The exponential growth of Internet of Things (IoT) devices necessitates fog computing architectures that balance network performance with energy efficiency and environmental sustainability. Traditional fog node placement algorithms decouple energy considerations from optimization processes, leading to excessive grid dependency and substantial carbon emissions. This research introduces Green D-OXA, a novel energy-efficient algorithm for dynamic fog node placement with integrated renewable energy harvesting in sustainable IoT networks. Green D-OXA extends the bio-inspired OX optimizer through four adaptive mechanisms: energy-aware warm-start initialization, adaptive iteration control, intelligent triggering with renewable energy prediction, and explicit solar-wind harvesting models with battery management. The algorithm formulates continuous multi-objective optimization integrating connectivity, coverage, movement costs, energy consumption, renewable utilization, and carbon reduction. Comprehensive experimental evaluation across five dynamic scenarios mobile fog nodes, equipment failures, time-varying traffic, network expansion, and combined dynamics demonstrates superior performance compared to three established baseline algorithms (SPP-TLBO, CSA-FSPP, SPP-DEA). Green D-OXA achieves 97.8% connectivity, 98.4% coverage, 68.5% renewable energy utilization, and 43.4%-56% CO₂ emission reduction. Scalability analysis from 50 to 1000 nodes confirms practical deploy ability with minimal performance degradation and 3.8%-4.9% energy overhead. Results establish Green D-OXA as an effective solution for sustainable large-scale IoT-fog computing infrastructures, advancing green computing initiatives through intelligent renewable energy integration.

Federated Learning (FL) enables collaborative model training on decentralized data, offering privacy advantages but struggling with data quality variations and adversarial attacks. This paper introduces FEDMAD (Federated Learning for Medical Data with Enhanced Defense), a novel framework designed to enhance robustness in such environments. FEDMAD integrates Homomorphic Encryption (HE) for model update privacy with a quality-aware aggregation mechanism based on a client’s local training loss (1/loss). Our key contribution is the robust aggregation of these quality scores using Median Absolute Deviation (MAD)-based clipping to defend against dishonest score reporting by adversaries. We evaluated FEDMAD on a real-world smoker prediction task using the TenSEAL HE library. Results demonstrate that FEDMAD’s quality-aware mechanism effectively mitigates the impact of noisy clients. More importantly, MAD-based score aggregation is essential for neutralizing dishonest score reporting attacks and preventing model collapse, a scenario where simpler percentile-based clipping fails. While FEDMAD shows significant resilience, our study highlights remaining challenges with sophisticated model poisoning attacks, suggesting directions for future research.

Security in Fog based IoT networks has a major problem where some IoT devices may be compromised due to attacks. This creates vulnerability to the sensitive data flow between Fog and IoT devices. Also, due to heterogeneity in network structure, the probability of attacks in the network is more. To analyze this dynamic and complex structure of fog networks, few shot learning is be a good solution to learn patterns more accurately and identify the maleficent or normal behavior of a device. In deep learning, few shot learning technique is a technique that uses less amount of labeled data for data processing that is more efficient than the traditional deep learning models. In this work, a hybrid MAML (Model Agnostic Meta Learning)-Reptile based few shot learning approach is proposed that secures the fog based IoT infrastructure from variety of attacks by detecting the attacks more accurately. The few shot models considered are Prototypical Networks, Matching Networks, MAML, and Reptile for selection of best model to be run at the fog server for attack detection. The fog node uses the best model to detect the attacks and broadcasts the local behavior list (LBL) to cloud and other fog nodes to generate the global behavior list (GBL) for sharing more attack information to the IoT device layer. Here, we consider standard datasets such as UNSW-NB-15, NSL-KDD, and CICIDS for performing implementations. The performance of the models is analyzed using accuracy, recall, f1-score, precision, inference time, training time, AUC-ROC, cost, energy consumption, and processing latency. From the results, it is observed that the proposed MAML-Reptile hybrid model performs better than other standard models in detecting maleficent behaviors more accurately.

