Search

The security of wireless information transmission in large-scale multi-input and multioutput (MIMO) is the focus of research in wireless communication. Recently, a new artificial noise—SCO-AN which shows no orthogonality to the channel, is proposed to overcome the shortcomings of traditional artificial noise. In the previous research, the optimization function of SCO-AN is not convex, and its extremum cannot be obtained. Usually, nonconvex optimization algorithms or iterative relaxation algorithms are used to get the maximum value of the optimization objective function. Nonconvex optimization algorithms or iterative relaxation algorithms are greatly affected by the initial value, and the extremum cannot be obtained by a nonconvex optimization algorithm or iterative relaxation algorithm. In this paper, we creatively apply the strong law of large numbers to obtain the optimal value of the optimization function of SCO-AN under the condition of large-scale MIMO: the strong law of large numbers is applied to obtain the ergodic lower bound (ELB) expression of SC for SCO-AN. The power allocation (PA) problem of the SCO-AN system is discussed. We use a statistical method to get the formula for calculating the optimal power distribution coefficient of the SCO-AN system. The transmitter can use the optimal power ratio of PA to distribute the transmitted power without using the PA algorithm. The effect of imperfect channel state information is discussed. Through simulation, we found that more power should be generated for SCO-AN if the channel estimation is imperfect and the proposed method can achieve better security performance in the large-scale MIMO system.

The COVID-19, novel coronavirus or SARS-Cov-2, has claimed hundreds of thousands of lives and affected millions of people all around the world with the number of deaths and infections growing exponentially. Deep convolutional neural network (DCNN) has been a huge milestone for image classification task including medical images. Transfer learning of state-of-the-art models have proven to be an efficient method of overcoming deficient data problem. In this paper, a thorough evaluation of eight pre-trained models is presented. Training, validating, and testing of these models were performed on chest X-ray (CXR) images belonging to five distinct classes, containing a total of 760 images. Fine-tuned models, pre-trained in ImageNet dataset, were computationally efficient and accurate. Fine-tuned DenseNet121 achieved a test accuracy of 98.69% and macro f1-score of 0.99 for four classes classification containing healthy, bacterial pneumonia, COVID-19, and viral pneumonia, and fine-tuned models achieved higher test accuracy for three-class classification containing healthy, COVID-19, and SARS images. The experimental results show that only 62% of total parameters were retrained to achieve such accuracy.

Convolutional neural networks have an immense impact on computer vision tasks. However, the accuracy of convolutional neural networks on a dataset is tremendously affected when images within the dataset highly vary. Test images of plant leaves are usually taken in situ. These images, apart from the region of interest, contain unwanted parts of plants, soil, rocks, and/or human body parts. Segmentation helps isolate the target region and a deep convolutional neural network classifies images precisely. Therefore, we combined edge and morphological based segmentation, background subtraction, and the convolutional neural network to help improve accuracy on image sets with images containing clean and cluttered backgrounds. In the proposed system, segmentation was applied to first extract leaf images in the foreground. Several images contained a leaf of interest interposed between unfavorable foregrounds and backgrounds. Background subtraction was implemented to remove the foreground image followed by segmentation to obtain the region of interest. Finally, the images were classified by a pre-trained classification network. The experimental results on two, four, and eight classes of datasets show that the proposed method achieves 98.7%, 96.7%, and 93.57% accuracy by fine-tuned DenseNet121, InceptionV3, and DenseNet121 models, respectively, on a clean dataset. For two class datasets, the accuracy obtained was about 12% higher for a dataset with images taken in the homogeneous background compared to that of a dataset with testing images with a cluttered background. Results also suggest that image sets with clean backgrounds tend to start training with higher accuracy and converge faster.

Since the content of soil organic carbon (SOC) is significantly correlated with the soil reflection spectrum, hyperspectral remote sensing technology provides an effective new choice for the estimation of soil properties. At present, deep learning method relies on its strong feature extraction ability and has been successfully applied to the field of data analysis. This paper attempts to apply the deep learning method to the estimation of SOC content and proposes a method combining the convolutional neural networks (CNN) and quantile regression (QR). This method constructs a three-layer CNN and adjusts the network structure with the idea of QR. The experimental results are presented for two soil datasets, LUCAS (Land Use/Land Cover Area Frame Survey) and AfSIS (Africa Soil Information Service), and compared with several advanced deep learning and traditional machine learning regression models. The experimental results show that this method performs well in estimation.

