ABSTRACT The need for effective spectrum allocation to guarantee highly reliable, minimal‐latency, and sustainable connections has increased due to the exponential expansion of wireless communications and the introduction of 6G technologies. In order to cope with these challenges, the proposed research is Efficient Resource Management in 6G to Sustainable Spectrum Allocation using a Dual‐Channel Temporal Convolutional Frequency‐Separated Attention Network with Generation Rights Trading Blockchain (ERMSA‐6G). This objective is achieved by means of a layered network model that combines spectrum sensing, dynamic allocation and security. The Communication Layer is based on a Generation Rights Trading Blockchain (GRT‐ BC ) with a smart contract algorithm to allocate resources transparently, and the Security Layer uses an Ultra‐Lightweight Encryption Algorithm (ULwEA). This is improved by the use of the Dual‐Channel Temporal Convolutional Network (DCTCN) of short‐ and long‐term demand trends, and Frequency‐Separated Attention Network (FSAN) of fine‐tuning attention modeling, collectively to create the Dual‐Channel Temporal Convolutional Frequency‐Separated Attention Network (DCTCN‐FSAN). Hyperparameters are further tuned using Cleaner Fish Optimization (CfO), and verified results are stored on GRT‐ BC . Experimental results demonstrate that ERMSA‐6G achieves a 99.18% spectrum efficiency, 99.44% blockchain security, a 29.8‐dB signal‐to‐interference‐plus‐noise ratio (SINR) margin, and an 85‐ms consensus delay, while reducing transaction and servicing latency to 2060 ms. Its statistical tests prove that it is significantly better than the current techniques. To sum up, ERMSA‐6G provides a secure, transparent, and sustainable framework of 6G spectrum allocation, which is more effective than the state‐of‐the‐art techniques and guarantees scalability, reliability, and efficiency in future wireless networks.
ABSTRACT Internet of Things (IoT) applications are becoming increasingly popular owing to their extensive application in a variety of real‐world services. Many devices in the IoT ecosystem are connected to each other over the internet, making IoT networks vulnerable to different sorts of cyberattacks; network security and user privacy are important concerns in its deployment. In this paper, an advanced framework leveraging contrastive multilevel graph neural networks with snow avalanches algorithm to enhance IoT network security for intrusion detection (CMGNN‐SAA‐IoT‐ID) is proposed. Initially, the raw network traffic is collected from CICIoT2023 and CIC‐MalMem‐2022 datasets. The collected data are given to the preprocessing, where the Bayesian boundary trend filtering (BBTF) is applied for data cleaning, missing values, and data normalization. Then the preprocessed data are fed to feature selection using banyan tree growth optimization (BTGO) to select best features. Then the selected features are fed to intrusion detection using contrastive multilevel graph neural networks (CMGNN), which is used to detect the attacks as recon, DoS, DDoS, brute force, web‐based, spoofing, and Mirai from CICIoT2023. Additionally, the CIC‐MalMem‐2022 dataset is used to detect attacks as benign and malware. Finally, snow avalanches algorithm (SAA) is proposed to optimize the CMGNN for precisely classifying the intrusion detection. The proposed CMGNN‐SAA‐IoT‐ID is implemented, and performance metrics such as accuracy, recall, precision, computational time, and detection rate are analyzed. Finally, the performance of the CMGNN‐SAA‐IoT‐ID provides 26.68%, 25.75%, and 26.16% high accuracy and 29.08%, 30.70%, and 16.26% high precision compared to the existing models.
