Search

Lifting-Automated Guided Vehicle (L-AGV) has been widely used in the fourth-generation automated port for its advantages on time-savings. Optimize the port L-AGV scheduling problem has dramatically significant impacts on port working efficiency. An improved Artificial Bee Colony algorithm (IABC) was proposed to greatly solve the problem by introducing ideas of genetic algorithm (GA) and adaptive factors. Compared with original Artificial Bee Colony algorithm (ABC) and GA, IABC made better performances on solutions and convergence, which also provided a direction on decision optimization for port AGV scheduling problem related.

The opening of the Arctic route will greatly shorten the distance of navigation and save costs. This research was mainly focus on hierarchical hub locations for rail-sea container transportation network system planning considering the Arctic Route. According to the characteristics of integrated network including the Arctic route, a new mathematical model was established by juxtaposing the traditional idea of hierarchical hub and hub arc location. Benders decomposition method was used to transform and solve the mixed integer nonlinear model. Based on the analysis of the Arctic route example, it could be indicated that the number of central hubs directly affects the transportation cost and location of hubs. This study provides a certain direction for a shipping company to choose hubs after the opening of the Arctic route in the future and also made sense for the rail-sea intermodal transportation system from a strategic consideration.

In this paper, we propose a novel and simple transmitter selection criterion to enhance the reliability performance of downlink transmission in a wireless vehicular network. In this network, the Manhattan-type urban street model is adopted such that the location of horizontal and vertical streets is generated by two independent and identical Poisson Point Processes (PPPs). Moreover, the location of vehicles on each street is modeled by a one-dimensional PPP and base stations (BSs) are located at the intersection of streets. According to the proposed criterion, the vehicle receives signals from either the nearest front vehicle via dedicated short range communications or the nearest front BS via cellular communications. Considering the generalized Nakagami- $m$ fading, we derive a new easy-to-compute expression for the coverage probability of the signal-interference-plus-noise ratio of the vehicle at the origin. Aided by numerical results, we demonstrate the accuracy of the derived expression and explicitly show the performance advantage of the proposed criterion. In addition, we find that there is an optimal vehicle intensity to maximize the coverage probability.

In this paper, we propose a new transmission scheme in the multi-source system to improve the freshness of the received packets. In our proposed scheme, each source generates its packet through an independent and identical Bernoulli process at the beginning of each time slot, and then randomly selects one channel from multiple orthogonal frequency channels to perform repeat transmission of its packet across multiple time slots. We then analyze the average age of information (AoI) to characterize the freshness of the received packets. Specifically, we derive new closed-form expressions for the average AoI achieved by the proposed transmission scheme for three packet management policies. The simulation results corroborate our analysis and demonstrate that the proposed transmission scheme can achieve a lower average AoI than the existing scheme for the three considered packet management policies. In addition, the impact of system parameters, such as the packet generation probability, the numbers of time slots and orthogonal frequency channels, and the channel error probability, on the average AoI is examined.

The scarcity of spectrum resources in current wireless communication systems has sparked enormous research interest in the terahertz (THz) frequency band. This band is characterized by fundamentally different propagation properties resulting in different interference structures from what we have observed so far at lower frequencies. In this paper, we derive a new expression for the coverage probability of downlink transmission in THz communication systems within a three-dimensional (3D) environment. First, we establish a 3D propagation model which considers the molecular absorption loss, 3D directional antennas at both access points (APs) and user equipments (UEs), interference from nearby APs, and dynamic blockages caused by moving humans. Then, we develop a novel easy-to-use analytical framework based on the dominant interferer analysis to evaluate the coverage probability, the novelty of which lies in the incorporation of the instantaneous interference and the vertical height of THz devices. Our numerical results demonstrate the accuracy of our analysis and reveal that the coverage probability significantly decreases when the transmission distance increases. We also show the increasing blocker density and increasing AP density impose different impacts on the coverage performance when the UE-AP link of interest is in line-of-sight. We further show that the coverage performance improvement brought by increasing the antenna directivity at APs is higher than that brought by increasing the antenna directivity at UEs., Accepted to appear in IEEE ICC 2020

Interference detection using deep neural network has recently received increasing attention due to its capability in learning rich features of data. In this paper, we proposed a low-complexity blind interference detection method. Our method operates on Time-frequency overlapped interference signal and can separate which from received signal. The proposed method uses AutoEncoder network to reconstruct the transmitted signal and separate interference signal. The AutoEncoder network consists of several layers of recurrent neural network (RNN) which is well-suited for learning representations from time-correlated data. Simulation results show that the separated interference signal has the same features of original interference, so it not only achieves good interference detection performance, but also can realize interference recognition.