Automatic modulation classification (AMC) for single signals has been widely studied in the past two decades. Meanwhile, the demand for multi-signal AMC is also increasing; however, single-signal AMC performs poorly and few methods were proposed for multi-signal AMC. To address this problem, a blind AMC method for overlapped co-channel signals applying capsule networks is proposed. Multiple transmitters transmit signals simultaneously and signals are received by a single-antenna receiver. The method handle signals at intermediate frequency directly, and is able to detect the number of signals automatically. Numerical results demonstrate the effectiveness of our proposed method, and excellent generalization ability provided by capsule network.

In this letter, an unmanned aerial vehicle (UAV) is employed as relay to build temporary communication links for disconnected ground nodes. To improve the fairness and energy efficiency, two scenarios are investigated. First, for given transmission power budget, we consider maximizing the minimum information rate by joint power and trajectory optimization over a time horizon. Then, for given required information rate, we consider minimizing the sum-power consumption. These two objectives are both non-convex, which is intractable to solve. To address these objectives, we propose iterative algorithms to approximate the solution by successive convex optimization techniques. Numerical results show the effectiveness of the proposed approaches.

Modern intelligent transport systems focus on the integration of multiple sensors to obtain hybrid navigation schemes. A key issue of a hybrid scheme is distribution of the information sharing coefficients (ISCs) of subsystems and the fusion of parallel multiple observations of navigation sensors. Recently, deep learning methods, particularly convolutional neural networks (CNNs), have achieved great success in image processing tasks. However, there has been limited work in using deep learning for multisensor-based integrated navigation solutions. In this letter, we propose an ensemble learner-based classification and information fusion method, in which estimation error covariance matrices provided by local adaptive filters are used as input for the classifier, and the triple numbers of ISCs are determined by the proposed scheme. The results validate the effectiveness of the proposed scheme, in which the adequately trained ensemble learner can detect the degradation of a subsystem that may suffer atypical observations or faults and consequently can adjust the corresponding ISC in real time.

In multiple-input multiple-output wireless communication system, artificial noise (AN) is designed to be aligned into the null space of the legitimate channel so that it has no effect on the legitimate channel. Meanwhile, AN will reduce the secrecy capacity of the eavesdropping channel so that AN increase the secrecy capacity of the wireless communication system. Nevertheless, AN technology only pays attention to the CSI of the legitimate channel but does not pay attention to the CSI of the eavesdropping channel so that the secrecy capacity of the system can be further improved. In this paper, we analyze the performance of traditional AN. A secrecy capacity optimization artificial noise for the whole communication system to further improve the system's secrecy capacity is proposed. The secrecy capacity optimization artificial noise is considered as a kind of artificial noise which is disadvantageous to the legitimate receiver but more disadvantageous to the eavesdropper. Furthermore, we further analyzed the technical details of AN technology and found the key factor affecting the secrecy capacity. Meanwhile, the secrecy capacity optimization artificial noise is effectively designed by applying these conclusions and the key factors of secrecy capacity optimization artificial noise affecting the secrecy capacity of the wireless communication system are described. The analysis and simulation results in practical environments show that the proposed method has a good performance on improving the secrecy capacity.

Modulation recognition is widely used in both military and civil fields. With the complexity of electromagnetic environment, modulation recognition technology is needed to adapt to the situation of large dynamic SNR and multi signal coexistence. In this paper, a modulation recognition algorithm based on deep learning is proposed, which directly extracts the features of IF signals and simplifies the design of artificial features. Two classifiers are used to ensure the correct rate with lower complexity in the case of single signal, while the accuracy rate and noise robustness are higher than traditional methods when multiple signals with time-frequency aliasing coexist.

Space formation aircraft communication system needs to meet strong anti-interference ability, low power consumption and high transmission rate performance. Ultra-wideband (UWB) has high data transfer rate, low device complexity, high interference resistance and confidentiality. Therefore, UWB technology is used in space formation aircraft communication system, which is direct sequence ultra-wideband (DS-UWB) space formation aircraft communication system. This paper, we put forward a multi-user detection (MUD) algorithm based on Gaussian Mixture Model (GMM) in DS-UWB communication system and its field-programmable gate array (FPGA) implementation. We adopt a way which based on MATLAB and FPGA platform, and results show that MUD algorithm based on GMM has better bit error rate (BER) performance compared to classical multi-user detection algorithm.