The Elliptic dataset is widely used in Bitcoin anti-money laundering research, yet its original anonymized features have limited forensic interpretability. Much of the existing Elliptic-based literature relies on these opaque benchmark variables, leaving insufficient attention to semantically explicit and interpretable graph representations for illicit transaction detection. To address this gap, this article proposes a combined approach that integrates transaction-level feature reconstruction with interpretable forensic descriptor engineering. First, the benchmark’s original feature space is replaced with a semantically explicit reconstructed representation derived from public on-chain transaction data and metadata after resolving benchmark node identifiers to transaction hashes. Second, the proposed approach extends this reconstructed representation with interpretable forensic descriptors that capture local transaction abnormality, outgoing value redistribution behavior, and deviations from upstream transaction history. The empirical design isolates the contribution of the proposed descriptors by comparing the reconstructed representation against its descriptor-augmented variant. Across eight classifiers evaluated under a whole-snapshot train-test protocol that preserves within-snapshot graph structure, the descriptor-augmented representation consistently improves illicit class retrieval. CatBoost achieves the best results, increasing the area under the precision recall curve for the illicit transaction class from 85.10% to 90.27%, precision from 77.49% to 87.04%, recall from 75.57% to 81.11%, and F1-score from 76.44% to 83.90%. The article also discusses how the predictive component can be embedded into a hybrid analytical framework that separates machine learning classification from address-level forensic interpretation. This structure supports explainable prioritization and expert review while preserving the distinction between predictive evidence and forensic interpretation. Overall, the findings demonstrate that semantically explicit and forensically interpretable representations can substantially improve illicit transaction retrieval while supporting transparent post hoc analysis in Bitcoin anti-money laundering research.

Given the rapid spread of images in the digital domain via open networks, maintaining the confidentiality of information and the secrecy of transmission has become a major challenge. Keeping this in view, the present paper proposes a new hybrid security framework, Cartesian 2D Logistic Chaotic Map Steganography and Vernam-style XOR operation Shannon Cryptography (CLCMS-VCSC), for the secure transmission of images. In the proposed framework, a chaotic map, namely the Cartesian four-quadrant 2D logistic chaotic map, has been used for embedding the encrypted data by utilizing the mechanisms of confusion and two-stage diffusion, while the Vernam-style XOR operation and Shannon entropy analysis provide robustness to the encryption technique. Deterministic symmetric-cipher analysis of the generated ciphertext has also been performed for evaluation only, i.e., to calculate entropy and conditional probabilities, without affecting the encryption technique’s deterministic nature. Evaluation of the proposed framework has been performed on the BOSSBase v1.0.1 dataset, comprising 10,000 grayscale images of size 512×512, achieving a maximum Peak Signal-to-Noise Ratio (PSNR) of 45.7 dB and Structural Similarity Index (SSIM) of 0.98, outperforming existing methods under the same experimental conditions. In addition, the proposed framework also exhibits low execution time and a key storage cost of about 20-21.9 bits. The results verify the effectiveness of the CLCMS-VCSC framework in terms of security, visual quality, computational cost, and key management; thus, the framework is more appropriate for secure and covert image communication in contemporary digital settings. The ablation analysis also validates the significance of each proposed module in improving the framework’s performance, thereby verifying the architectural novelty of the CLCMS-VCSC framework.