Due to the randomness of wireless channels, it is possible to achieve secure communications by designing communication systems according to the characteristics of wireless channels. In this work, a physical layer security communication mechanism based on end-to-end learning is introduced. The proposed mechanism achieves secure communications by designing the loss function of end-to-end learning. We evaluate secrecy performances of the proposed mechanism through block error rate (BLER) under three classic circumstances which different decoding models are applied by the eavesdropper. Self-training decoding model, the legitimate decoding model and unsupervised learning (K-means clustering) model are alternative for the eavesdropper. The experiments results show that K-means clustering model can get the best BLER performance for the eavesdropper with some prior information, but its BLER is still higher than 0.5. Even if the eavesdropper has a much higher than the legitimate users, there is no improvement in its BLER traditional physical layer security E s /N 0 . Compared to methods, the proposed mechanism can achieve secure communications without extra key extraction.

Histopathology image analysis plays a critical role in cancer diagnosis and treatment. To automatically segment the cancerous regions, fully supervised segmentation algorithms require labor-intensive and time-consuming labeling at the pixel level. In this research, we propose CAMEL, a weakly supervised learning framework for histopathology image segmentation using only image-level labels. Using multiple instance learning (MIL)-based label enrichment, CAMEL splits the image into latticed instances and automatically generates instance-level labels. After label enrichment, the instance-level labels are further assigned to the corresponding pixels, producing the approximate pixel-level labels and making fully supervised training of segmentation models possible. CAMEL achieves comparable performance with the fully supervised approaches in both instance-level classification and pixel-level segmentation on CAMELYON16 and a colorectal adenoma dataset. Moreover, the generality of the automatic labeling methodology may benefit future weakly supervised learning studies for histopathology image analysis., 10 pages, 9 figures, accepted by ICCV 2019

Suffering from complex sideslip angles, path following control of an under actuated surface vehicle (USV) becomes significantly challenging and remains unresolved. In this paper, a finite-time observer based guidance and control (FOGC) scheme for path following of an USV with time-varying and large sideslip angles and unknown external disturbances is proposed. The salient features of the proposed FOGC scheme are as follows: 1) time-varying large sideslip angle is exactly estimated by a finite-time sideslip observer, and thereby contributing to the sideslip-tangent line-of-sight guidance law which significantly enhances the robustness of the guidance system to unknown sideslip angles which are significantly large and time-varying; 2) a finite-time disturbance observer (FDO) is devised to exactly observe unknown external disturbances, and thereby implementing FDO-based surge and heading robust tracking controllers, which possess remarkable tracking accuracy and precise disturbance rejection, simultaneously; and 3) by virtue of cascade analysis and Lyapunov approach, global asymptotic stability of the integrated guidance-control system is rigorously ensured. Simulation studies and comparisons are conducted to demonstrate the effectiveness and superiority of the proposed FOGC scheme.

The performance of traditional deep learning-based signal detection methods have been challenged by real world signal impaired by random noise and interferences through wireless channel, they quite often appeal to a prior filter for interference elimination before deep learning model. To deal with the problem, we introduce an embeddable convolutional layer-based filter (Conv-based filter), which takes raw time domain signal as input and can adaptively learn characteristics of the band-pass filter whose center frequency and bandwidth are compatible with the target signal. To enhance the performance of learned filters, the attention mechanism is introduced by using Squeeze-and-Excitation block (SE-block) after Conv-based filters. As an embeddable block, the filter is trained end-to-end in a deep learning network, no explicit assumptions about the raw data are made. For application of signal detection, compared with the way of signal preprocessing and then training separately, embedding the preprocessing as a block into the deep learning network works better.