ABSTRACT Wireless Sensor Network‐Internet of Things security in Cyber‐Physical Systems (CPSs) focuses on protecting interconnected sensor networks and devices from cyber threats while ensuring reliable operation of physical processes. However, these systems face challenges such as limited device resources, high vulnerability to attacks, and difficulties in maintaining data integrity and privacy. This research introduces a novel blockchain‐based wireless sensor network and Internet of Things (WSN‐IoT) security framework that makes use of a Directed Acyclic Graph Triplet Attention Network coupled with the Draco Lizard Optimizer (DAGTAN‐DLO) as an optimization technique. It first collects network traffic data from the UNSW‐NB15 and NSL‐KDD datasets, followed by pre‐processing using Adaptive Multiple Imputation of Missing Values with Class Center (AMIMVC), which reduces noise and normalizes features. Then the Empowering Decision Transformer (EDT) is used to extract compact and discriminative features, and the Wonderful Fairy Wren Optimization Algorithm (SFOA) is used to select the most relevant features. Ultimately, these features are classified by Directed Acyclic Graph Triplet Attention Network (DAGTAN) into normal or anomalous traffic, with classification accuracy improved through weight optimization using the Draco Lizard Optimizer (DLO). The encrypted data is then securely stored using Blockchain‐Based Ultra‐Low Storage overhead Practical Byzantine Fault Tolerance (ULS‐PBFT), thus allowing for privacy‐preserving storage with the integration of smart contracts and cloud services. The proposed network traffic monitoring framework not only effectively and accurately detects cyber threats but also manages data securely and efficiently in the WSN‐IoT‐enabled CPS environments. The proposed technology achieves up to 99.89% accuracy, 99.43% F1‐score, and exceptionally high specificity and precision on UNSW‐NB15 and NSL‐KDD datasets, offering a scalable, precise, and blockchain‐secured solution for cyber threat detection in WSN‐IoT‐enabled CPS environments.
ABSTRACT The use of underwater wireless sensor networks (UWSNs) has become a major security concern due to their increased use in environmental monitoring, military surveillance, and offshore applications. The networks are highly susceptible to various cyber threats due to their dynamic topology, limited energy resources, and unreliable communication networks. The existing intrusion detection techniques tend to exhibit low detection rates, high false alarms, and high complexity, making them inapplicable to real‐time and resource‐constrained UWSN systems. Moreover, many conventional and deep learning‐based models cannot effectively model feature interactions and emerging attack patterns in heterogeneous IoTs and underwater environments. To address these weaknesses, this paper presents a new hybrid intrusion detection model, which integrates CoDeV based feature extraction algorithm and the HyPerion‐T model. To capture both the spatial and temporal characteristics of network traffic and leverage a transformer‐based structure to enhance intrusion detection by exploiting adaptive attention, the article presented here was based on sophisticated feature description and dependency modeling. These combinations will ensure greater accuracy, strength, and efficiency in the calculations. The results of the proposed methodology are tested on the benchmark datasets, i.e., WSN‐DS, IoT‐23, and N‐BaIoT. Among all models, CoDeV+HyPerion‐T has reached the highest average accuracy (98.96%) with an accuracy of more than 98% and recall of more than 98% and also achieved the lowest average inference time (9.5 ms), demonstrating great real‐time performance. The inference energy is also measured to be 3.42 mJ for per inference, achieving up to 50.1% energy reduction compared with basis approach models.
ABSTRACT Intelligent reflecting surfaces (IRS) have emerged as a promising solution to enhance signal propagation and mitigate interference in next‐generation wireless networks. The purpose of this study is to develop and evaluate an optimized beam routing framework for IRS‐assisted wireless communication in complex urban environments. To achieve this, this study proposes an IRS‐assisted beam routing framework that leverages metaheuristic optimization to determine optimal signal paths in complex urban environments. The key contribution lies in the design of a server‐controlled IRS network that dynamically adjusts phase shifts based on environmental factors and user locations, enabling adaptive and efficient signal routing. A server‐controlled IRS network dynamically adjusts phase shifts to ensure efficient transmission, overcoming obstacles and improving network coverage. Extensive simulation results show that the proposed method significantly enhances signal reliability, reduces path loss, and outperforms existing IRS‐based approaches in terms of path optimization and communication efficiency. Specifically, the proposed framework achieves a 40% improvement in path efficiency compared to prior state‐of‐the‐art methods, as confirmed through comparative analysis. Comparative analysis with prior research confirms a 40% improvement in path efficiency, validating the system's effectiveness. These findings highlight the potential of IRS‐based beamforming as a scalable and robust solution for sixth‐generation (6G) wireless networks and beyond. Future work will focus on real‐world validation and the integration of machine learning techniques for adaptive IRS configurations.
