A multi band integrated antenna scheme for the fifth generation (5G) mobile phone is proposed, that is, one antenna integrates five bands, namely Global Positioning System (GPS) Level 1 (L1), GPS Level 5 (L5), Wi-Fi 2.4G, Wi-Fi 5G and Long Term Evolution (LTE) band 32, the antenna volume for the whole mobile phone is effectively saved. This five-in-one antenna is located in the upper left corner of the back of the mobile phone, the upper clearance and side clearance of mobile phone are 1mm and 0.2mm respectively. The size of five-in-one antenna is 38.5mm by 26mm. The form of five-in-one antenna is combining inverted F antenna (IFA) and parasitism. This antenna is radiated by metal frame and laser direct structring (LDS), measurement results show that, passive efficiency in free space of GPS L1, GPS L5, Wi-Fi 2.4G, Wi-Fi 5G and LTE Band 32 could reach -4.5dB, -8.5dB, -6dB, -4.5dB, -6.5dB respectively, also the specific absorption rate (SAR) value of the transmission band is lower than Federal Communications Commission (FCC) specification. A mobile phone equipped with this five-in-one antenna has been commercially shipped.
A multi-channel (9*X) data acquisition system based on FPGA and MCU is developed, which can be used for signal field strength detection of base station equipment such as RFID. The system consists of data acquisition array, central controller, embedded firmware, and upper computer. Using the modular design method, the data acquisition array can be designed by changing the number of channels. The single-channel data acquisition module has a sampling resolution of 16 bits, a sampling rate of 200 kSPS/ch and dynamic range 70dB. The master-slave architecture is used to realize multi-channel synchronous acquisition of data. When the central controller communicates with the upper computer, it controls the operation of the single-channel data acquisition module and obtains the sampling data in a parallel manner. The embedded firmware is written in Verilog and C language to realize data acquisition, buffering, and sending and receiving of uplink and downlink signaling. The upper computer has the functions of system control and communication, establishes a data link with the central controller through the Ethernet cable, and obtains multi-channel sampling data through signaling. The system verifies the performance of the multi-channel data acquisition system through Analog-to-digital converter, single-channel data acquisition module and system performance test.
Since December 2019, COVID-19 has ravaged the world, severely affecting the quality of life and physical health of human society. Computed tomography (CT) imaging is an effective way to detect solid lung lesions as well as pulmonary ground-glass nodules and is an effective way to diagnose COVID-19. The automatic and accurate segmentation of COVID-19 lesion areas from CT images can determine the severity of the disease, which is essential for the diagnosis and treatment of COVID-19. A new model CAE-UNet(Combine-ASPP-ECA-UNet) is proposed in this paper for COVID-19 CT image segmentation based on UNet. The coding structure of UNet is replaced with the improved ResNet50 and incorporated with ECA attention module and atrous spatial pyramid pooling(ASPP). Fusing different sensory fields, global, local and spatial features to enhance the detail segmentation effect of the network. The experimental results on the CC-CCII show that the mIoU of the proposed CAE-UNet reaches 79.53%, which is better than some other mainstream methods. The proposed method achieves automatic and efficient segmentation of COVID-19 CT images.
Traditional ad-hoc is infrastructure-less. UEs (User Equipment) can realize rapid networking, but cannot guarantee the reliability of D2D (Device to Device) connection. 5G provides a network infrastructure of reliable connection, based on this, D2D provides visualized connection i.e. an autonomously controllable "virtualization ad-hoc". In order to integration of 5G D2D (ie. DCE) with DTE (Data terminal equipment) in power grids’ automation control and protection in the field area, dual-plane redundancy in substation and heterogeneous hand in hand connection in distribution power lines are proposed. Cross-layer connection failure detection and autonomous maintenance are further discussed. The applicable test shows the method can significantly improve the reliability of the power grids’ D2D network.
