International Journal of Communication Systems最新文献

In this paper, a combination model based on complementary ensemble empirical mode decomposition (CEEMD) is proposed. First, CEEMD is applied to decompose original network traffic to generate high-frequency component, low-frequency component, and residual component. Then, the high-frequency components are modeled and predicted using bi-directional long short-term memory (BiLSTM). The low-frequency components and the residual component are modeled and predicted using autoregressive integrated moving average (ARIMA). Meanwhile, considering that the BiLSTM model is influenced by the hyperparameters, an Improved Bald Eagle Search (IBES) algorithm is proposed and applied to optimize three hyperparameters of BiLSTM, avoiding the blindness and subjectivity of manual selection of parameters. Finally, the prediction values of BiLSTM and ARIMA model are summed to obtain the final predicted value of network traffic. The comparisons with other models proved that the proposed network traffic prediction model is closer to the real data, with the optimal performance indicators, which is very suitable for high precision occasions.

We investigate an iterative sparse interference cancelation (ISIC) algorithm in massive multiple input multiple output (MIMO) uplink systems, which includes a multilayer implementation consisting of a channel sparse estimation layer using an improved Sparsity Adaptive Matching Pursuit (SAMP) algorithm, sorting layer and filtering layer with noise power threshold. The theoretical bound for noise power threshold is also addressed. To optimize sparse interference cancelation, we analyze its feasibility and robustness with an iterative scheme detecting the symbols sequentially and eliminating interference from all other users at different multiuser access conditions. Additionally, we provide theoretical proof for iteration termination condition. Analysis and simulation also demonstrate the performance of our proposed sparse interference cancelation approach ideal maximum likelihood (ML) detection under different multiuser access conditions.

With the rapid advancement of the Internet, indoor localization technology has gained increasing importance across various fields. However, the complexity of indoor environments presents significant challenges for achieving precise positioning using GPS or BeiDou systems. As a result, there is a growing demand for innovative localization methods that deliver high accuracy, improved security, and cost-effectiveness. In this study, a dataset comprising 9291 fingerprints collected from a building was processed and split into training and test sets in a 7:3 ratio. To facilitate feature extraction, four algorithms—UMAP, LDA, PCA, and SVD—were employed. Subsequently, six machine learning models (KNN, Random Forest, ANN, SVM, GBDT, and XgBoost) were trained on the training set and evaluated on the test set to compare their performance with different feature extraction algorithms. The objective was to identify the most effective feature extraction method. Model performance was assessed using three metrics: average error, coefficient of determination, and accuracy. Finally, a stacking ensemble model was developed, incorporating the six models as primary learners and selecting the five models with superior predictive performance as secondary learners. This approach aimed to enhance the localization accuracy. UMAP feature extraction significantly improved the prediction accuracy of the indoor localization model, whereas the stacking ensemble model, combining KNN, GBDT, XgBoost, ANN, Random Forest, and SVM as primary learners and Random Forest as the secondary learner, achieved the highest localization accuracy, with an error of approximately 1.48 m.

The predictive analysis of mobile network traffic is important for future generation cellular networks. Knowing user requests in advance enables the system to allocate resources in the best way possible. In this manuscript, Real-Time Mobile Data Traffic and Noise monitoring System for AI Data Prediction Using open Source Frame Work (RMTNMS-OSF) is proposed. Unlike previous studies that primarily remained theoretical, this research aims to identify areas with the highest demand for 5G internet service and also promptly provide the information to IT professionals. This is significant because of the high demand for internet services among tech professionals working from home in rural areas. This developed software now utilizes HTML, OpenLayers, and real-time spatial location data along with the Google Satellite Map API as its base layer to detect user locations as well as to ensure uninterrupted high-speed internet service. The innovation of this proposed RMTNMS-OSF model lies in the integration of AI-driven predictive models with real-time geospatial data processing to optimize network performance in rural areas by dynamically predicting network demand, detecting congestion, and preventing data loss using cost-effective open-source technology, and this mark up a significant advancement in mobile network traffic prediction and resource allocation. The performance of the proposed RMTNMS-OSF method is evaluated with existing methods.

Software-defined networks (SDN) provide an efficient network architecture by enhancing global network monitoring and performance through the separation of the control plane from the data plane. In extensive SDN implementations for the Internet-of-Things (IoT), achieving high scalability and reducing controller load necessitates deploying multiple distributed controllers that collaboratively manage the network. Each controller oversees a subset of switches and gathers information about these switches and their interconnections, which can lead to imbalances in link and controller loads. Addressing these imbalances is crucial for improving quality of service (QoS) in SDN-enabled Industrial Internet-of-Things (IIoT) environments. In this paper, we present the NP-hardness of the link and controller load balancing routing (LCLBR) problem within IIoT. To tackle this issue, we propose an enhanced flow control and load balancing approach for SDN-enabled Industrial Internet-of-Things (EFLB-IIoT). EFLB-IIoT is an approximation-based technique that effectively maintains network activity among distributed controllers. Simulation results indicate that our proposed strategy reduces the maximum link load by 76% and the maximum controller response time by 85% compared to existing techniques, demonstrating superior performance over state-of-the-art methods.