ABSTRACT The need for efficient routing arises in Internet of Things (IoT)–based networks that comprise low‐power and lossy nodes. Routing Protocol for Low‐Power and Lossy Networks (RPL) has multiple limitations, including rank manipulation attacks, inefficient energy utilization, poor link quality, and vulnerability to malicious nodes. Moreover, conventional routing metrics typically evaluate only a limited set of parameters and are unable to capture the multiple risks and dynamic behaviors encountered in practical IoT environments. To overcome these limitations, this research proposes a Blockchain‐enabled Risk‐Aware RPL (Bc‐RARPL) framework that enhances routing security, adaptability, and performance. The proposed framework establishes a unified routing architecture by integrating real‐time rank verification, Multi‐Attribute Utility Theory (MAUT)–based risk assessment, trust‐aware consensus, and blockchain‐based route validation into a continuous routing decision process. Each node evaluates candidate parent nodes using a composite risk utility score derived from multiple factors, including residual energy, link quality, latency, and node failure probability, thereby enabling secure and intelligent parent selection. Routing path optimization is performed using a risk‐aware decision‐making strategy supported by the Starfish Optimization Algorithm, which jointly minimizes energy consumption, improves route stability, and mitigates dynamic network risks. In addition, a lightweight Zigbee blockchain securely stores routing histories, node rankings, and risk values, whereas a Trust‐Weighted Byzantine Fault Tolerance (TW‐BFT) consensus protocol restricts validation to trusted nodes, thereby reducing the impact of malicious attacks. Experimental results demonstrate that the proposed framework achieves a packet delivery ratio of up to 96%, reduces end‐to‐end delay and energy consumption, and improves routing reliability while providing enhanced resilience against rank manipulation and insider attacks.
ABSTRACT Vehicular ad hoc networks (VANETs) are essential components of intelligent transportation systems that facilitate real‐time communication between vehicles (V2V) and between vehicles and infrastructure (V2I). Despite their importance, VANETs face challenges, such as high node mobility, energy limitations, security risks, and ever‐changing network topologies. Existing clustering and routing algorithms often struggle to manage the instability caused by mobility, energy disparities, and secure congestion‐free communication simultaneously. To address these challenges, this work introduced an integrated cross‐layer framework featuring three innovative algorithms: mobility‐aware black hole clustering (M‐BHC), energy‐aware piranha optimization algorithm (EPOA), and cross‐layer multi‐attribute blockchain routing with congestion control (CL‐MABRC). The M‐BHC algorithm enhances the stability of clusters and counters black‐hole attacks by forming clusters dynamically based on real‐time vehicle mobility patterns. EPOA optimizes the selection of cluster heads (CHs) by reducing energy consumption through a bio‐inspired resource allocation strategy modeled on piranha predation behavior. CL‐MABRC addresses network congestion and security using blockchain‐based verification and cross‐layer routing decisions informed by multi‐attribute metrics. Extensive simulations were conducted with a setting of 100 veh/km 2 . The proposed framework showed significant performance improvements over benchmark protocols, such as optimal security‐aware cluster‐based hybrid geographical and opportunistic routing (OSC‐GOR), enhanced location‐aided ant colony routing (ELAACR), trust‐based multi‐objective honey badger algorithm (TMOHBA), and robust cryptographic scheme for reliable data communication (RCSRC). It achieved a throughput of 99.89 Kbps, end‐to‐end delay of 3.9 ms, collision rate of 21.8%, energy consumption of 41.98%, and jitter of 0.05 ms. Together, the M‐BHC, EPOA, and CL‐MABRC algorithms create a robust, energy‐efficient, and secure communication framework for VANETs, enhancing scalability, reliability, and real‐time performance in transportation systems.
ABSTRACT A wideband coaxial‐fed ring dielectric resonator antenna (DRA) design is presented, offering uniform high gain over its wide operating bandwidth. The proposed antenna uses a simple, compact, and cost‐effective coaxial edge‐feeding to achieve a wideband broadside radiation pattern. A comprehensive theoretical analysis is carried out to investigate the role of the substrate in bandwidth enhancement by accurately estimating the resonating frequencies of the ring DRA. To improve the antenna gain, a compact metallic wall is introduced using a small antenna footprint. Experimental results demonstrate an impedance bandwidth of 49.5% (11.55–19.15 GHz). The broadside gain consistently exceeds 15 dBi over the band, reaching a peak value of 15.65 dBi. The proposed antenna exhibits a highly compact profile (2.5λ × 2.5λ × 0.7λ) while simultaneously offering high gain and wide bandwidth. In addition, the antenna exhibits excellent polarization purity, with cross‐polarization isolation better than 45 and 35 dB in the E‐ and H‐plane, respectively. Owing to its compact size, wide bandwidth, and high gain features, the proposed ring DRA is well suited for Ku‐band high‐gain applications and can be extended to higher frequencies.
