Search

To improve spectral efficiency of Device-to-Device multicast (D2DM) communication, we study the joint optimization of subcarrier and power allocation in D2DM communications underlaying cellular system with extremel large-scale multi-input-multi-output (XL-MIMO) in this paper. Specifically, we aim to maximize the aggregate multicast data rate of all D2D groups (DGs) while guaranteeing the data rate requirements of CUs, where each subcarrier can be allocated to multiple DGs and each DG can reuse multiple subcarriers. This formulated problem leads to mixed integer non-linear programming (MINLP), which is usually NP-hard and computationally intractable. Therefore, we decompose it into two sub-problems, i.e., subcarrier assignment and power allocation. we design a heuristic subcarrier assignment scheme for allocating subcarriers to CUs and DGs by equally distributing the transmit power of CUs and DGs over all subcarriers. After that, we design an effective power allocation algorithm by adopting the “difference of two concave functions” (D.C.) programming to transform the nonconcave functions into concave ones. Extensive simulations and numerical results demonstrate that our joint resource allocation algorithms outperform the other baseline methods.

Aiming at the problem that the existing algorithms are difficult to segment efficiently in indoor scenes due to the high similarity of stacked and closely adjacent objects, this paper proposes an indoor point cloud segmentation algorithm based on hierarchical feature fusion. Firstly, according to the results of stratification, the principal component analysis method is adopted to reduce the dimension of each layer point clouds, and then the decision rules on the two-dimensional plane are established based on the structural characteristics of each layer objects. Combined with the analysis of hierarchical connectivity, fusing the features of each layer, so as to obtain the global positioning of the object. Last, for further division, the density segmentation method based on inter cluster constraints are taken into consideration. The experimental results show that the location and segmentation effect of stacked objects is fine, it is worth mentioning that this method has high accuracy, and the average mIoU can reach 0.811. In addition, another advantage of this method is that it takes less time to process point clouds, the average consumption time is about 6.59 seconds. These characteristics represent that this method can better complete the segmentation of indoor scenes.

A multimode resonance patch antenna with the attractive radiation gain and efficiency is investigated in this paper. A driven patch with both sides shorted similar to the cavity model is presented to generate TM03 and TM11 modes. TM21 mode is also excited by inserting five shorting parasitic patches on both sides of the radiation aperture to extended the impedance bandwidth and no additional feeding structure is required. Meanwhile, a novel metasurface with the positive gradient of the reflection phase is placed directly above the proposed multimode antenna. Then the characteristics of high gain and better front-to-back ratio (FBR) are acquired. The measured results show that the characteristics of 20.7% impedance bandwidth from 5.08 GHz to 6.25 GHz, stable gain of up to 12.34 dBi, high radiation efficiency of up to 89% and cross-polarization level below −20 dB are obtained. These characteristics all imply that the proposed antenna can be applied to WLAN and Car-to-Car (C2C) communications.

Integrated methods are used in the modification of trajectory, including improved perturbation impact point deviation prediction, adaptive proportional guidance and adaptive proportional differential guidance, thus improving the firing accuracy of guided mortar shell. The six degrees of freedom of both the trajectory model and the control model were established, and their guidance laws were designed based on the three guidance schemes. Firstly, the perturbation impact point deviation prediction method is improved by setting up a discrimination factor in the rising phase of trajectory based on the principle of traditional perturbation impact point deviation prediction method and in combination with the trajectory characteristics of guided mortar, which further improves the correction efficiency. The adaptive proportional guidance law is designed in the longitudinal plane, while the adaptive proportional differential guidance law is designed in the transverse plane due to the fact that the constant proportional coefficient in the proportional guidance law does not conform to the requirements of actual trajectory. In this paper, Monte Carlo simulation method and ammunition flight test are used to verify the designed guidance law, and the simulation results illustrate that the integrated guidance method is both reasonable and effective. As for the error of the guidance tool and actuator, the method is available for the reduction of impact point deviation and the improvement of accuracy. The circular error probability (CEP) not under control decreases from 126.317m to 10.1284m when control is applied. Besides, the feasibility of the designed guidance law is verified by the flight tests of guided mortar projectile in large, medium and small range respectively from the perspective of engineering application. It can be seen from the test trajectory impact point data that the guidance law is available for the effective correction of trajectory deviation in the actual hardware operation and site environment with reliable guidance. As the outcome, the CEP reaches 10.86m, and the impact point deviation of some guided missiles is within 2m.