In this paper, we propose a transmission control scheme and design an approximate message passing (AMP) algorithm for the joint user identification and channel estimation (JUICE) in massive connectivity networks. In the proposed transmission control scheme, a transmission control function is designed to determine a user’s transmission probability, when it has a transmission demand. By employing a step transmission control function for the proposed scheme, we derive the channel distribution experienced by the receiver to characterize the effect of transmission control on the design of AMP algorithm. Based on that, we propose a minimum mean squared error denoiser and design the AMP algorithm. We demonstrate that the proposed scheme can significantly improve the JUICE performance and reduce the average delay, compared to the conventional scheme without transmission control.

Subject to complex uncertainties, a novel surge-varying adaptive LOS (SALOS) based path following control scheme is addressed for underactuated marine vehicles (UMVs) in this paper. Main contributions are as follows: (1) The surge velocity and heading angle are guided with path following errors, simultaneously, and the unknown sideslip angle is estimated by an adaptive term online. (2) Complex unknown uncertainties including the internal parameter perturbations, external environmental disturbances and unmodelled dynamics are compositely estimated by the constructed fuzzy uncertainty observer (FUO); (3) Robust surge and heading tracking controllers based on the FUO are developed to ensure that the tracking errors and uncertainty observation errors are uniformly ultimately bounded (UUB). Simulation studies are conducted to demonstrate the effectiveness and remarkable performance of the proposed SALOS scheme.

As the most common non-deadlock concurrency bugs, atomicity violations are extremely hard to detect during testing since the exhaustive testing of a multi-threaded program is impossible because of the large number of interleavings. The studies in recent years have mainly focused on single-variable atomicity violation. However, those methods are unable to predict or find atomicity violations with multiple variables involved. Many variables are inherently correlated and need to be accessed together with their correlated peers in a consistent manner. These variables need to be either updated together consistently or accessed together to avoid inconsistent update or reading. This paper presents a method for predicting two-variable atomicity violation, based on access correlation between variables and atomicity violation pattern of variable accesses, including algorithms to infer access correlation between variables and to predict atomicity violation using model checking. The effectiveness of our method is evaluated with several real-world systems.

Subject to complex unknowns, a novel surge-guided adaptive LOS (SALOS) based path following control scheme is addressed for underactuated marine vehicles (UMVs) in this paper. Main contributions are as follows: (1) Based on the LOS guidance method with adaptive sideslip compensation, the guided surge velocity adapted to cross-track error is innovatively incorporated, and thereby significantly improving the decision-making level and robustness of the guidance system; (2) Complex unknowns are compositely observed by the constructed finite-time unknown observer (FUO), and the unknown observation errors are stabilized to be origin in a finite time; (3) Combined with finite-time unknown observation, the robust surge and heading tracking controllers are developed with high accuracy and finite-time unknown rejection. Simulation studies are conducted to demonstrate the effectiveness and remarkable performance of the proposed SALOS scheme.

Suffering from complex unknown disturbances, a novel surge-varying LOS (SVLOS) based path following control scheme of underactuated marine vehicles (UMVs) is proposed in this paper. Main contributions are as follows: (1) The surge velocity guided with cross-track error is incorporated into previous line-of-sight guidance laws, and thereby significantly enhancing robustness and agility of the guidance system; (2) Complex unknown disturbances are compositely estimated by the constructed finite-time disturbance observer (FDO), and the observation errors are ensured to be zero in a finite time; (3) Robust surge and heading tracking controllers based on the FDO design and backstepping technique are developed so that reference surge and heading can be exactly tracked with complex unknown disturbances. Simulation studies are conducted to demonstrate the effectiveness and remarkable performance of the proposed FDO-SVLOS scheme.

For an underactuated marine vehicle (UMV) with unmodelled dynamics and disturbances, a novel fuzzy uncertainty observer based path following control (FUO-PFC) scheme is proposed. Main contributions are as follows: (1) A surge-varying line-of-sight (SVLOS) guidance law is devised innovatively where the guided surge velocity adapts to cross-track error, and thereby significantly enhancing robustness and agility of the guidance system; (2) Unknown nonlinearities including unmodelled dynamics and disturbances are compositely estimated by the constructed fuzzy uncertainty observer (FUO); (3) Robust surge and heading tracking controllers based on the FUO are developed to ensure that tracking and observation errors are uniformly ultimately bounded (UUB). Simulation studies are conducted to demonstrate the effectiveness and remarkable performance of the proposed FUO-PFC scheme.