Nowadays, the rapid development of Internet technology drives the advance of the big data era, and along with the widespread use of smart devices and social media, multimedia data is growing explosively. With the increasingly complex and diverse needs of information exchange, collection and storage, the types of information have also evolved from traditional text information to diverse data forms such as pictures and video and audio, bringing different degrees of convenience to people's work and life and other scenarios. However, the huge amount of multimedia data also makes information storage and retrieval more cumbersome. How to realize the effective storage and efficient retrieval of data, so as to better utilize the value of multimedia data, is one of the challenges that academia and information industry are tackling nowadays. In this paper, we study the cross-media retrieval technology of text and image by Contrastive Language-Image Pre-training model based on natural language processing method. The cross-media pre-training idea proposed in this paper can be applied not only to text-image processing, but also theoretically to mutual retrieval of modal information of video and audio, etc.
The researchers aimed at creating an App-Based Gym Workout Instructor using Image Recognition via Artificial Neural Network which can recognize the body type of a male person using images and show the workout for the body type. The input is a whole-body image of a male person and the output is the workout for the detected body type. Using MATLAB, the researchers created an Artificial Neural Network that is trained to recognize body types and C# platform to implement the ANN. The results of the study showed that the developed system was able to determine the body type of the user. In terms of the over-all accuracy of the developed igym instructor for all of the body type defined, it was fairly moderate with an average of 64.38%. The effectivity and accuracy of the iGYM does not only depend on the number of training data but also with the quality of the data set.
Computer vision has been aiding various industries in making work efficient. In the case of marine and ocean-related industries, climate change, greenhouse gasses, fishing exploitation, and coastal contamination are all causing significant effects on human life. In the Philippines, the same problem that burdens most coastal countries exists. The system implemented to aid this problem is limited to those with access to the SAR and has no local or small-scale implementation. Different studies focus on the utilization of algorithms to detect and classify ships in the sea. Therefore, counting and classification based on the type of ships are essential. Using the YOLOv5 and DeepSORT algorithm, the system was able to achieve a model, prototype, and counting accuracy of 98.65%, 98.11%, and 100% respectively. Some of the misclassifications are due to the close similarities of the different classes and the under representation of some classes. It can be concluded that the produced model is accurate in detecting, classifying, and counting ships based on type.
At present, the fault identification method of glass insulators has the problems of difficult feature extraction and poor generalization ability of the model, which leads to the low accuracy of fault identification of glass insulators. Based on the Yolov5 network, this paper introduces a lightweight general sampling operator CARAFE to solve the problem of difficult feature extraction. At the same time, the attention mechanism module SENet is added to give different channels different weights to improve recognition accuracy. In addition, this paper makes further improvements in the network structure to make the network fit the above improvements. The experimental results show that the fault recognition rate of glass insulators is significantly improved compared with the unimproved network.
Slotted ALOHA implemented in the internet of things (IoT) uses the LoRaWAN media access control (MAC) protocol to improve its performance. Various backoff algorithms have been proposed in LoRaWAN to evaluate the delay, throughput, and packet loss rate (PLR). However, an adaptive backoff algorithm has been implemented in this paper to examine the performance metrics for electronic shelf labels (ESLs). The use of adaptive backoff optimizes the delay and the bandwidth (BW) for more efficient and meaningful communication with a certain degree of data loss. The results illustrate that the intra-slicing model and adaptive backoff estimate the optimal delay for each slice, starting with the best slicing priority for the end device (ED) which brings mobility into the network.
The traditional operation and maintenance platform is dependent on the static rules set manually, which can not better cope with the dynamic and complex changing scene. Nowadays, with the rapid development of machine learning and artificial intelligence, intelligent operation and maintenance system can make more efficient and accurate decisions in the face of dynamic changing scenarios through big data accumulated in business scenarios, and can also automatically monitor services, detect abnormal events, and deal with faults in emergency. This paper carefully analyzes the necessity of constructing an intelligent operation and maintenance system, and the application of machine learning in the analysis and fault detection of intelligent operation and maintenance system.