Starting from the car suspension system, the nonlinear characteristics of a class of two-degree-of-freedom oscillators with strong nonlinearity under a periodic excitation are discussed by the switching theory of flow in discontinuous dynamical systems. Based on the discontinuous forces and different motions of the two masses, the phase plane of each mass is composed of stick domain, nonstick domain (or free domain) and separation boundary in absolute and relative coordinates, respectively. The swithching criteria between the stick and nonstick motions and the conditions of grazing motion in two different regions are developed via the G-functions and switching control laws. The mapping dynamics theory is used to give the four-dimensional transformation set and four-dimensional mapping, and the conditions for periodic motions are explored. In addition, the stick motions, two kinds of grazing motions, periodic motions for this system and a comparison of the velocities, accelerations (or forces responses) of the two masses under the two conditions of control force are simulated numerically. The results show that the stability and comfort of the vehicle can be improved by adjusting the control force, which is generated by the control unit of system or exerted by the external excitation. For further investigating the influence of system parameters on dynamical behaviors, the stick and grazing bifurcation scenarios varying with driving frequency or amplitude are also developed, which can provide useful information for parameter selection of vibration systems with clearance and the optimal design of vehicle suspension systems. This paper also has important reference value for practical applications in other industries or machinery with elastic impacts.

Diffusion Tensor Imaging (DTI) reveals subtle abnormalities associated with stroke, multiple sclerosis, schizophrenia and dyslexia, which has a broad application prospect in the medical field. The densely sampled 3-D DTI fiber tracts in biological specimens have high geometric, spatial and anatomical complexity. To provide users with more immersive and convenient interactions in exploring DTI fibers, we design specific interactions based on the APIs of Leap Motion. Leap Motion is a somatosensory interaction device focusing on hand tracking. We design four different interaction modes for users to analyze the data in different interaction stages and scenarios, in order to better explore the DTI fibers which users are interested in by Leap Motion gestures. They are Normal Mode, Box Basic Interaction Mode, Box Logic Operation Mode, and Cluster Exploration Mode. Users can conduct tradition manipulations over the whole DTI fiber data in Normal Mode. Boxes are employed to filter out uninteresting tracts in Box Basic Interaction Mode, and expression-based queries are further designed to get logic set operations based on multiple boxes in Box Logic Operation Mode, e.g., the intersection, union and complement of boxes. Complex logic combination queries are allowed to perform to reduce visual clutter and help users explore DTI fibers more precisely. In Cluster Exploration Mode, DTI fibers can be classified by clustering them into spatially and anatomically related tracts and then different clusters can be explored individually by designed gestures. Compared with the explorations through traditional input devices, the evaluation tests show that the proposed approach is more intuitive and efficient in 3-D explorations and provides an immersive experience for users to explore the DTI fiber data.

This paper studies the constrained optimization problem for nonlinear diesel blending. A new hybrid algorithm called cultural harmony search algorithm is presented to solve the proposed optimization problem, which uses cultural knowledge in the belief space of the cultural algorithm to guide the evolving and searching process of the harmony search algorithm. Then, an improved harmony improvisation in the population space of cultural algorithm is developed for new harmony generation to enrich the population diversity. Moreover, in order to accelerate convergence, the domain of decision variables is scaled down by a simplex method at the beginning of the algorithm, and a simplex improved cultural harmony search algorithm is provided. Finally, benchmark functions and the results of application in nonlinear diesel blending of a real-world refinery show the feasibility and effectiveness of the proposed algorithms. The contrasted experiments show that our proposed hybrid algorithm is better than other hybrid algorithms, especially in diesel blending optimization problem.