ABSTRACT A two‐port dielectric antenna for mm‐wave frequency range has been constructed and examined in this study. The alumina ceramic resonator is excited by the structure via aperture coupling. The design's originality is emphasized by three salient features: (a) Circularly polarized waves may be produced in the working range of 31.8–32.45 GHz using polarization convertor suspension; (b) 1 × 2 array formation improves measured gain close to 9.5 dBi; and (c) antiparallel placement of antenna terminals/SNG MS suspension reduces the coupling level to below −35 dB. An effective working range of 31.45–33.25 GHz is confirmed by experimental validation. Broadsided radiation is guaranteed by the excitation of the basic hybrid HEM 11δ mode. Due to stable radiation feature and diversity parameters, the suggested antenna is applicable for the lower portion of the mm‐wave spectrum and future FR2 extensions.
ABSTRACT Reconfigurable intelligent surfaces (RIS) represent a transformative technology for enhancing coverage and energy efficiency in sixth‐generation (6G) wireless networks. However, the substantial pilot overhead required to estimate cascaded channels remains a critical bottleneck, particularly in low signal‐to‐noise ratio (SNR) environments where conventional estimation techniques degrade significantly. This paper proposes a robust two‐timescale channel estimation framework that decomposes the cascaded channel into a quasi‐static base station (BS)‐to‐RIS link estimated via coordinate descent (CD) and a rapidly varying RIS‐to‐User Equipment (UE) link estimated via a bias‐aware minimum mean square error (MMSE) filter. We prove rigorously, via a majorization‐minimization (MM) surrogate construction, that each CD coordinate subproblem is strictly convex under the bilinear observation model arising from hardware phase noise and mutual coupling, guaranteeing monotone cost reduction and convergence to a stationary point. The proposed MMSE estimator explicitly absorbs the residual error covariance of the CD estimate into a Woodbury‐structured filter, formally distinguishing it from standard Wiener filtering and eliminating the high‐SNR bias floor present in prior art. We provide a comprehensive comparison with recent works (2020–2024), including least squares (LS), minimum variance unbiased (MVU), orthogonal matching pursuit (OMP)‐based compressive sensing (CS), and a deep learning (DL) benchmark, under identical hardware‐realistic conditions: 2‐bit phase quantization, dBc phase noise, and dB mutual coupling. Ablation studies isolate the individual contributions of the CD and MMSE components. Sensitivity analyses with respect to RIS size and phase quantization bits are reported. We further provide a detailed discussion of implementation complexity, real‐time feasibility on modern FPGAs and ASICs, and hardware deployment challenges to support practical adoption. The results confirm that CD‐MMSE maintains superior normalized mean square error (NMSE) performance while drastically reducing pilot overhead, offering a scalable, spectrally efficient solution for massive multiple‐input multiple‐output (MIMO) and Internet of Things (IoT) deployments in future 6G networks.
ABSTRACT Free‐space optical (FSO) communication systems face significant performance degradation due to atmospheric turbulence and pointing errors, which reduce spectral efficiency, limit data rates, and affect link reliability. This work proposes an Ocotillo‐optimized modulated FSO (OM‐FSO) system that employs the Málaga distribution to accurately represent turbulence and integrates the Ocotillo optimization algorithm (OcOA) with a low‐intermediate frequency (low‐IF) receiver to enhance system performance. The proposed design improves spectral efficiency, increases data transmission rates, and reduces bit error rate (BER) while maintaining robustness under severe atmospheric variations. Simulation results show that for a 20‐km link, the system achieves a reduction in required transmission power of 20.4 dB under clear conditions and 20.53 dB under strong turbulence at a BER of 10 –7.2 . Furthermore, it attains a spectral efficiency of 154 bits/s/Hz and a data rate exceeding 870 Gbps. These results demonstrate that the optimized OM‐FSO system offers a resilient and efficient solution for long‐distance FSO communication.