Compressive sensing (CS) is a viable source of an innovative 5G system, what's more, it's an effective technology to deal with the data redundancy problem of massive machine-to-machine communication (MMC), since it enables the recovery of sparse and approximately sparse signals with significantly fewer samples than demanded by Nyquist-Shannon sampling theory. Interference in signal will lead a series problem to signal processing, which will be a recovery error proportional to the interference energy in CS. This paper tries to mitigate interference by proposing a compressive interference pre-filter based on linear minimal mean square error (LMMSE). The main contributions are that two practical estimation methods that based on autocorrelation function and auto-covariance function respectively, are applied to estimate the LMMSE-based filtering matrix. The estimation and pre-filter algorithms can be practically integrated into the compressive signal processing framework with improved robust property to noise and interference.

The pilot real-time perception of the external terrain environment is improved to avoid the collision of civil aircraft. The Synthetic Vision System(SVS) using the three-dimensional(3D) computer image generation technology is applied to provide pilots with surrounding terrain image and comprehensive information display. During the flight, a quick, accurate and effective technology of invocation of terrain data is required for large-scale data transfer. In this paper, a dynamic terrain-data invocation method including block partition call, pre-call and visual cone tailoring technology is proposed. The terrain elevation data is invoked from the corresponding blocks according to block partition call. The butter technology is used to speed up for rendering in the pre-call. Data is selected before display with the visual cone tailoring technology. The result shows that the efficiency of SVS is obviously raised, 3D terrain is showed in real time, and the improvement of visual perception is achieved.

In hyperspectral image classification, labeled samples are usually limited and unlabeled data are available in large quantities. This paper presents a novel semi-supervised feature reduction methods, Semi-supervised Locality Preserving Discriminant Analysis (SLPDA), in order to solve the classification problem of hyperspectral imagery. The distinguished feature of the algorithm is the combination of the two graph-based dimensionality reduction methods, supervised LSDA and unsupervised NPE, in a way without extra tuning parameters. The proposed method preserves local geometrical structure in the labeled and unlabeled data, and discriminates data of different classes in labeled samples as far as possible. Experimental results demonstrated that our method had better performance than other traditional dimensionality reduction methods in data sets Indian Pine and Kennedy Space Center.

High level Petri nets (HLPNs) have been widely applied to model concurrent and distributed systems in computer science and many other engineering disciplines. However, due to the expressive power of HLPNs, they are difficult to analyze. In recent years, a variety of new analysis techniques based on model checking have been proposed to analyze high level Petri nets in addition to the traditional analysis techniques such as simulation and reachability (coverability) tree. These new analysis techniques include (1) developing tailored model checkers for particular types of HLPNs or (2) leveraging existing general model checkers through model translation where a HLPN is transformed into an equivalent form suitable for the target model checker. In this paper, we present a term rewriting approach to analyze a particular type of HLPNs -- predicate transition nets (PrT nets). Our approach is completely automatic and implemented in our tool environment, where the frontend is PIPE+, a general graphical editor for creating PrT net models, and the backend is Maude, a well-known term rewriting system. We have applied our approach to the Mondex system -- the 1st pilot project of verified software repository in the worldwide software verification grand challenge, and several well-known problems used in the annual model checking contest of Petri net tools. Our initial experimental results are encouraging and demonstrate the usefulness of the approach.

In this paper, we propose a novel detection algorithm for random multiple access, which combines slot based decoding and frame based decoding to recover users' messages. For the slot based decoding, we exploit a linear physical-layer network coding (LPNC) algorithm to decode the linear combinations of collided packets at each slot. By designing the coefficient matrix of linear combinations, we have the decoded linear combinations independent of each other in each slot. At frame level, we have equations constructed by linear combinations from all slots within one frame, and we adopt Gaussian elimination aided successive interference cancellation (GE-SIC) decoding method to recover users' messages. We show that our LPNC algorithm can increase the number of linear combinations per slot by minimizing the decoding error probability. In addition, we demonstrate that the proposed scheme can significantly improve the system performance in terms of throughput and packet loss rate.

