International Journal of Communication Systems最新文献

Visual sensor network (VSN) captures and transmits the visual data from various locations in real-time monitoring for context aware decision making. Visual sensors have characteristics that make them interesting as sources of information for any process. However, sensors are endowed with low-power cameras for visual monitoring, and it may also impact the outcome that a visual application delivers to the user, such as spatial coverage, lifetime, and dependability. Additionally, due to the restricted resources available to sensor nodes, meeting both coverage and network lifespan criteria may be impossible. In order to address these drawbacks, we developed a hole detection based on grid-based triangle approach and optimal link state routing algorithm for efficient data transfer in VSN. Initially, the node is deployed and captures the visual information across the environmental area. After node deployment, the network undergoes hole detection using grid-based triangle approach–starfish optimization algorithm (GT-SFOA) to identify the uncovered regions. Then, graph learning–based spatial–temporal graph convolutional neural network (GLSTGCN) is used to overlap detection of sensor nodes based on the similarity calculation. Subsequently, the hole recovery mechanism using optimal localizable k-coverage (OLKC) algorithm for reroute similar sensing nodes to cover the holes. Next, low latency optimal link-state routing (LL-OLSR) algorithm is used to transmit the sensor data to the base station effectively. The proposed approach achieves a hole discovery time of 3.3 ms for 100 nodes, an average coverage degree of 81.8%, a packet loss of 2% for 100 nodes, and a routing overhead of 96.49%. The proposed approach improves the coverage accuracy in dynamic environments to enhance the reliability of the visual sensor network.

Software-defined networking (SDN) is revolutionizing network architecture by offering incredible scalability, flexibility, and programmability. SDN's applications have expanded into various domains, including cloud computing, edge computing, enterprise networks, data centers, and the Internet of Things. One of its standout features is its ability to enhance traffic engineering. Traffic engineering involves optimizing and dynamically controlling traffic flows to reduce congestion and minimize the maximum link utilization, all while managing network traffic from a centralized point. SDN boosts overall network performance, improves user experience, and ensures that resources are used effectively. Minimizing maximum link utilization means distributing network traffic across different links or paths to make use of the most available resources and avoid congestion. This process is often referred to as load balancing or congestion minimization within the network. The primary aim is to prevent any single link from becoming overloaded. Over the years, various algorithms have been proposed to tackle this issue, but still such algorithms struggle with effectively minimizing the maximum link utilization. The proposed work achieves approximately 24% and 16% improvement in load balancing; it reduces congestion by 94% and 60% when compared to tabu search and enhanced tabu search.

Social multimedia (SM) platforms such as Instagram, LinkedIn, YouTube, Facebook, and Twitter enable users to communicate, share ideas, and exchange information within virtual networks. The widespread use of multimedia-based applications has led to a substantial increase in social media traffic, often containing sensitive personal and interaction data. These data are vulnerable to security threats, including identity theft and information misuse, highlighting the need for effective anomaly detection mechanisms. This research proposes an enhanced binary spotted hyena optimized dependent linear regression (EBSHO-DLR) method to detect anomalies in SM networks, ensuring secure data transmission. The framework begins with collecting a Twitter dataset from India, followed by pre-processing using Gaussian blur (GB) to filter anomalies and histogram of oriented gradients (HOG) for feature extraction, removing irrelevant information. The proposed method integrates software-defined networking (SDN) to provide dynamic, programmatically efficient network management, improving performance and monitoring compared to conventional approaches. Experimental evaluation demonstrates the effectiveness of the method in detecting fraudulent activities such as identity theft, profile cloning, and unauthorized data collection. Key performance metrics include precision (96.7%) and accuracy (97.8%), confirming the robustness and reliability of the EBSHO-DLR approach for securing social multimedia communication systems.