ABSTRACT The rapid integration of Internet of Things (IoT) in the healthcare domain has led to the emergence of the Internet of Medical Things (IoMT), which introduces significant benefits in patient monitoring and real‐time medical services. However, IoMT networks are inherently vulnerable due to resource constraints, heterogeneous devices, and sensitivity of medical data. In this paper, we propose a novel federated learning‐based anomaly detection system (Fed‐ADS) designed specifically for IoMT networks. Our system leverages local training of lightweight ML models on resource‐constrained IoMT devices and employs secure model aggregation at the gateway to preserve privacy and avoid centralized data collection. To address real‐world challenges, we implement and evaluate our system on a real IoMT testbed using Raspberry Pi devices under various attack scenarios. Furthermore, we examine the impact of privacy‐preserving techniques such as differential privacy on detection accuracy and system overhead. The runtime evaluation shows that our approach achieves high detection accuracy (over 94%) with minimal CPU and memory usage (under 3%), making it suitable for practical deployment in medical environments.
ABSTRACT Secure resource allocation in future sixth‐generation (6G) networks is a significant challenge due to the increasing demand for efficient spectrum utilization and advanced multimedia services. Emerging 6G technologies, including cognitive radios (CRs) for dynamic spectrum access, hybrid multiple access (H‐MA) that integrates orthogonal multiple access (OMA) and non‐orthogonal multiple access (NO‐MA), and clustering techniques for efficient user grouping, offer promising solutions to address these challenges. The collective integration of these technologies within cognitive radio networks (CRNs) is referred to as a 6G CRN framework. In this paper, a novel cluster‐assisted CR‐enabled downlink hybrid multiple access (CCRDHMA) scheme is proposed in the presence of eavesdroppers to maximize the sum secrecy rate (SSR). The formulated optimization problem incorporates minimum quality of service (QoS) requirements while satisfying false detection (FD) and missed detection (MD) constraints. To efficiently solve the resulting mixed‐integer nonlinear programming (MINLP) optimization problem, a low‐complexity ‐optimal outer approximation algorithm (OAA) is employed. The performance of the proposed CCRDHMA scheme is evaluated against conventional OMA‐assisted CRN, NO‐MA‐assisted CRN, and other existing secure resource allocation approaches. Extensive simulation results demonstrate that the proposed CCRDHMA framework significantly improves SSR while achieving superior performance in Key Performance Measures (KPMs), including secondary mobile handsets (SMHs) admission within clusters, association with secondary tower (ST), QoS satisfaction, fair power allocation (PA), FD, and MD. Furthermore, the ‐optimal OAA achieves near‐optimal solutions with , highlighting its computational efficiency and practical applicability for secure resource allocation in future 6G CRNs.
ABSTRACT Internet‐of‐Things (IoT) and wearable devices require compact, wideband, and efficient antennas that can be used in close proximity to the human body. Traditional ultra‐wideband (UWB) antennas are seriously affected by performance degradation in the vicinity of human tissues because of detuning, low gain, and high specific absorption rate (SAR). In order to address these drawbacks, this paper introduces a miniaturized UWB antenna with a metamaterial‐inspired multi‐resonant artificial magnetic conductor (AMC). The antenna is fabricated on an FR4 substrate and optimized through careful parametric analysis in order to attain broadband operation between 3.2 and 9.7 GHz. A novel UWB antenna and AMC unit cell is modeled as an equivalent LC circuit that offers an in‐phase reflection at three resonant frequencies. The UWB antenna, with the support of the AMC, shows a high gain boost of about 3 dB, increasing from 1.9 to 4.7 dB at 5 GHz and 1.5 to 3.5 dB at 8 GHz. At 5 GHz, the standalone antenna is 73.5% radiation efficient, and at 8 GHz, the radiation efficiency is 76.0%. With the addition of the AMC, the efficiency decreases to 65.3% and 75.0% respectively, because of the extra dielectric losses in the AMC substrate, although the realized gain increases by about 3 dB because of increased forward directivity. Multilayer human phantom SAR assessment shows a significant decrease compared with that of standalone antennas, with 1.57 W/kg at 5 GHz and 0.163 W/kg at 8 GHz measured, which meets both FCC and ICNIRP safety limits. The proposed design has been thoroughly compared against the AMC‐based wearable antennas recently reported, which allows the conclusion that the given solution reaches a higher gain‐to‐size ratio and safety level, making it one of the most promising solutions to be implemented in the next‐generation wearable and IoT applications.