© 2016 IEEE. In this paper we investigate the coded slotted ALOHA (CSA) schemes with repetition codes and maximum distance separable (MDS) codes over erasure channels. We derive the extrinsic information transfer (EXIT) functions of the CSA schemes over erasure channels, which allow an asymptotic analysis of the packet recovering process. Moreover, we define a traffic load threshold provided that the recovered probability is more than a given recovery ratio. The optimal distribution of the codes chosen by users in the CSA schemes is then designed to maximize the peak throughput and traffic load threshold. By performing the asymptotic analysis, we show that our optimal distributions improve the traffic load threshold by 60% for ϵ = 0.1 and 86% for ϵ = 0.135 compared to the optimal distribution for collision channels. Using repetition codes as an example, simulation results show that the obtained distributions enhance the peak throughput for erasure channels when both packet erasure channels and slot erasure channels are considered.

Compressive Sensing (CS) enables us to exactly reconstruct a signal from a small number of observations if it has a sparse representation in a known basis. But in the situation where the sparsity is not strictly satisfied as the signal contains Gaussian noise, the general practice tries to directly recover the noisy signal, and then filters out the noise from the recovered signal. Because the existence of noise may reduce the reconstruction accuracy, to overcome this drawback, we propose a noise reduction method which preprocesses the signal in compressed domain based on minimum mean square error (MMSE). The method can efficiently suppress the noise but still keep plenty of energy of the desired signal. Both analyses and simulations indicate that the modified method can improve the performance of reconstruction.

Longitudinal brain image series offers the possibility to study individual brain anatomical changes over time. Mathematical models are needed to study such developmental trajectories in detail. In this paper, we present a novel approach to study the individual brain anatomy over time via a linear geodesic shape regression method. In our method, we integrate separate pairwise registrations between the baseline image and the follow-up images into a unified spatial registration plus temporal regression framework. Different from previous geodesic shape regression approaches, which use the LDDMM framework to estimate the brain anatomical change over time, our method is based on the LogDemons method to decrease the computation cost, while maintaining the diffeomorphic property of the deformation over time. Moreover, a temporal regression constraint is explicitly implemented in each optimization iteration to make sure that the entire developmental trajectory can be compactly represented by the baseline image and an optimal stationary velocity field. Our method is mathematically well founded in the Alternating Direction Method of Multipliers (ADMM), which for our image regression application is interpreted in diffeomorphic space instead of Euclidean space. We evaluate our new method on 2D synthetic images and real 3D brain longitudinal image series, and the experiments show promising results in regression accuracy as well as estimated deformations.

With the widespread use of networked control system, the research on ship networked is becoming increasingly important. In this paper, the ship networked control system based on semi-physical simulation platform is introduced, and simulates it by truetime network toolbox. To solve the problems of random delay and packet dropout in the networked control system, the fuzzy PID algorithm is used to simulate the model, and the simulation results demonstrate the feasibility of this algorithm.

Compressed sensing theory can be applied to reconstruct the signal with far fewer measurements than what is usually considered necessary. While for the classification of modulated signals, we only expect to acquire some characteristics rather than the original signal. However, to select the feature used for modulation classification with sparsity is the main challenge. In this paper, we propose a method to identify the linear modulation format of an unknown single carrier linear digital signal using compressive samples, without reconstructing the original signal. In our method, we construct a compositional feature of multiple high-order moments of the received data as the identification characteristic. From simulations we can see that the method is effective, even at a relatively low signal-to-noise ratio.

For hyperspectral data classification, feature reduction techniques have become an apparent need to extract information from original data. In this paper, we introduce Locality Sensitive Discriminant Analysis (LSDA) to perform feature reduction for classification of hyperspectral imagery. By preserving both the discriminant and local geometrical structure in the data, the proposed method can obtain better classification accuracy with less computational load. Experiments carried out on two data sets, Indian Pine and Pavia City, show that results of our method outperform that of other traditional dimensionality reduction methods.

We introduce a method for the automatic generation of magnetic resonance imaging brain atlases from a set of brains without any prior on their structure and number. This is a bottom-up, data-driven approach whose goal is to generate brain atlases that mirror as closely as possible the properties of arbitrary brains. This is important, for example, in the segmentation of brain structures of highly deformed brains as that of advanced Alzheimer's disease patients. The method is as follows. First, we register all brain pairs and obtain similarity weights. Second, we construct a graph whose nodes represent volumes and the links the similarity weights. Third, we use Dijkstra's method to select atlas candidates. Fourth, we iteratively merge clusters of atlas candidates until a good subset of them remains. We applied our method to a set of 61 ADNI brain volumes and found that a good selection could be made.