Unmanned aerial vehicle (UAV) is suitable to improve connectivity of wireless communication networks in harsh environment due to its flexible deployment and high mobility. In this paper, we consider the case that a UAV is employed as a mobile relay to ferry data between two disconnected ground nodes. The amplify-and-forward mobile relay technique is studied. By exploiting the channel state variation induced by the movement of UAV, we consider maximizing the end-to-end throughput of this mobile relay system by optimizing the source/relay power allocation, as well as the UAV’s trajectory. However, the formulated problem is non-convex and intractable to solve. To address this non-convex problem, we propose two iterative algorithms to optimize the source/relay power allocation with the fixed UAV’s trajectory and the UAV’s trajectory with fixed power allocation, by successive convex optimization, respectively. Then, an iterative algorithm is proposed to optimize the source/relay power allocation and UAV’s trajectory in an alternate manner. By exploiting the predictable channel variation via the proposed power and trajectory optimization scheme, numerical results show that significant throughput gains can be achieved.

Arbitrary-order orbital angular momentum multiplexing (AOAMM) systems utilize OAM modes with both integer and fractional topological charges which are non-orthogonal. In this paper, the transmission matrix and the capacity per unit bandwidth, i.e., the spectral efficiency (SE) of an AOAMM system is derived based on the spatial cross correlations of the OAM submodes under different aperture conditions. The results show that in limited apertures, the SEs of AOAMM systems increase dramatically as the interval of two adjacent OAM submodes decreases by losing orthogonality. AOAMM systems are effective choices for satisfying the explosive growth of the communication requirements. This paper provides insight into the selection of spatially multiplexing approaches and the design of interference mitigation techniques for AOAMM systems with increased SEs.

Automatic modulation classification (AMC) plays a key role in non-cooperative communication systems. Feature-based (FB) methods have been widely studied in particular. Most existing FB methods are deployed at a fixed SNR level, and the pre-trained classifiers may no longer be effective when the SNR level changes. The classifiers may also need to be re-trained to be suitable for the varying channel environment. To address these problems, a robust AMC method under varying noise conditions is proposed in this paper. The method attempts to select noise-insensitive features from a large feature set to ensure that the trained classifiers will be robust to SNR variations. First, a feature set consisting of 25 types of features is extracted, and 4 features that are insensitive to noise are chosen through a feature selection method based on rough set theory. The generalizability of an SVM classifier trained on the 4 chosen features is evaluated based on numerical results. The classification accuracy remains reasonable when the SNR varies between 5 and 20 dB, indicating that the proposed method can be deployed under varying noise conditions.

Agriculture is not only China's primary industry but also the foundation of the national economy. The amount and quality of agricultural products are inextricably linked to people's daily life. The outbreak of pests and diseases in the field has a great impact on agricultural production, so it can be seen that the prevention and control of pests and diseases are very important. In order to control crop diseases and pests, this paper combines emerging machine learning techniques based on a large number of crop pest and disease pictures, and introduces two kinds of convolutional neural networks------AlexNet and GoogleNet to detect crop pests and diseases. Much work has been done to improve the algorithm, including proposing an improved network based on migration learning and data expansion, which greatly improves the accuracy of detection. Compared with the manual detection method and the traditional algorithm, the improved detection algorithm based on the deep convolutional neural network has the highest detection accuracy rate of 98.48% for 38 pests and diseases, which has higher efficiency, practicability, and accuracy.

The Chinese Haiyang-2A (HY-2A) altimetry satellite is equipped with a calibration microwave radiometer (CMR) for correcting atmospheric water vapor path delay in radar altimeter observations. To evaluate the satellite-borne CMR, we retrieved 1 Hz high-frequency precipitable water vapor (PWV) using shipborne GPS and GLONASS from a two-month cruise in the Indian Ocean. This open-sea evaluation of the satellite-borne radiometer is free of the contamination of the satellite footprints induced by coastal lands, which occurs inevitably in ground-based or coastal stations. The estimates and errors of the retrieved PWV from shipborne GNSS kinematic precise point positioning were analyzed and then compared to the CMR-retrieved PWV. The results show that the shipborne GNSS kinematic precise point positioning can obtain marine PWV with an uncertainty of about 2.8 mm. When the HY-2A sub-satellite point and the ship cross, the HY-2A CMR-retrieved PWV is in good agreement with the GNSS PWV with the difference of about 0.8 mm, which demonstrates that the HY-2A satellite can contribute high-precision precipitable water vapor measurements to weather and atmospheric studies.

