Mobile Networks and Applications最新文献

With the rapid development of Internet of Things technology, wearable optical imaging devices can monitor the status and performance of athletes in real time, but the image quality is affected by environmental factors, often resulting in information loss. This study aims to improve the effectiveness of wearable optical imaging devices in sports image analysis by using wireless sensor networks and fuzzy image recovery algorithms, so as to achieve more accurate motion state monitoring. Wireless sensor network architecture combined with mobile network technology is used to realize data acquisition and transmission in motion scenes. In this paper, a fuzzy image recovery algorithm is designed and implemented to process fuzzy image data collected by equipment. In the experiment, the algorithm is trained and verified by using the image data in different motion scenes, and its recovery effect is analyzed. Experiments show that the proposed fuzzy image recovery algorithm can effectively improve the clarity and detail capture of images, and make the status monitoring of athletes more timely and reliable combined with the real-time data transmission of wireless sensor networks. Therefore, wearable optical imaging equipment based on wireless sensor network and fuzzy image recovery algorithm shows a good application prospect in sports image analysis, which can provide important support for athletes’ training and performance evaluation, and promote the intelligent process in the field of sports.

Traditional health monitoring methods rely on wired transmission, which limits the flexibility and real-time data acquisition. Therefore, technology combining optical sensors and wireless mobile networks offers new opportunities for health monitoring. This study aims to explore the application of deep Convolutional neural network (DCNN) algorithm based on optical sensor and wireless mobile network in real-time health monitoring, improve the accuracy and real-time monitoring, and support personalized health management. A monitoring system integrating optical sensor and wireless mobile network is designed. Deep convolutional neural network is used to process the data collected by sensor. The system realizes real-time data transmission through the mobile network, and uploads the user’s physiological data to the cloud for analysis. During the experiment, we conducted a series of tests, including the monitoring of physiological parameters such as heart rate and blood oxygen saturation, and compared it with traditional methods. The experimental results show that the monitoring system based on DCNN has a high identification accuracy in multiple health parameters, and the application of wireless mobile network reduces the data transmission delay to the millisecond level, ensuring the real-time and effectiveness of health monitoring information. In addition, the data acquisition effect of the user in the mobile state is good, which fully demonstrates the portability and convenience of the system.

With the rapid development of the Internet of Things technology, wireless network and mobile network are increasingly widely used in various fields, especially in the evaluation of motion technology, through the real-time acquisition of sensor data, accurate monitoring of motion status can be achieved. However, there are still challenges in the transmission stability and data acquisition accuracy of traditional sensors. The aim of this study is to develop a wireless network data acquisition system based on optical topology sensor to improve the accuracy and real-time performance of motion technology evaluation. Through this system, we hope to achieve efficient monitoring of sports conditions and provide data support for athletes' training and rehabilitation. The research adopts optical topology sensor technology to achieve accurate acquisition of motion data. The sensor transmits data through wireless network and adopts advanced mobile network protocol to ensure the integrity and real-time information in the process of data acquisition. The performance differences between the new system and the traditional sensor in data transmission speed, accuracy and delay are compared and analyzed. The experimental results show that the wireless data acquisition system based on optical topology sensor can improve the data transmission speed significantly, and the system can still work stably in the signal interference environment, which proves its reliability in practical application.

The rapid advancement of mobile network technology has led to an increasing popularity of virtual reality (VR) systems in fashion design. However, this proliferation has also introduced significant network security vulnerabilities. This paper presents a discussion on establishing an effective network intrusion detection system tailored to the unique aspects of mobile networks, aiming to safeguard the security and reliability of VR applications in clothing design. We propose a deep learning-based intrusion detection algorithm that leverages the features of wireless networks and mobile applications to monitor and analyze traffic data in real time. Training and validation datasets are utilized to assess the model's detection performance across various scenarios. Experimental findings indicate that the proposed intrusion detection system can proficiently identify multiple types of network attacks, achieving a high detection rate coupled with a low false positive rate. The system demonstrates strong real-time performance and accuracy, allowing it to adapt to the dynamic nature of mobile network environments. The mobile network-based intrusion detection system holds significant application potential in the realm of VR fashion design, providing a secure and dependable platform for designers.

With the rapid development of Internet of Things (IoT) technology, wireless sensor networks are increasingly used in environmental monitoring and management. In the protection and restoration of lake ecological environment, real-time monitoring of water quality, water temperature and other environmental factors becomes particularly important. The purpose of this study is to explore the application of artificial intelligence image processing technology based on wireless sensor network in lake environment landscape monitoring, in order to improve monitoring efficiency and strengthen environmental protection measures. A network of wireless sensor nodes was constructed to collect data on lake water quality and environment in real time. At the same time, the image processing algorithm and deep learning model are combined to analyze the lake image to identify and evaluate the ecological state. Mobile devices are used for remote access and analysis of data. Through comparative experiments, the data collection method based on wireless sensor network has significantly improved the accuracy and timeliness of data compared with traditional water quality monitoring methods. The results of image processing show that the change trend of lake ecological environment can be quickly identified, and the change of multiple environmental indicators can be successfully predicted. Therefore, the artificial intelligence image processing technology based on wireless sensor network has a broad application prospect in the lake environment landscape monitoring.

