can improve the energy-efficiency with respect to DQN method by 21.8% in dense scenarios. Moreover, the proposed MADRLQC methodology can improve the energy-efficiency with respect to the DQN and DRLQC methods by 24.5% and 3.38% respectively in ultra-dense scenarios.

The rapid growth of IoT in healthcare demands reliable, secure, and energy-efficient communication solutions. We propose CLEHTO, a novel cross-layer optimization framework that dynamically adapts to network conditions by integrating real-time energy monitoring, joint mobility-security assessment, and adaptive congestion control. Unlike conventional approaches, CLEHTO introduces a unified reliability scoring system that simultaneously evaluates physical channel quality, link reliability, node mobility, and transport-layer congestion. Experimental results demonstrate CLEHTO's exceptional performance: a 92.8% Packet Delivery Ratio under 20% link failure while maintaining 4.5 Mbps throughput, a 98.6% authentication success rate using SSL/TLS (outperforming IPSec's 98.3%), and optimal energy consumption of 0.55 mAh for battery-powered devices. CLEHTO maintains 105 ms latency (vs. IPSec's 95 ms) for secure medical data flows, showing significant improvements over single-layer approaches. These results establish CLEHTO as a robust and efficient solution for IoT-based healthcare systems.

Ensuring reliable communication and energy efficiency remains challenging in Internet of Everything (IoE) environments, where devices are heterogeneous, mobile, and energy-limited. This paper presents a cross-layer optimization framework that leverages the physical, data link, network, and transport layers to enable adaptive route selection, balanced traffic management, and energy-aware data transmission. Coordinating these layers improves data delivery, reduces latency, and extends device operational lifetime. Experimental results show a 24% increase in delivery rate, 0.85 ns reduction in latency, 37% longer device activity, and up to 35% energy savings. These findings demonstrate the framework’s effectiveness for robust, high-performance, and energy-efficient communication in heterogeneous IoE systems. Overall, this work provides a holistic cross-layer design integrating routing, traffic adaptation, and energy management for reliable and sustainable IoE networking.

Over recent years, research has continued on the subject of Ambient Backscatter Communication (AmBC), owing to the low-power, passive nature of this technology. It shows great promise for the growing networks of the Internet of Things (IoT), which now include billions of devices. Despite the potential, AmBC system applications are still limited by low data rates and short range. To help overcome these challenges, Intelligent Reflective Surfaces (IRS) have been suggested as a means to improve the quality of the backscattered signal. This paper investigates the effect of integrating IRS with an AmBC system to enhance the quality of the backscattered signal. The simulation results show that the implementation of IRS significantly improves the bit error rate (BER), as well as the received signal power. These findings suggest that IRS shows promise for scalable and robust AmBC networks.

The Internet of Things (IoT) and its industrial counterpart, the Industrial Internet of Things (IIoT), have transformed sectors such as home automation, healthcare, and manufacturing by enhancing data management through advanced networking. However, the rapid growth of IIoT has introduced significant cybersecurity challenges, necessitating a comprehensive approach to securing data across the TCP/IP model. This paper presents a novel cybersecurity investment strategy formulated as a bi-objective optimization problem, validated through genetic and iterative algorithms. The strategy effectively balances security and cost, achieving nearly 50% efficiency in solution effectiveness. By utilizing these optimization techniques, the approach provides a practical and cost-effective solution to improve IIoT security within budget constraints, offering valuable insights for cybersecurity professionals seeking robust and economically viable solutions.

The Internet of Things (IoT) facilitates real-time connectivity of objects, allowing for access from anywhere at any time. For IoT Low-Power and Lossy Networks (LLNs), the Routing Protocol for Low-Power and Lossy Networks (RPL) has been introduced. In RPL-based topologies, the rank of nodes reflects their positions within the network, calculated by adding the rank of a node's preferred parent to the link metric between them. However, due to inaccuracies in assigning link metric values to neighboring nodes, frequent changes in preferred parent selection occur, resulting in significant control overhead, increased energy consumption, higher latency, and degraded Packet Delivery Ratio (PDR). This paper presents an optimized path selection method that ensures the most stable and optimal choice of preferred parents for nodes. Using the Cooja simulator under various network densities, the proposed approach demonstrates a 73% reduction in preferred parent changes, a 49% decrease in control overhead, and a 50% reduction in total energy consumption. Additionally, it improves PDR by 46% and reduces latency by 2.81 seconds.

Federated Learning (FL) has emerged as a promising approach for privacy-preserving collaborative model training, especially in high-stakes applications such as e-health. However, FL is highly vulnerable to data poisoning attacks, where malicious clients introduce corrupted updates to compromise model integrity. This paper proposes a Federated Ensemble Learning (FEL) framework to detect and mitigate such poisoning attempts using ensemble learning techniques. FEL leverages multiple independent models to identify adversarial behavior while preserving privacy. Our method predicts and filters poisoned data to enhance disease detection accuracy. We demonstrate that FEL offers superior robustness against labelflipping attacks compared to baseline FL models. Experiments on clinical datasets confirm FEL's effectiveness in rejecting poisoned updates while maintaining high prediction accuracy (97%). FEL enhances FL security mechanisms and provides a scalable solution for decentralized healthcare systems.