Cloud computing forms the basis for the emerging technologies in various fields, providing a reliable framework for managing resources to meet the needs of different applications. The rapidly increasing energy requirements inherent to cloud computing pose a real problem concerning sustainability. Energy efficiency, fair resource sharing, and performance consistent across the dynamic and heterogeneous cloud computing system are essential since existing approaches introduce inefficiency, energy consumption, and unfair distribution of loads. This research introduces Adaptive Osprey-Bowerbird Optimized Green Cloud Computing with Randomized Attention Coupled Fair Resource Distribution in Scalable Systems (AO-BO-RNCN-MAN) to address these challenges. The proposed framework integrates the Randomized Neural Coupling Network to learn diverse data representations, with the Multi-instance Attention Network to prioritize tasks, and Adaptive Osprey-Bowerbird Optimization, which is a combination of the Osprey Adaptive Algorithm and the Adaptive Bowerbird Optimization for further fine-tuning of the system. By optimizing the placement of virtual machines and scheduling of tasks, the proposed framework guarantees fairness and high utilization of energy with low turnaround time. Performance assessments indicate that the proposed framework outperforms the existing systems with energy efficiency of 99.82%, precise task scheduling of 99.61% and fair resource allocation of 99.74%. AO-BO-RNCN-MAN not only proposes a new way of addressing green computing challenges but also opens the gates to sustainable, adaptive, and scalable designed cloud infrastructures for resource management in cloud ecosystems and establishes the proposed conceptual framework as a new standard.

Vehicle to Vehicle communication (V2V) is the foundation of intelligent transportation systems, but due to high mobility and frequent topology changes, reliable and secure routing is still a challenge, and it is further exacerbated when vehicles are potentially malicious. The existing trust-aware routing protocols, e.g., Trusted Context-aware Opportunistic Routing (TCOR), rely heavily on heuristic and deterministic trust aggregation mechanisms, which are less effective in achieving optimal tradeoff between trust and routing efficiency as vehicular environments change dynamically. In order to overcome these limitations, we model the trust-aware V2V routing as a multi-objective optimization problem and design a new routing scheme based on a Quantum Inspired Trust-Aware Opportunistic Routing (Q-TAOR). The proposed method adoptedly choose secure forwarding paths in the face of malicious by taking into account trust maximization and routing efficiency. An effective quantum-inspired probabilistic representation is employed to extend the solution search space and generate reliable routes more efficiently via convergence yet does not depend on static decision rules. Therefore, the routing scheme integrates both direct and indirect trust observations and embeds optimization within a robust path selection process under highly dynamic scenarios. Results obtained using OMNeT++ show the effectiveness of the proposed approach under realistic vehicular mobility and attack circumstances. Simulation outcomes are valid for the proposed quantum-inspired trust-aware routing algorithm to be optimal for secure V2V communication: when attacker nodes are produced, highest-performing packet delivery ratio and robustness in comparison to TCOR, TCOR-Rec and conventional routing protocols.

This study introduces a hybrid machine learning framework for Wireless Sensor Networks (WSNs) designed to enhance energy efficiency and extend network longevity. The model integrates Q-learning for adaptive routing, hybrid clustering through Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), and decision tree regression for predictive energy depletion analysis. By dynamically balancing energy consumption and rerouting data to circumvent nodes approaching exhaustion, the framework improves reliability and operational stability.
Simulation results demonstrate notable improvements over conventional protocols such as LEACH and PEGASIS, achieving a 40% reduction in energy consumption and a 37.76% extension of network lifespan. Statistical validation (t-test, p < 0.0001) confirms the significance of these results. The proposed approach holds promise for deployment in real-world WSN and IoT applications, where optimized energy utilization and extended network lifetime can reduce maintenance costs and ensure continuous, reliable data acquisition.

Mobile Edge Computing (MEC) handles energy constraints and enhances performance by facilitating the effective offloading of applications that are delay-sensitive and computationally demanding from mobile devices. Nevertheless, high computing complexity, network limits, and the possibility of task failures brought on by user mobility and resource constraints make efficient task scheduling difficult. To address the limitations, the Q-Optimize OffloadPro Framework (QOOPF) is proposed as a task scheduling and offloading system designed to manage high virtual machine utilization, reduce latency, and improve resource efficiency in MEC. The framework incorporates the OffloadPro Scheduling Method (OPSM), which optimizes task assignment by prioritizing tasks based on a critical path approach to ensure effective offloading. To ensure that task offloading choices in edge computing settings are made dynamically, this technique is augmented by a Deep-Q-Driven Policy-Value Optimizer that has been trained on large amounts of task data. QOOPF dynamically balances computational loads, reduces task failures, and increases resource consumption by combining Policy Value Optimization (PVO) with Q learning. The experimental findings demonstrate QOOPF achieves a makespan of 720 seconds and variance of 30.03 for 300 tasks, with VM results showing a makespan of 445.88 seconds and variance of 4.58 for 16 VMs, scaling efficiently with up to 608.54 seconds and 6.08 variance for 32 VMs for high-demand MEC situations. This method provides an efficient, scalable solution for dynamic computing requirements while successfully addressing scheduling constraints.