The impact of network anomaly on data transmission and system operation cannot be ignored, so an effective anomaly detection method is needed to ensure the stability of the system. This study aims to improve the anomaly detection ability of the cardiopulmonary exercise monitoring system by constructing artificial intelligence algorithms based on wireless sensor networks, ensure the accuracy and reliability of real-time data, and provide support for sports health management. In this study, an integrated learning algorithm was adopted, combined with network traffic monitoring and sensor data analysis, and through data preprocessing, feature extraction and anomaly detection model construction, real-time monitoring of cardiopulmonary monitoring data was realized. Simulation platform is used to evaluate the performance of the algorithm in different network environments, especially in wireless networks and mobile networks. The experimental results show that the proposed algorithm can effectively identify abnormal data under abnormal network conditions. Compared with traditional detection methods, the proposed method significantly improves detection efficiency and response speed, and can adapt to complex wireless sensing environment.

With the rapid development of Internet of Things (IoT) technology, optical sensors, as an important data acquisition tool, can accurately monitor the physiological state and athletic performance of athletes, and provide data support for personalized training. This study aims to explore the simulation effect of optical sensors in the Internet of Things sports training system combined with wireless sensor network and cloud technology, so as to improve the science and effectiveness of sports training. This paper uses simulation model to build a wireless sensor network-based motion training system for the Internet of Things, and focuses on analyzing the performance of optical sensors in the process of data acquisition. Through the three stages of sensor deployment, data transmission and cloud processing, the accuracy and reliability of the sensors in real-time monitoring of athletes’ athletic ability and physical status are evaluated. The simulation results show that the optical sensor can effectively collect motion data in the system and quickly transmit it to the cloud for analysis through wireless network. The response time of the system is significantly reduced, the stability and accuracy of data transmission are improved, and the real-time feedback of athletes during training is realized. The combination of wireless sensor network and cloud technology provides a new solution for sports training, and the effective application of optical sensor can significantly improve the training effect.

In mobile edge computing (MEC), randomly offloading tasks to edge servers (ES) can cause wireless devices (WD) to compete for limited bandwidth resources, leading to overall performance degradation. Reinforcement learning can provide suitable strategies for task offloading and resource allocation through exploration and trial-and-error, helping to avoid blind offloading. However, traditional reinforcement learning algorithms suffer from slow convergence and a tendency to get stuck in suboptimal local minima, significantly impacting the energy consumption and data timeliness of edge computing task unloading. To address these issues, we propose Parallel Exploration with Asynchronous Training-based Deep Reinforcement Learning (PEATDRL) algorithm for MEC network offloading decisions. Its objective is to maximize system performance while limiting energy consumption in an MEC environment characterized by time-varying wireless channels and random user task arrivals. Firstly, our model employs two independent DNNs for parallel exploration, each generating different offloading strategies. This parallel exploration enhances environmental adaptability, avoids the limitations of a single DNN, and addresses the issue of agents getting stuck in suboptimal local minima due to the explosion of decision combinations, thereby improving decision performance. Secondly, we set different learning rates for the two DNNs during the training phase and trained them at various intervals. This asynchronous training strategy increases the randomness of decision exploration, prevents the two DNNs from converging to the same suboptimal local solution, and improves convergence efficiency by enhancing sample utilization. Finally, we examine the impact of different parallel levels and training step differences on system performance metrics and explain the parameter choices. Experimental results show that the proposed method provides a viable solution to the performance issues caused by slow convergence and local minima, with PEATDRL improving task queue convergence speed by more than 20% compared to baseline algorithms.

Malware has now grown into one of the most important threats on the Internet. To meet this challenge, researchers regard malware classification as an effective method in malware analysis, which can classify the malicious samples with similar features into the same family. Although machine learning based malware classification models have great performance, they rely heavily on large-scale labeled datasets. In the real world, many malware families only have a small number of samples, which makes the traditional data-driven models perform poor results. In this paper, we propose an attention-based transductive learning network to solve the problem. In order to extract features, our approach first converts malware binaries into gray-scale images, and encodes them into feature maps using an embedding function. Then, we build a Gaussian similarity graph based on attention mechanism to transfer information from labeled instances to unknown instances. Through the end-to-end training, we demonstrate the effectiveness of the proposed approach on a malware dataset containing 11,236 samples with 30 different malware families. Comparing with state-of-the-art approaches, the experimental results show that our approach achieves a better performance.

To improve the performance of the blind image watermarking technique, a novel blind and robust image watermarking scheme in the spatial-domain by applying a two-level partition is proposed in this work. Specifically, the host image is firstly partitioned into non-overlapping 4 × 4 sub-blocks, and the standard deviations of all sub-blocks are computed in the meanwhile. Then, the sub-blocks with lower standard deviation are selected, and each selected block is partitioned into four non-overlapping 2 × 2 sub-blocks. Thereafter, the direct current coefficients of three 2 × 2 sub-blocks (i.e., top-left, top-right, and bottom-left sub-blocks) are computed in the spatial-domain without carrying out the two-dimensional discrete cosine transform. Lastly, utilizing the correlation principle between adjacent 2 × 2 sub-blocks, two bits of watermark are embedded into a 4 × 4 sub-block via adjusting the direct current coefficients of the three 2 × 2 sub-blocks. Experimental results show that the proposed image watermarking scheme is suitable for gray-scale and color images, and possesses a good performance in terms of invisibility and robustness. In particular, all peak signal noise ratios are greater than 42 dB, all structural similarity index measures are more than 0.96, and the normalized correlations are greater than 0.85 under various attacks.