Automatic modulation classification (AMC) is a core technique in noncooperative communication systems. In particular, feature-based (FB) AMC algorithms have been widely studied. Current FB AMC methods are commonly designed for a limited set of modulation and lack of generalization ability; to tackle this challenge, a robust AMC method using convolutional neural networks (CNN) is proposed in this paper. In total, 15 different modulation types are considered. The proposed method can classify the received signal directly without feature extracion, and it can automatically learn features from the received signals. The features learned by the CNN are presented and analyzed. The robust features of the received signals in a specific SNR range are studied. The accuracy of classification using CNN is shown to be remarkable, particularly for low SNRs. The generalization ability of robust features is also proven to be excellent using the support vector machine (SVM). Finally, to help us better understand the process of feature learning, some outputs of intermediate layers of the CNN are visualized.

With the popularization and development of cloud computing technology and intelligent terminal, the “Educational Cloud Service + Cloud Terminal” (referred to as the “cloud + terminal” mode) mode, which is composed of educational cloud service and intelligent learning terminal has opened up a new teaching scene and learning situation for digital teaching materials. First of all, this paper introduces the educational cloud service system and the “cloud + terminal” model; Then, on the basis of deeply analyzing the elements and characteristics of personal learning environment, this paper constructs a model of personal learning environment oriented to digital teaching materials under the mode of “cloud + terminal”, at the same time, the paper also introduces its main functions; Finally, the paper designs a learning system for the digital teaching materials under the mode of “cloud + terminal” from four aspects, that is, the terminal equipment, the tool software, the contents of digital teaching materials and the educational cloud service platform. The research contents of this paper have reference value and guiding significance for the design, development of digital educational materials and the construction of learning environment.

Automatic Modulation Classification (AMC) is of paramount importance in wireless communication systems. Existing methods usually adopt a single category of neural network or stack different categories of networks in series, and rarely extract different types of features simultaneously in a proper way. When it comes to the output layer, softmax function is applied for classification to expand the inter-class distance. In this paper, we propose a hybrid parallel network for the AMC problem. Our proposed method designs a hybrid parallel structure which utilizes Convolution Neural Network (CNN) and Gate Rate Unit (GRU) to extract spatial features and temporal features respectively. Instead of superposing these two categories of features directly, three different attention mechanisms are applied to assign weights for different types of features. Finally, a cosine similarity metric named Additive Margin softmax function, which can expand the inter-class distance and compress the intra-class distance simultaneously, is adopted for output. Simulation results demonstrate that the proposed method can achieve remarkable performance on an open access dataset.

The prohibitive computational complexity caused by redundant dictionary which is used in signal sparse decomposition has always been perplexing researchers of signal processing all the time. However, large over-complete dictionaries are essential to approximate the signal. In this paper, a decomposition algorithm based on modulation correlation partition (PBMC) is introduced, which improves the process of searching matched atoms in the redundant dictionary of functions. Through analyzing the structure of signals, we unite the frequency factor and phase factor to obtain the 2D modulation factor. The over-complete dictionary is partitioned into several sub-dictionaries according to the modulation correlation of atoms. Each sub-dictionary is represented by a single selected atom which is used in the greedy algorithm. At the end of this paper, experimental results show that the computational complexity of signal sparse decomposition can be adequately reduced by partitioning over-complete dictionary without impacting the decomposition result. HighlightsAn algorithm that raises efficiency of matching pursuit is proposed.Partition strategy is introduced to improve structure of redundant dictionary.The proposed approach is more effective than the traditional ones and has robustness.

Space formation flying has attracted a lot of attention recently. In this paper, we using direct-sequence ultra-wideband (DS-UWB) as the physical layer of inter-spacecraft communications. The traditional multiuser detection method like soft-input soft-output (SISO) cost too much computation. The proposed multiuser detector no longer uses interleaving, and the computational complexity has a linear relationship with the number of users. The bit error ratio (BER) simulation result shows that the proposed method approaches the single-user system with increasing signal-to-noise ratios (SNRs).