The Internet of Medical Things (IoMT) allows ongoing monitoring and automatic analysis of physiological signals, e.g., electrocardiogram (ECG) or similar ones. Nevertheless, the high level of classification, feature representation, and computational viability in the IoMT resource-constrained environment remains a challenge. Traditional machine learning algorithms have been characterized by poor scalability and poor inter-feature modeling in ECG signals. To overcome these constraints, the present research proposes an ECG classification model based on a Similarity Directed Graph Neural Network (SDGNN) that encodes ECG features as graph-structured data to model their relationships explicitly. To improve classification efficiency and convergence stability, a Mountaineering Team-Based Optimization (MTBO) algorithm is used to optimise parameters and fine-tune models. The experimental assessment of the benchmark ECG datasets shows that the suggested SDGNN-MTBO framework is even more accurate and precise than the regular methods, while consuming less computing resources. The framework achieves 99% classification accuracy, indicating its suitability for conducting a reliable analysis of the ECG signal in a healthcare monitoring system that employs the IoMT.

Mobile Ad-hoc Networks (MANETs) play a crucial role in defense, disaster relief, and autonomous operations but remain highly exposed to threats such as blackhole, wormhole, and Sybil due to their decentralized topology, while traditional centralized trust mechanisms collapse under dynamic scenarios. This work presents the Blockchain-Fick Gradient Model for Secure MANET Routing and Threat Analytics (FiGRO-CoDpAT), combining blockchain consensus, gradient-based routing, and intelligent intrusion detection. The process begins with Network Initialization using Converged Blockchain Media Consensus (Co-BM-Co) for decentralized node verification. Fick’s Gradient Route Optimizer (FiGRO) then establishes congestion-free, attack-resistant routing. Following this, intrusion detection is performed through the Cosine Dual Phase Aggregator Transformer (CoDpAT), merging Cosine Convolutional Neural Network (CoCNN) and Dual Phase Aggregator Transformer (DpAT) for accurate packet analysis. Blockchain Trust Updates consistently maintain node credibility, while the Mountaineering Team Adaptive Optimizer (MtAO) enhances network efficiency in fluctuating topologies. Simulation findings prove the framework’s effectiveness, reaching an Accuracy of 99.5%, a Packet Delivery Ratio of 99.6%, a Packet Loss of only 0.4%, and a very low delay of 99.72 ms. In summary, FiGRO-CoDpAT provides secure, adaptive, and efficient communication in hostile MANET conditions.

Mobile Ad-Hoc Networks (MANETs) are self-organizing networks without any fixed infrastructure, which are decentralized and, thus, find applications in a dynamic and infrastructure-less environment. Nevertheless, these networks face significant challenges, including excessive energy consumption, malfunctioning routing, and security risks posed by malicious nodes. Such issues tend to cause higher communication overhead and a shorter network lifetime, particularly in highly mobile environments. To overcome these challenges, this paper proposes a dynamic, energy-saving routing architecture that could improve the network's Security and reliability. The suggested solution assesses the reliability of network nodes through energy performance and communication reliability, and continuously analyzes network traffic to identify malicious activity in real time. Data delivery is guaranteed through secure route selection and smart intrusion detection, allowing the framework to reduce superfluous energy consumption. The obtained simulation outcomes show that the suggested approach achieves 92.8% and 96.5% in the packet delivery ratio and intrusion detection, respectively, which is a strong indication of an impervious defense against attacks. Moreover, the method is highly energy-efficient, has a longer network lifetime, and is thus a good fit for the practice of MANET use, such as emergency response, military communications, and mobile IoT networks.