The IEEE 802.11 MAC layer lacks the limitation of QoS guarantee, and the current research and improvement of IEEE 802.11 differentiated services are mostly based on single rate, and rarely consider the research in the case of multi-rate. Aiming at this situation, this paper analyzes the IEEE 802.11 protocol mechanism in depth., and established a multi-rate IEEE 802.11 protocol model. Based on this, a IEEE 802.11 differentiated service solution is proposed: dynamic timeslot base on probability of channel access (DTPCA). By detecting the network information, the strategy can update the network node-related protocol parameters adaptively in real time, which can effectively realize the multi-rate differentiated service performance and improve the system throughput at the same time. Through theoretical analysis and detailed simulation experiments, the effectiveness of the proposed strategy is confirmed. Moreover, the strategy is easy to control and simple to implement, and it is easy to be applied and promoted in the actual network.
This paper proposes a limit-achievable joint range and direction of arrival estimation method in pulse-Doppler radar, named sampling a posteriori (SAP). To evaluate the performance of the SAP, entropy error is introduced and it is proved that the empirical entropy error of the SAP achieves the theoretical entropy error when the number of snapshots tends to infinity. In order to reduce the computation burden of the SAP, a relaxation technique is adopted, which is based on a random number generator and the cumulative distribution function. Numerical simulation is carried out and results indicate that the relaxation technique can effectively reduce the computation burden of SAP at the cost of a insignificant estimation performance loss.
Most existing behavioral models of the radio frequency (RF) power amplifiers assume that power amplifiers are memoryless devices. But in broadband communication systems, memory effects of the RF power amplifiers are significantly observed. The traditional memoryless models can no longer accurately depict the input-output relationship of power amplifiers. This paper investigates power amplifiers memory effects and a pre-distortion method is proposed for linearizing RF power amplifiers with memory effects. A valid behavioral model based on the Volterra series is suggested to treat memory effects. Since the characteristics of amplifiers change during transmission time, a recursive least squares algorithm with size- fixed observation matrices is developed to update the parameters of the proposed power amplifiers model. The identification algorithm can decrease computational complexity and data storage space and facilitate real-time online identification. Simultaneously, the pre-distortion model is constructed to simulate. Simulation results show that the proposed method can effectively compensate for the nonlinear distortion and memory effects of RF power amplifiers. The model's accuracy is used to evaluate power amplifiers memory effects by the enormalized mean square error (NMSE). In the end, simulation and evaluation of this pre-distortion system proceeded, and the power spectral density estimation was also applied to calculate the adjacent channel power ratio.
The wireless network technology represented by IEEE 802.11 technology plays an increasingly important role in practice. However, in the multi-node and multi-rate network scenario, this technology will produce abnormal performance, resulting in high-speed and medium-speed nodes. The throughput drops to the same as the throughput of the low-speed node, which reduces the total throughput of the system and degrades the performance. Aiming at this problem, a solution (TRCWAA) is designed. Through the minimum contention window, the size of the minimum contention window of each node is inversely proportional to its transmission rate, so as to realize the reasonable allocation of channel resources. Theoretical analysis and simulation experiments show that the TRCWAA strategy can well solve the problem of abnormal performance and unfairness, greatly improve the network fairness, and improve the total throughput of the system. In addition, the significant advantage of this method is that it only needs to make minor modifications on the basis of 802.11, that is, to adjust the minimum contention window size according to the transmission rate. Its physical implementation is very simple and convenient. No modification to the hardware, so it is easy to implement and generalize in real networks.
Alzheimer’s disease is a degenerative disease of the nervous system. If the doctor can detect the disease early, he can treat the patient in advance to slow down the deterioration of the health. We propose a network 3D_ResNeXt_Bi-LSTM fused with ResNeXt and Bi-LSTM, which uses MRI brain images to classify and recognize AD (Alzheimer disease) and NC (Normal Contrast) from neuroimaging. We use a 3D convolution kernel to replace the 2D convolution kernel and flatten the feature of the final ResNeXt into one-dimensional data and send it to Bi-LSTM. So that the network can thoroughly learn the spatial information of the 3D brain image data, finally we send the features to the classifier for classification. Experiments on the ADNI dataset show that our network’s highest classification accuracy for AD and NC is 98.97%.