ABSTRACT The wireless sensor networks (WSNs) that exist in large and dynamic systems demand routing systems that are able to make real‐time decisions with stringent hardware limitations. Although several deep learning (DL)–based routing algorithms have been proposed, their high computational complexity and software‐based inference limit their deployment in low‐power FPGA‐based WSNs. To overcome these challenges, a hybrid multistage routing and prediction framework integrating Ad hoc On‐Demand Distance Vector–Multipoint Relay (AODV‐MPR) routing, Improved Chaotic Brownian Kite Optimizer (ICBKO), and AtSCS‐Net based on Chaotic Hybrid Crab‐Starfish Optimization (CHCSO) is proposed. A minimum redundancy maximum relevance (MRMR) feature selection is used to approximate the most informative routing parameters, which results in reducing computational complexity. Experimental results demonstrate the effectiveness of the proposed framework, achieving a 99.76% success rate, 98.95% positive predictive value (PPV), 0.10 mean square error (MSE), and 0.5 mean absolute error (MAE), thereby outperforming existing methods in routing accuracy and error reduction.
ABSTRACT Compact and wider bandwidth designs of proximity fed circular and equilateral triangular microstrip antennas backed by rectangular slots cut ground plane are proposed. Slot on the ground plane reduces the fundamental mode frequency on each of the patches that yield wideband characteristics in a lower frequency band. Among the proposed designs, circular patch and its multi‐resonator variations employing slots cut ground plane achieve optimum results. The single circular patch achieves 22.5% reduction in fundamental mode frequency, 47.66% reduction in patch area, 0.027λ g11 reduction in thickness of the substrate supported with 10% increase in reflection coefficient bandwidth. Gap‐coupled circular patch designs achieve 10% reduction in the start frequency of the bandwidth, more than 30% reduction in the total patch area, with more than 15% increase in the bandwidth. For all these improvements, slot cut ground plane antenna exhibits only 1 dBi reduction in the peak broadside gain. Considering all the antenna parameters together, slots cut ground plane present substantial improvement, so as to achieve a compact wideband solution. Design methodology for circular patch employing slot cut ground plane is presented. It is useful in realizing similar configuration as per specific frequency band of wireless application. Measurements in the proposed designs have been carried out that show close agreement against the simulation.
One of the most well-known routing systems is cluster-based routing, where the head node collects information from every other cluster node, performs specific aggregation functions, and afterwards delivers the aggregated information towards the base station. Motivated by that, we have developed a novel cluster-based routing approach that is comprised of two working phases which are optimal cluster head selection as well as routing. Cluster-based routing reduces the energy consumption of sensor nodes by organizing them into clusters, allowing data to be aggregated before transmission, which minimizes redundant data communication. The suggested approach allows the network to dynamically react to shifting conditions, such as node failures, mobility, and changes in ambient parameters. In this first phase, the suitable cluster head is selected using the hybridized optimization named customized chimp with blue monkey optimization (CCBMO), by considering the parameters like energy, distance as well as delay. After the selection of the optimal cluster head, routing is done which uses factors such as packet delivery ratio (PDR) and security for the optimal routing process.