Automatic modulation classification (AMC) plays an important role in cognitive radio and is widely studied recent years. However, most existing AMC schemes must be deployed under their training SNRs, which makes them highly dependent on the accuracy of channel estimation. The classifiers may need to be re-trained to fit the varying channel condition. To address this problem, a feature combination method aiming to find noise-robust features under fading channels is proposed in this paper. Stacked auto encoder is deployed to explore robust features from an extracted feature set, and these new features is then used to train a support vector machine (SVM). Numerical results shows that the generalization ability of SVMs trained with new features can be significantly improved; therefore the method is robust to SNR variation.

CFAR is a widely used algorithm for target detection in SAR images. The simplicity of computation and stable performance make it a key role in practice. However, for man-made target detection, conventional CFAR has a limited performance because of the non-adaptive processing window and the varieties of categories, sizes and structures of targets. In order to detect man-made targets with different size and complex structures in high resolution SAR images, an improved CFAR algorithm with an adaptive processing window named Adaptive-Window CFAR is proposed. A global guard window obtained by pre-detection adaptively is used to take place of the guard window in conventional CFAR makes AW -CFAR an algorithm with both adaptive threshold and adaptive processing window. In this case, the size of detectable targets is not fixed anymore and targets of different sizes and complex structures are detectable in AW-CF AR. Images with different resolutions and environments, which contain different categories of man-made targets, are used in the experiments. The experimental results show that the AW -CFAR inherits the simplicity of computation and stable performance of the conventional CFAR and has a better performance of man-made target detection in high resolution SAR images.

Recently, millions of persons research e-learning which has important social significance and extensive application background. E-learning influences the global higher education significantly. Based on the applied superiority of MOOC (the chief kind of e-learning), the main research aim of this paper is to provide effective learning strategy and to increase the engagement and the learning effect in e-learning. This paper utilizes the acquired big data during e-learning to analyze three dimensions of teacher-support and five dimensions of learners’ learning engagement in detail. An effective model is proposed to analyze the relation between teacher-support and learning engagement. Based on the acquired big data, the mean, variance, correlation coefficient and other datum are analyzed to compare teacher-support and learning engagement in the effective model, all path coefficients of the result achieve significant level: p > 0.05. This paper proposes improved recommendations and supplies effective learning strategies.

As a subject of the quantum information science, Quantum Secure Direct Communication (QSDC) has been different from Quantum Key Distribution (QKD), QSDC makes the cryptograph to be transmitted directly in the quantum channel. Quantum non-cloning theorem and the indeterminacy principle insure its security. So QSDC has an important research value for the quantum communication. This letter proposed one division-multiplexed of the control code based on QSDC, the simulation is realized, utilizes the division-multiplexed of the control code to encrypt the quantum subsequence, and further improves the security. Two steps of the quantum communication are optimized to be one step. The division-multiplexed of the control code improves the efficiency of QSDC.

In this paper, a new non-coherent chaos-based communication scheme, named orthogonal frequency division multiplexing based different chaos shift keying (OFDM-DCSK), is presented inspired by multi-carrier direct sequence code- division multiple access (MC-DS-CDMA) system, multi-carrier DCSK (MC-DCSK) system, and channel estimation of OFDM system. In this scheme, all of the occupied subcarriers are grouped into several groups. In each group, one chaotic reference is transmitted over the central subcarrier, while multiple modulated data streams are transmitted over the remaining subcarriers. Therefore, this transmitter structure increases the spectral efficiency, compared with the conventional DCSK system. At the receiver side, no radio frequency delay circuits are needed to demodulate the received data, which makes the system easy to implement in practice. The bit error rate (BER) performance of the OFDM-DCSK system is derived based on Gaussian approximation under additive white Gaussian noise (AWGN) channel. Simulation results are given under AWGN and two-path Rayleigh flat fading channel, which confirm the effectiveness of this new design.

Various vehicle applications in the future will require reliable and accurate vehicle positioning techniques. Nowadays, hybrid schemes combining multiple sensors have been promising solutions for high precision positioning. However, positioning error can be remarkably affected by the temporal alignment and fusion algorithms in practice. In this paper, we propose a decentralized fusion structure containing an inertial navigation system (INS), a GPS receiver, a RFID reader, and an odometer. The update rates of the sensors are different, and the INS/GPS integration presents severe performance degradation in urban area. To achieve an effective alignment and fusion of the sensors, we propose a concept of confidence field to indicate the confidence levels of subsystems for changing driving environments. A confidence field-based alignment and fusion algorithm and its simplification are proposed when we use the weighted least squares curve method. Time biases of the sensors are also considered in local adaptive filters. Simulation results demonstrates the performance of the proposed scheme with the proposed algorithms, especially in GPS- denied environments.