The rapid development of 5G communication systems has further increased the popularity of wireless personal communication. However, for improving existing wireless personal communication networks, meeting rigorous broadcast rapidity and QoS requirements is an obstacle. Non-orthogonal multiple access and massive multiple-input multiple-output are two emerging technologies with enormous potential. RA at the BS is crucial to ensure fair distribution of services among users. Increasingly, as the number of users goes up, it becomes indispensable for WNs to elevate their capacity to address increasing demands. One of the significant ways to do this is by the integration of MIMO-NOMA. This manuscript proposes a novel deep learning-based resource allocation approach to maximize RA efficiency in MIMO-NOMA. First of all, the users are grouped. For that purpose, a highly correlated K-means clustering technique has been introduced, which helps control interference across and within clusters. Further, the aim is to improve the spectral efficiency, maximize the system capacity, and ensure fairness in resource allocation using a Multi-Granularity Gated Temporal Convolutional Neural Network. Also, the LBO technique has been proposed to optimize the weight parameter of MG-TCNN to improve the SE performance. A decay-based movement adjustment was also incorporated into the LBO to maintain exploration and exploitation in better terms for improved convergence behavior. The proposed framework has been administered and simulated via the Python platform. Several assessment metrics, such as ASR, SE, and EE have been computed and associated with other frameworks. A comprehensive ASR of 88.76 Mbps, SE of 119.40 bits/s/Hz, and EE of 19.43 bits/joules/Hz is obtained by the proposed framework for allocating the resources to the 5G WNs.

The Industrial Internet of Things (IIoT) represents a transformative force in modern industries, enabling unparalleled levels of automation, efficiency, and data-driven decision-making. However, the pervasive connectivity and heterogeneity of IIoT devices introduce significant security challenges, with authentication and access control emerging as critical concerns. This paper proposes a comprehensive authentication scheme specifically tailored for IIoT environments, encompassing system initialization, device registration, mutual authentication, and dynamic device addition phases. At the heart of this scheme lies an innovative approach to key exchange, leveraging an improved elliptic curve cryptography (ECC) model enhanced by a novel self-improved white shark optimization (SI-WSO) algorithm. This inspiration from the acute sensory capabilities of great white sharks, the SIWSO algorithm optimizes key generation efficiency, strengthening the security and reliability of IIoT authentication processes. Through a seamless integration of cryptographic principles and nature-inspired optimization techniques, the proposed scheme aims to fortify IIoT infrastructures, mitigating security threats and fostering trust in industrial deployments. This paper presents a rigorous exploration of the proposed authentication scheme, highlighting its efficacy in enhancing IIoT security and resilience. As IIoT continues to evolve, robust authentication mechanisms play a pivotal role in safeguarding critical assets and sustaining the momentum of digital transformation in industrial settings.

Intrusion detection holds a significant application in identifying malicious packets and maintaining cybersecurity. The multiplex distribution of network traffic and the evolving nature of threats impact the effectiveness of the detection methods. Moreover, the prevailing approaches faced complexities due to the excessive false alarms and feature losses. The research proposes grouped probabilistic update optimized transfer learning with drift-enabled distributed neural network (GPrUO-TD2NN) to address the challenges and to exhibit effective outcomes for diverse network threats. The internal covariance of the data distributions is identified using the drift mechanism, and the generalization of targeted data is enhanced by including transfer learning. In addition, the grouped probabilistic update optimized synthetic minority oversampling (GPrUS2M) approach refines the data by minimizing the imbalances, and neighbor sets are chosen with a definiteness over the attributes of each instance. The grouped probabilistic update optimization (GPrUO) algorithm enhances the selection of better neighbor sets and also assists in parameter tuning of the detection model. Moreover, the effectiveness of the GPrUO-TD2NN in intrusion detection is computed using various metrics that exhibit superior outcomes with higher 96.92% precision, 97.29% recall, 97.16% accuracy, and 97.10% F1 score using 90% training.

The advancement of sixth-generation (6G) wireless communication demands intelligent, adaptive solutions to meet the increasing requirements of ultra-reliable low-latency communication (URLLC), massive connectivity, and terahertz (THz) data rates. Intelligent Reflecting Metasurfaces (IRMs) have emerged as a revolutionary paradigm that reconfigures the wireless propagation environment by intelligently controlling the behavior of electromagnetic waves. However, traditional beamforming methods face significant challenges in dynamically adapting to complex, time-varying wireless environments, especially in the presence of high user mobility and signal obstructions. These limitations hinder the efficiency of signal directionality and energy focusing, resulting in sub-optimal communication performance. To overcome these issues, a novel framework, the Model-Level Multi-Scale Edge-Guided Attention graph neural Network (M-LMSedGAN), is introduced, enabling efficient learning from complex spatial relationships and heterogeneous signal interactions. Optimization of the reflective phase shifts is achieved using the Biruni Earth Radius Optimization (BERO) algorithm, which enhances the global search capability and convergence efficiency of the system. The proposed method enables real-time and context-aware beamforming adjustments, achieving superior signal quality and improved network performance. Experimental evaluations demonstrate a peak beamforming accuracy of 99.9%, indicating robust generalization across dynamic environments and diverse user distributions. This intelligent approach establishes a foundation for deploying energy-efficient, high-performance IRM-assisted 6G communication infrastructures.