In recent years, the methods of no-reference stereoscopic image quality assessment (NR-SIQA) have been well investigated, but there still remain challenges due to the inaccurate extraction of binocular perception information. In this paper, we propose an NR-SIQA method based on visual attention and perception. We combine saliency and just noticeable difference (JND) to model visual attention and perception, respectively, and weight the global and local features extracted from the left and right views. Meanwhile, in order to obtain the accurate binocular perception information, the global structural features reflecting spatial correlation are extracted from the cyclopean map that is synthesized by the left and right views. Then, a regression model is learned based on a support vector machine regression (SVR) to evaluate the quality of stereoscopic images. The experiments on popular SIQA datasets demonstrate that the proposed NR-SIQA method has better and more reliable performance than the state-of-the-art methods.

Single-layer graphene (SLG) sheets can exhibit thermoelectric properties under the control of gate voltage. The controlled factors and regulation mechanism of SLG thermoelectric properties have become research hotspots. In this paper, a SLG thermoelectric parameter model considering carrier concentration and mobility with temperature and gate voltage is proposed. Based on the proposed model, the square resistance (Rs) and Seebeck coefficient (S) of the SLG are calculated. The results show that the maximum value of Rs decreases from 5.8 KΩ to 3.2 KΩ at the Dirac voltage when the temperature increases from 100 K to 500 K. A large and stable S can be obtained at high voltages and temperatures. The maximum value of S can reach 161.3 μV/K at T = 500 K, exhibiting a more obvious thermoelectric characteristic. Simultaneously, the saturation law of the power factor (Q) with the change of gate voltage and the amplitude regulation of Q by temperature are obtained. This work can provide a theoretical basis for analyzing the thermoelectric characteristics of SLG.

We present an efficient approach for real-time continuous human action recognition with depth sensors. Instead of using the powerful but quite complex deep neural networks, our approach uses a lightweight discriminative frame-level descriptor, which is called the context pose (CP). The objective is to make our approach realistic in mobile depth sensor applications. CP integrates the context of the motion within a depth video. CP can increase the discriminative power of the frames and enable more frames to represent actions. CP is incorporated with a new part-based random decision forest (PRDF) method. The PRDF is designed to automatically select the optimal combination of body parts to represent and distinguish each action. We evaluate our approach on three classical single-action benchmark datasets. The experiments show that our approach has 200 frames/s and wins superior performances to the existing frame-level descriptors and classifiers in terms of accuracy.

Traditional recommender systems often suffer from the problem of data sparsity, because most users rate only a few of the millions of possible items. With the development of social platforms, incorporating abundant social relationships into recommenders can help to overcome this issue, because users' preferences can be inferred from those of their friends. Most existing social recommenders are based on matrix factorization, a collaborative filtering model that has been proven to be effective. In this paper, we introduce a novel social recommender based on the idea that distance reflects likability. Compared with matrix factorization, the proposed model enables us to obtain a spatial understanding of the latent factor space and how users and items are positioned inside the space by combining the factorization model and distance metric learning. In our method, users and items are initially mapped into a unified low-dimensional space. The positions of users and items are jointly determined by ratings and social relations, which can help to determine appropriate locations for users who have few ratings. Finally, the learned metrics and positions are used to generate understandable and reliable recommendations. Experiments conducted on real-world data sets have shown that compared with methods based on only matrix factorization, our method significantly improves the recommendation accuracy.

As a significant application of energy, smart grid is a complicated interconnected power grid that involves sensors, deployment strategies, smart meters, and real-time data processing. It continuously generates data with large volume, high velocity, and diverse variety. In this paper, we first give a brief introduction on big data, smart grid, and big data application in the smart grid scenario. Then, recent studies and developments are summarized in the context of integrated architecture and key enabling technologies. Meanwhile, security issues are specifically addressed. Finally, we introduce several typical big data applications and point out future challenges in the energy domain.