With the capacity of overcoming radio spectrum shortages for wireless communications in smart grids, cognitive radio enabled Advanced Metering Infrastructure (CR-AMI) networks are expected to enhance the efficiency and practicability of future smart grids. As an integral component of the smart grid ecosystem, CR-AMI networks are practically deployed as a static multi-hop wireless mesh network. This paper focuses on the investigation of an novel RPL-based routing protocol for enhancing the energy efficiency in CR-AMI networks. In accordance with practical requirements of green communications in smart grids, the proposed routing protocol adopts the energy efficiency over virtual distance as the core of routing mechanism such that the energy-efficient route can be achieved. In addition, the protocol has the mechanism for primary (licensed) users protection whilst meeting the utility requirements of cognitive radio users. System-level evaluation shows that the proposed routing protocol has better performances compared with existing routing protocols for cognitive radio- enabled AMI networks.

Tiangong-2 is a space laboratory of China, which carries out various the Earth observations and space experiments. Controlling of the Tiangong-2 and completing these experiments rely on high precision orbit information. Hence, study on perturbed motion, which can be switched to analyze perturbation forces is the basis and core tasks. Firstly, in order to analyze the influence to the orbit by the perturb forces, initial orbit state is obtained by two line elements. Furthermore, several integral orbits are achieved under different perturb forces using numerical integration method, including the Earth potential perturbation, tides, three body perturbation, atmospheric drag perturbation, the solar radiation pressure and the Earth albedo. The experimental result shows that the atmospheric drag perturbation is one of the key factors to constraint the orbit accuracy and much more complex, due to the fact that Tiangong-2 is a low Earth orbit spacecraft. Moreover, atmospheric drag perturbation is discussed in details for comparing different upper atmosphere density models. And its orbit affect level is concluded, which is benefit for precise orbit determination.

Aimed at lack of temperature and pressure data in retrieving Precipitable Water Vapor (PWV) from Precise Point Positioning (PPP), a Layer Interpolation Method (LIM) used the NCEP/NCAR layer reanalysis data is proposed. 14 IGS stations located at different altitudes are selected to validate the method with its meteorology files. The interpolated temperature and pressure values from the LIM and those from the classical Surface Interpolation Method (SIM) have been compared with the observed data, further for evaluating effect on the derived PWV. The experimental results show that (1) two interpolation methods perform well in pressure at the GPS sites below 2000 m with RMS 1.79 mbar (LIM) and 3.83 mbar (SIM), LIM is better than SIM. While the interpolated temperature values nearly have the same accuracy with RMS 3.63 K (LIM) and 3.69 K (SIM). (2) The LIM and the SIM exist the large pressure errors for the GPS site above 3000 m as LHAZ. The interpolated pressure error reach 8 mbar, which contributes 3 mm PWV error. While the accuracy is still better than the accuracy of PWV without meteorological data, especially for LIM with the improvement from 4.36 to 3.29 mm. (3) The derived PWV accuracy with interpolation data shows obvious positive correlation with the accuracy of the interpolated pressure data, while the accuracy of PWV has no correlation with temperature data, which means the accuracy of retrieving PWV is mainly effected by pressure. (4) The accuracy of the PWV retrieved with the interpolated meteorological data has no obvious coupling relation with altitude and latitude.

Convolutional neural network has a huge partake and is still a dominating tool in the field of computer vision. In this study, we introduce a model with depthwise separable convolution architecture for plant disease detection based on images of leaves. We present two versions of depthwise separable convolution comprising two varieties of building blocks. Training and testing of the models were performed on a subset of publicly available PlantVillage dataset of 82,161 images containing 55 distinct classes of healthy and diseased plants. These depthwise separable convolutions achieved less accuracy and high gain in convergence speed. Several models were trained and tested, of which Reduced MobileNet achieved a classification accuracy of 98.34% with 29 times fewer parameters compared to VGG and 6 times lesser than that of MobileNet. However, MobileNet outperformed existing models with 36.03% accuracy when testing the model on a set of images taken under conditions different from those of the images used for training. Thin models were also introduced, which showed effective trade-off between latency and accuracy. The satisfactory accuracy and small size of this model makes it suitable for real-time crop diagnosis in resource constrained mobile devices.