This work presents an advanced federated learning (FL) framework that integrates a CNN–GRU + FedTrust model for precise anomaly recognition in resource-constrained wireless sensor networks (WSNs) with unreliable connections. Unlike traditional centralized and heterogeneous FL approaches, the proposed system achieves higher accuracy, efficiency, and robustness by combining lightweight CNN–GRU feature extraction with trust-weighted aggregation through the FedTrust mechanism. The model leverages top-k gradient sparsification to reduce communication overhead and energy-aware client selection (EAC) to optimize energy use, ensuring sustainable network performance. Evaluations on real-world datasets—WUSTL Wireless Sensor Data and Intel Lab Data—demonstrate the model's superiority, achieving 98.4% accuracy, 0.98 F1 score, and faster convergence than Autoencoder-FL, GNN-FL, and hierarchical FL methods. It also exhibits enhanced robustness under client dropout (96.1%) and noise (92.7%) conditions, significantly outperforming existing FL techniques. By efficiently capturing both spatial and temporal patterns while maintaining privacy and energy balance, the CNN–GRU + FedTrust framework delivers reliable and scalable anomaly detection across diverse IoT and smart-industry environments. This hybrid design establishes a new benchmark for energy-efficient, trustworthy, and high-precision FL-based anomaly detection in next-generation WSNs.

Wireless sensor networks (WSNs) are widely used for data collection in environments with limited infrastructure and resource-constrained sensor nodes (SNs). To meet these demands, utilizing multiple wireless channels has emerged as a promising solution to increase network capacity and reduce interference. However, exploiting multiple channels in energy-constrained WSNs introduces significant challenges. Thus, this research introduces an elk skill optimizer (ESO) for dynamic channel allocation (DCA) in WSN. Initially, the CA system model is simulated, and channel assignment is accomplished. The channel assignment with joint power allocation, sampling rate, and transmission rate is performed employing ESO. Finally, channel assignment is conducted utilizing a deep Kronecker network (DKN) to determine the allocated channel at the next time interval. In addition, ESO obtained high performance results compared to existing schemes with maximum values of achievable rate, energy efficiency, network utility, and sum rate about 6.013 Mbits/s, 0.242 Mbits/J, 282.172, and 121.022 Mbits/s as well as minimum value of bit error rate (BER) about 0.776. Moreover, the performance gained by the proposed scheme when considering the metrics achievable rate is 30.191%, 17.354%, 28.032%, 16.520%, 14.215%, and 1.317% higher than the existing schemes used for comparison.

This paper presents an antenna consisting of a bent monopole and a reactive impedance surface (RIS) reflector for smart helmets that can be used by rescue personnel for disaster management. The bent monopole and RIS concepts are introduced to achieve multiband performance with improved radiation characteristics. The bent monopole's resonant modes and the RIS reflector, which operates in its inductive region, are combined to produce hepta-band behavior. Unlike perfect electrical conductors (PECs) and perfect magnetic conductors (PMCs), RIS can operate over a wide frequency range with reflection phases ranging from 0° to 180°. Therefore, the challenges associated with PEC/PMC, such as focused image contribution and low efficiency, can be mitigated using RIS. The phase distribution can be controlled by varying the array size of the RIS. To achieve multiband performance, a RIS of array size of 18 × 21, with periodic square patches, is placed underneath the bent monopole. The designed antenna functions at the GPS L5 (1.176 GHz), GSM (1.8 GHz), LTE (2.6/3.7 GHz), WLAN (5.5/5.8 GHz), and MBAN (7.2 GHz) bands. The addition of the metasurface layer increases the gain of the antenna significantly over all resonant bands. With the RIS layer, the antenna achieves a peak gain of 8.7 dBi. The total efficiency values obtained at the seven resonating bands are 97.21%, 97.35%, 97.49%, 97.7%, 98.12%, 98.33%, and 98.81%, respectively. The monopole and RIS reflector are fabricated separately on the RT/duroid 5880 substrate with permittivity of 2.2. Also, a thick foam (λ₀/8) is sandwiched between the antenna and the RIS reflector to simulate (outer and inner surfaces) a realistic helmet scenario.