Wireless sensor networks (WSNs) are considered a key foundation for high-level Internet of Things (IoT) practices. Moreover, WSNs depend on transmission for data transfer among sink modules and sensor nodes or intermediate points in the system. However, in WSN, there are several interruptions during the transmission of data. Further, storage space, network bandwidth, and processing power are limited, and hence, it is significant to enhance data distribution to improve network performance. Therefore, this paper devised a new approach known as the Flamingo Lyrebird Optimization Algorithm (FLOA) to improve radio frequency identification (RFID)-WSN integrated network lifetime. Firstly, the network topology of WSN-RFID is simulated, and then cluster head (CH) selection is performed by FLOA in terms of multiobjective fitness, namely, energy, network lifetime, and inter- and intracluster distance. Here, FLOA is formed by integrating Flamingo Search Optimization (FSA) and Lyrebird Optimization Algorithm (LOA). After this, the energy-efficient multipath routing is performed by utilizing FLOA, where link life time (LLT) and energy are predicted using a gated recurrent unit (GRU). Furthermore, FLOA attained the maximum performance with network lifetime and energy of 820.11, 0.923 J, and minimum delay of 0.211 s.

Data dissemination is a promising use in Vehicular Ad hoc NETworks (VANETs), where messages are jointly carried and delivered by vehicles toward their destinations at defined points. Vehicles find it challenging to select the target uplink node among the roadside units (RSU) put in VANETs due to the RSU coverage, traffic intensity, and other sophisticated and dynamic affecting factors. In this study, a new adaptive vehicle clustering method is proposed that attempts to reduce the power consumption of automobiles. It dynamically distributes the computational assets of every virtual machine within the automobile. The mimosa pudica clustering optimization algorithm (MPCOA) is used to identify the ideal clustering number to reduce the overall energy consumption of the cars, and the clustering head is chosen based on the direction the vehicles are going, their weighted mobility, and their entropy. A cluster head (CH) oversees all intercluster and intracluster communication. Some factors to gauge the effectiveness of a network include the load on every CH, the lifespan of the cluster, and the overall number of clusters in the network. The new cluster is well suited for this type of huge data, in which they are separated into similarity, variations, and neighborhoods as different types of VANET features along with their combinational groups. Using a test bed and numerous simulations, the suggested system efficacy is assessed. Conducting a performance study and comparing the test bed findings to the simulation results provides a thorough understanding of the performance and viability of the proposed MPCOA-based system.

Multiple-input multiple-output (MIMO) technology plays a crucial role in the field of wireless communications. As one of the key technologies, MIMO signal detection ensures the reliability of communication systems and achieves high throughput transmission. In this paper, to find a high-performance and low-complexity detection algorithm, a hybrid detection algorithm is proposed based on the K-best detection algorithm. The hybrid algorithm employs the minimum mean square error (MMSE) linear detection algorithm combined with sorted QR decomposition (SQRD) for preprocessing, followed by signal detection using the K-best detection algorithm. Compared with the traditional K-best detection algorithm, the proposed method shows significant performance improvement. To address the computational complexity issue caused by the fixed K value in each layer of the hybrid detection algorithm, an adaptive threshold algorithm is introduced to select an appropriate K value for each layer, significantly reducing the algorithm's complexity. On the hardware implementation level, not only is the overall algorithm architecture optimized but a lookup table (LUT) based sorting algorithm is also proposed to address the sorting delay issue in the hybrid detection algorithm. Comprehensive analysis shows that this detector, implemented in a 28-nm process, achieves a throughput of 4.8 Gbps at a clock frequency of 769 MHz, presenting a significant advantage compared with other literature.

In this manuscript, an Advanced Estimation and Feedback of Wireless Channels State Information for sixth generation (6G) Communication via Recurrent Conditional Wasserstein Generative Adversarial Network (AEF-WCSI-6G-RCWGAN) is proposed. Deep Learning (DL) based channel estimation algorithm using Recurrent Conditional Wasserstein Generative Adversarial Network (RCWGAN) is estimated the channel parameters in 6G, such as channel gains and delays from received signals, which is crucial for effective communication and resource allocation. The primary purpose of this paper is to discuss key issues and possible solutions in DL-based wireless channel estimation and channel state information (CSI) feedback including the DL model selection, training data acquisition and neural network design for 6G. The deep learning-dependent channel estimator refines the predicted channel output, which is subsequently used for increase the efficacy and dependability of the communication scheme. The proposed AEF-WCSI-6G-RCWGAN is implemented and the performance metrics, like Detection Success Probability, Mean Square Error (MSE), and Normalized Mean Square Error (NMSE) are analyzed. Finally, the performance of the proposed AEF-WCSI-6G-RCWGAN method achieves 30.73%, 28.35%, and 29.62% higher Detection Success Probability, 25.73%, 28.05%, and 24.62% lower MSE when compared with existing methods: towards DL-assisted wireless channel estimate and CSI feedback for sixth generation (WCE-CSI-6G-GAN), an effectual deep neural network channel state estimate for Orthogonal frequency-division multiplexing (OFDM)wireless systems (CSE-WS-BiLSTM), and distributed machine learning dependent downlink channel estimate for reconfigurable intelligent surfaces supported wireless communications (DCE-AWC-HDCENet) methods, respectively.

Packet routing is one of the critical factors that affect network performance and security, with the goal of finding the optimal path for network packets from the source node to the destination node. However, with the diversification of network architectures, the differences in network application requirements, and the time-varying characteristics of network topologies, the limitations of traditional model- and rule-based routing algorithms in terms of computational overhead and flexibility are becoming increasingly apparent. This paper designs a packet routing strategy based on multiagent deep reinforcement learning (MIMRL). In MIMRL, each router node is abstracted as an independent agent with its own neural network. Multidimensional data such as the current location of the data packet, the number of nodes in the network, the length of the data packet received at the current location node, and the set of neighboring nodes are used as inputs to the neural network. Combined with a segmented reward function, the optimal routing action is determined. Experimental results under different network loads in static and dynamic networks show that the MIMRL algorithm significantly outperforms the benchmark algorithm in multiple metrics such as average delivery time and proportion of full capacity nodes.