Due to the increasing complexity of electromagnetic signals, there exists a significant challenge for recognizing radar emitter signals. In this paper, a hybrid recognition approach is presented that classifies radar emitter signals by exploiting the different separability of samples. The proposed approach comprises two steps, namely the primary signal recognition and the advanced signal recognition. In the former step, a novel rough k-means classifier, which comprises three regions, i.e., certain area, rough area and uncertain area, is proposed to cluster the samples of radar emitter signals. In the latter step, the samples within the rough boundary are used to train the relevance vector machine (RVM). Then RVM is used to recognize the samples in the uncertain area; therefore, the classification accuracy is improved. Simulation results show that, for recognizing radar emitter signals, the proposed hybrid recognition approach is more accurate, and presents lower computational complexity than traditional approaches.

The performance of Global Satellite Positioning System / Inertial Navigation System (GPS/INS) integrated navigation system based on Kalman Filter (KF) is greatly influenced by measurement information related to GPS. However, it can be unreliable: it can be lost and the statistical characteristics of the measurement noise can change. Thus, the performance of navigation will get worse. Therefore, a filter algorithm for the integrated navigation system based on the Interacting Multiple Model-Adaptive Filter (IMM-AF) is proposed in this paper. Two measurement noise models for small Gaussian noise and non-small Gaussian noise are designed respectively to be applied to the algorithm; one step prediction algorithm for the case of GPS signal loss is also combined. The results of the experiment of the integrated navigation system of mobile robot show that, compared with KF or IMM, IMM-AF algorithm presents higher accuracy and better robustness, with almost the same update time.

Tracking and Data Relay Satellite System (TDRSS) is a space-based telemetry, tracking, and command system, which represents a research field of the international communication. The issue of the dynamic relay satellite scheduling, which focuses on assigning time resource to user tasks, has been an important concern in the TDRSS system. In this paper, the focus of study is on the dynamic relay satellite scheduling, whose detailed process consists of two steps: the initial relay satellite scheduling and the selection of dynamic scheduling schemes. To solve the dynamic scheduling problem, a new scheduling algorithm ABC-TOPSIS is proposed, which combines artificial bee colony (ABC) and technique for order preference by similarity to ideal solution (TOPSIS). The artificial bee colony algorithm is performed to solve the initial relay satellite scheduling. In addition, the technique for order preference by similarity to ideal solution is adopted for the selection of dynamic scheduling schemes. Plenty of simulation results are presented. The simulation results demonstrate that the proposed method provides better performance in solving the dynamic relay satellite scheduling problem in the TDRSS system.

In this paper, a Differential Chaos Shift Keying system based on Generalized Frequency Division Multiplexing (GFDM-DCSK) is proposed to achieve a system that has higher energy efficiency than DCSK system, solves the delay problem of DCSK system and decreases the decoding complexity of the GFDM system. Neither channel state information nor channel estimation circuits are needed at the receiver side. Numerous simulation results show that the proposed GFDM-DCSK system has better BER performance than DCSK system.

In this paper, we have studied the multiuser detection (MUD) technology applied in multiaccess (MA) direct sequence ultrawideband (DS-UWB) systems. Some typical MUD algorithms were analyzed and a new approximate optimal MUD algorithm based on searching in the local space of suboptimal minimum mean-squared error (MMSE) solution was proposed. In this new algorithm, the suboptimal solution of MMSE is considered as the initial value for the local search, and then a local solution space of this suboptimal solution is constructed by Euclidean distance. Within this space, some likely solutions for the classical optimal MUD algorithm (OMUD) are searched and the optimal one of them is chosen as the solution of this new optimal MUD algorithm. Theoretical analysis and simulation results indicate that, the BER performance and the Near-far Effect resistant ability of this proposed algorithm are much better than those of MMSE and adaptive MMSE detections, which are very close to those of OMUD. Furthermore, its time com...

In this letter, a mutated ant colony optimization (MACO) cognitive radio engine is proposed, and it is the first time to apply ACO algorithm to this problem. The cognitive radio is a promising technology nowadays to alleviate the apparent scarcity of available radio spectrum, and the cognitive radio engine determines the optimal radio transmission parameters for the system. The cognitive engine problem is usually solved by genetic algorithm (GA), however, the GA converges slowly and its performance can still be improved. Hence, MACO algorithm with excellent performance is applied to the cognitive engine in this letter. Simulation results show that the fitness scores obtained by the MACO engine are much better than the ACO and GA engines in different scenarios.