The alignment of inertial measurement units(IMUs) to segment is an important step in inertial motion capture, which directly affects whether the imu data can fully represent the motion of the segment. Inspired by the gene crossover and mutation of Genetic Algorithm(GA), we propose a dynamic inertial weighted particle swarm optimization algorithm with cross factor to solve the joint constraint problem, and compared our algorithm with Particle Swarm Optimization(PSO) and Dynamic Inertial Weighted Particle Swarm Optimization(DPSO) algorithms to show the superiority of our algorithm during human lower limb movements. The experiment shows that introduced the random cross mechanism between particles with larger fitness and only the effective cross retained, makes the new algorithm show better search ability and convergence effect in this project, the stability and effectiveness are also improved. Our current work provides a good support for accurate calculation of joint angles in the future.
In this paper, we propose an efficient estimation method for resolving the challenge related to the carrier frequency offset (CFO) estimator technique for multi-user OFDM systems when the base station (BS) employs the beam-scanned leaky wave antenna (LWA). We consider that the LWA can change the scanning frequency, which can perform the CFO estimation independently for each user, thereby reducing interference within the subcarriers. The proposed CFO estimator scheme utilizes the subspace MUSIC (multiple signal classification) estimation algorithms developed by computing correlation matrix through the eigendecomposition process. We further compute the CRLB to serve as a performance benchmark with the developed MUSIC scheme. Finally, different numerical results are available to justify the effectiveness of the proposed studies under different configurations.
In order to meet the increasingly frequent communication requirements between data link networks working in different frequency bands, a design of dual-band circularly polarized microstrip antenna is proposed. Based on the basic form of the microstrip antenna, this antenna has dual-band circularly polarized radiation characteristics by cutting four centrally rotationally symmetrical equal-arm U-shaped slots on the square-cut corner patch. The radiation mechanism and structure of this type of antenna are studied and analyzed using electromagnetic simulation software, and a dual-band circularly polarized antenna with operating at 1.85GHz and 4.24GHz is designed, which radiate left-handed circularly polarized waves in low and high frequency. The impedance bandwidth is 1.83~1.89GHz and 4.17~4.29GHz when S 11 less than -10dB, the gain reaches 5dB at the central low frequency, and reaches 7.1dB at the high frequency, the axial ratios at the two frequency points are both less than 3dB.
Along with the development of data scale and the high complexity of sound signals, feature extraction and classification methods of sound signals have become a major research hotspot. However, the current sound signal feature extraction methods are difficult to accurately and stably provide a high-precision classification effect for the sound signal due to the complex frequency distribution and the influence of noise. Therefore, a robust feature extraction method for sound signals based on multi-scale and multi-directional Gabor filters and Mel frequency cepstral coefficient (MtWGM) was proposed. This method performs preprocessing on the signal by mixing hard threshold and soft threshold wavelet denoising. The Gabor filter is used to the combined energy spectrum of the framed signal, to achieve the effect of relatively more balanced intra-class features and more prominent inter-class features, and finally improve the noise reduction performance and classification accuracy of sound signals. The experiments are conducted out on three different sound signal datasets. Three classifiers are trained to test the effectiveness of extracted features. The experimental results show that the proposed multiple wavelet Gabor_MFCC (MtWGM) method has obtained better classification accuracy and robustness than that of MFCC.
This systematic review aims to determine the audiovisual language trends in video clips according to the scientific literature from 2010 to 2020. Likewise, the articles obtained were found in the Ebsco, Redalyc, and ProQuest databases. The words “language” and “audiovisual” were used to carry out the exclusion and inclusion of the articles. The synthesis method was applied using categories to carry out the research, and the reading comprehension technique and underlining were used to obtain information. The results are related to the audiovisual language of music videos and it was found that the rhythm allows a relationship between the length of shots and the sequence. Likewise, visual elements such as shots, and angles, among others, significantly contribute to production. Likewise, the tools that are used allow the production to have the power to transmit the essence of audiovisual aesthetics. On the other hand, there were limitations such as articles in other languages like Portuguese and English, and restricted access to scientific research in databases. In conclusion, the use of rhythm can lead to the variation of balance in the sequences. Furthermore, the use of shots as a language is applied throughout the audiovisual production, which is enriched by the variety that makes up the set of images. In addition, the music videos present techniques where the production has exceptional dynamism, as previously stated.