Drones, also referred to as unmanned aerial vehicles (UAVs), have evolved into an effective solution for autonomous detection and data collection in indoor spaces without the need for human involvement. It was more suitable for crucial applications, but there are several impacts faced in indoor surveillance, like battery power constraints, unreliable communication, and poor maintenance of coverage across large indoor areas. In order to improve the performance of indoor surveillance, several AI-based approaches are used for detecting objects and ensuring drone pathing according to the targeted object in indoor environments with obstacle avoidance. However, these methods include the inability to accommodate very dynamic indoor environments, where they do not have the ability to find unexpected occurrences of obstacles in real-time and the lack of reliability in low light or sensor-degradation scenarios, which limit the resilience of indoor drone surveillance systems. To overcome these issues, a lightweight UAV detection-You Only Look Once based object detection with a deep reinforcement learning method was implemented for re-planning the drone's path to avoid collision. Initially, the drone was designed with no GPS for indoor applications. The drone contains two sensors, namely, stereo cameras and light detection and ranging. The stereo cameras are used for capturing images of indoor environments. The LIDAR generates a 3D point cloud of the indoor environments, and it shows the objects' distances and shapes for exact mapping. Here, the hybrid AI model was implemented for object detection and rephasing the drone. The lightweight UAV detection-You Only Look Once (LUD-YOLO [C3K2]) detects objects based on the stereo cameras' images. After that, semantically driven deep reinforcement learning (SD-DRL) was used for re-pathing the drone based on LIDAR distance measuring if the obstacles are detected in the moving path of the drone. Finally, the proportional-integral-derivative controller adjusted the drone's motion through its pitch, yaw, thrust, and roll for stable collision-free navigation in indoor environmental monitoring. The proposed LUD-YOLO (C3K2) model achieves accuracy, precision, recall, and specificity values of 97.01%, 89.72%, 84.93%, and 88.67%. Additionally, the proposed reinforcement learning-based deep deterministic policy gradient model attains an average reward comparison value of 320, a root mean squared error value of 1.2, and a success rate value of 86%. The proposed model attains accurate object detection and efficient re-pathing without collisions, ensuring reliable surveillance in indoor environments.
Intrusion detection systems (IDSs) for sixth generation (6G) internet of things (IoT) networks often suffer from low adaptability, high false alarm rates, and limited ability to handle complex and evolving traffic. Traditional deep learning models struggle to capture spatial and temporal attack patterns, and gradient-based optimizers often get stuck in local minima, reducing detection accuracy. To overcome these challenges, this work proposes an optimized cross-hierarchical structure-detail-aware cascaded capsule neural network (CS2C2N-GBFIO) for efficient and responsible attack detection in responsible artificial intelligence (AI)-enabled 6G-IoT networks. Initially, real-world IoT traffic datasets are collected. The Quality-aware fuzzy min-max neural network (QF2MNet) is used to ensure clean, normalized, and multiview feature representation. Subsequently, the Hybrid Divine Religions and Catch Fish Optimization (DRA-CFOA) algorithm is used to select the features that are the most discriminative and nonredundant. This is followed by the cross-hierarchical structure-detail-aware cascaded capsule neural network (CS2C2N) that learns both structural and fine-grained features of traffic behavior and Grizzly Bear Fat Increase Optimization (GBFIO) that learns hyperparameters to obtain optimal convergence. The experiments of Real-Time Internet of Things 2022 (RT-IoT2022) and Canadian Institute for Cybersecurity Internet of Things 2023 (CICIoT2023) datasets demonstrate superior performance, achieving 99.36% accuracy, 99.43% F1 score, and a false alarm rate (FAR) of 1.30%. The low computational cost, scalability of the model, and adaptability in real-time make it appropriate to securely implement the model at fog and edge layers of contemporary 6G-IoT infrastructures.
Software-defined wireless networks (SDWNs) have emerged as a groundbreaking approach to managing advanced network environments, enabling centralized control and dynamic traffic management to deliver optimal performance. The highly advanced distributed denial-of-service (DDoS) attacks and growing network demands are signs of constant scalability, latency control, and effective routing issues during high load. Given these difficulties, this paper proposes a new system to address the shortcomings of traditional approaches. QCD-Net (Quokka-optimized Clifford-steerable dynamic graph network) for wireless SDN traffic control proposes a new comprehensive approach that starts with adequate data collection from the DDOS attack SDN dataset and stringent preprocessing using min-max Z-score normalization to maintain homogeneity. The most important characteristics are acquired using the uncertainty-aware decision transformer, which identifies the major traffic behaviors and uncertainties. The inputs are fed into a software-defined networking (SDN) controller with a traffic-predicting unit based on a Clifford-steerable dynamic graph attention network, encompassing both spatial and temporal network behavior. Routing decisions are optimized by the traffic routing unit using Quokka swarm optimization for adaptive load balancing with low latency. The performance of experimental testing is excellent, with a prediction accuracy of 99.5, routing efficiency of 98.7, detection rates of 99.2, and overall network performance improvements of 99.0. In general, the QCD-Net framework should be regarded as a significant improvement in network resilience, scale, and efficiency in a dynamic, attack-susceptible environment, paving the way for further improvements in wireless SDN traffic management and laying the foundations for future adaptive network management solutions.