In this paper we present a population declining ant colony optimization (PDACO) multiuser detector for asynchronous CDMA communications. Ant colony optimization (ACO) algorithms have already been used in multiuser detection in CDMA systems; however, as the pheromone accumulates, we may not get a global optimum because it stops searching early. PDACO can enlarge searching range through increasing the initial population of the ant colony, and the population declines in successive iterations. So, the performance of PDACO is superior with the same computational complexity. PDACO is applied to multiuser detection in asynchronous CDMA systems over slowly multipath Rayleigh-fading channels in this paper. Via computer simulations it is shown that the performance of PDACO detector is much better in bit-error rate and near-far effect resistance than conventional detector, ACO detector and genetic algorithm detector, and is close the optimal multiuser detector.

In this paper, we present a robust H∞ distributed power control scheme for wireless CDMA communication systems. The proposed scheme is obtained by optimizing an objective function consisting of the user’s performance degradation and the network interference, and it enables a user to address various user-centric and network-centric objectives by updating power in either a greedy or energy efficient manner. The control law is fully distributed in the sense that only its own channel variation needs to be estimated for each user. The proposed scheme is robust to channel fading due to the immediate decision of the power allocation of the next time step based on the estimations from the H ∞ filter. Simulation results demonstrate the robustness of the scheme to the uncertainties of the channel and the excellent performance and versatility of the scheme with users adapting transmit power either in a user-centric or a network-centric efficient manner.

According to signal characteristic of Unified S-band (USB) measurement and control system, a method of spectrum sensing based on artificial neural network (ANN) is introduced in this paper. Spectrum sensing classifier is designed by the independent learning ability of ANN. The information collected by departed classifier is studied and stored to accurately judge that whether the master signal exists or not. With comparison and analysis, it can be proved that the spectrum sensing based on ANN possesses a higher accuracy.

This study mainly focuses on multiuser detection in tracking and data relay satellite (TDRS) system forward link. Minimum mean square error (MMSE) is a low complexity multiuser detection method, but MMSE detector cannot achieve satisfactory bit error ratio and near-far resistance, whereas artificial fish swarm algorithm (AFSA) is expert in optimization and it can realize the global convergence efficiently. Therefore, a hybrid multiuser detector based on MMSE and AFSA (MMSE-AFSA) is proposed in this paper. The result of MMSE and its modified formations are used as the initial values of artificial fishes to accelerate the speed of global convergence and reduce the iteration times for AFSA. The simulation results show that the bit error ratio and near-far resistance performances of the proposed detector are much better, compared with MF, DEC, and MMSE, and are quite close to OMD. Furthermore, the proposed MMSE-AFSA detector also has a large system capacity.

In this letter, a population declining mutated ant colony optimization (PDMACO) multiuser detector for MC-CDMA is proposed. The PDMACO algorithm is a hybrid ACO algorithm introducing population declining mechanism and mutation mechanism, which can achieve excellent performance by enlarging the searching range and avoiding local minima with almost the same computational complexity. Simulations show that its performance in reducing the bit-error rate and near-far effect is much better than that of the ACO multiuser detectors and close to that of the optimal multiuser detector.

In this paper, we propose a signal superposition layer multiplexing for the scalable video codec (SVC) coded video wireless broadcasting system. The SVC coded video layers are grouped into various groups and each group is encoded independently into a signal layer. The various signal layers are multiplexed using signal superposition and then transmitted. We introduce successive group decoder, originally used for interference channels, to decode the multiple signal layers from the received superimposed signals. Based on the optimal decoding order of the successive group decoder, we propose the signal layer allocation which guarantees that each receiver first decodes the signal layer encoded from the SVC base layer and then decodes those encoded from the SVC enhancement layers. From simulation results, the signal superposition SVC layer multiplexing shows significant peak signal-noise-ratio (PSNR) improvement for the reconstructed video, compared with the orthogonal channel SVC layer multiplexing, e.g., the time division multiple access (TDMA).

We present a traffic sign detection method which has won the first place for the prohibitory and mandatory signs and the third place for the danger signs in the GTSDB competition. The method uses the histogram of oriented gradient (HOG) and a coarse-to-fine sliding window scheme. Candidate ROIs are first roughly detected within a small-sized window, and then further verified within a large-sized window for higher accuracy. Experimental results show that the proposed method achieves high recall and precision ratios, and is robust to various adverse situations including bad lighting condition, partial occlusion, low quality and small projective deformation.