Search

With the computing power of High-Performance Computing (HPC) systems having stepped into the exascale era, more complex problems can be solved with scientific applications on a large scale. However, due to the significant performance gap between computing nodes and storage subsystems, suboptimal design for the Input/Output (I/O) module will significantly impede the efficiency of scientific applications, especially for the ubiquitous atmosphere applications. Two-phase I/O implemented in N-to-1 mode creates a serious bottleneck that hinders the scalability for the Model for Prediction Across Scales-Atmosphere (MPAS-A) on the new generation Sunway supercomputer. To address the I/O problem, we apply a custom data reorganization method to enable N-to-M I/O mode to exploit the parallel file system's performance and limit the data transfer among MPI ranks to a restricted scope to alleviate communication overhead. Moreover, we have conducted several methods to accelerate the computations, including the redesign for tracer transport, a hybrid buffering scheme, and a three-level parallelization scheme, which allows MPAS-A to use all heterogeneous computing resources efficiently. Experimental results show admirable scalability and efficiency of our I/O method, which achieves speedups of 41× and 58.9× for input and output compared with the raw I/O method on 30,000 MPI ranks. By scaling MPAS-A to 39 million heterogeneous cores, we demonstrate the necessity of a well-constructed I/O module for a real-world atmosphere application. Speed tests show that our optimization methods obtain good results for computations, and MPAS-A achieves a speed of 0.82 Simulated Day per Hour (SDPH) and 0.76 parallel efficiency of strong scaling with 600,000 MPI ranks. [ABSTRACT FROM AUTHOR]

This paper is the first to investigate both communication and control in traffic channel (TCH) and control channel (CCH) respectively when considering leader-follower formation keeping in UAV communication networks. In this paper, we analyze the relationship between the mutual interference and information exchange cost, and then formulate the joint channel and link selection problem as a two-way consensus game between CCH and TCH. To characterize the two-way choice of link selection, we creatively propose the generalized two-way consensus equilibrium (GTCE) to capture the stable state. Then, we prove that the formulated game has at least one pure-strategy GTCE which can maximize the UAV communication network utility. A distributed better reply based joint channel and link selection (BRJCLS) algorithm as well as two-dimensional minimum spanning tree (MST) based initialization (TMSTI) algorithm is proposed to achieve the GTCE. Simulation results are presented to show the convergence and effectiveness of the formulated two-way consensus game and proposed algorithms. [ABSTRACT FROM AUTHOR]

Aerial computing is a new paradigm where computing servers embedded in unmanned aerial vehicles are deployed to alleviate the offloading pressure of the core networks. In this article, we formulate an innovative aerial computing cycle to describe the aerial computing operations and classify the requirements and challenges in aerial computing. Then, Stackelberg game-based hierarchical framework is proposed to solve the inherent challenges in aerial computing networks from a distributed manner. A case study is proposed and discussed to verify the effectiveness of the proposed method. Finally, the future research directions are discussed to shed light on the potential applications. [ABSTRACT FROM AUTHOR]

Fast optical switches have the potential of lowering the cost, power, and latency of today’s data-center networks. The on-off switching of DFB lasers in arrays is poised to provide extremely fast switching because the on-off time of a DFB laser is the level of sub-nanosecond. However, the key of the DFB laser array based tunable laser lies in the high precision control of the lasing wavelength. In this letter, we proposed and experimentally demonstrated a tunable laser by integrating 16 DFB lasers based on the reconstruction-equivalent-chirp (REC) technique. Without using the conventional e-beam lithography, an equivalent $\lambda $ /4 phase-shift is fabricated in each DFB laser to ensure the single-longitudinal-mode (SLM) yield. Moreover, high precise wavelength spacing is realized. Wavelength deviations of 89.3% DFB lasers are within ±0.2 nm. The output power of all the channels is above 7 mW, and the relative intensity noise is below −130 dB/Hz. By on-off switching the lasers, sixteen channels with 2.4-nm spacing can be fast switched. The switching time between any two channels is less than 50 ns. In addition, a wide-band continuous wavelength tuning range of 38.4 nm can be obtained with a small temperature variation of 18 °C. Owing to the REC technique, the tunable laser via high-density integration of DFB lasers is demonstrated with simple fabrication, high SLM yield, and high wavelength precision, which opens up a path for wide-band and especially fast tunable lasers. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) have attracted great attention in the semantic segmentation of very-high-resolution (VHR) images of urban areas. However, large-scale variation of objects in the urban areas often makes it difficult to achieve good segmentation accuracy. Atrous convolution and atrous spatial pyramid pooling composed of atrous convolution can alleviate this problem by exploring multiscale contextual information. Unfortunately, atrous convolution causes gridding artifacts, where actual receptive fields are separated unit sets and fail to cover all the receptive fields. To address this problem, in this article, we first propose a self-smoothing atrous convolution (SS-AConv) that intrinsically improves atrous convolution, unlike existing methods. SS-AConv enhances sampling rates with low computational costs by adding several key parameters in the spatial dimension of its filter. Then, in the backbone, the Xception network, all max pooling operations are replaced with SS-AConvs for a large and effective receptive field. Moreover, to extract diverse features at multiple scales, we propose the SS-AConv cascaded multiscale structure (SCMS) by integrating SS-AConvs with different rates and the residual correction scheme (RCS) into a cascaded spatial pyramid. Finally, to extract diverse features at dense multiple scales, the SS-AConv convolutional network (SS-ACNet) is constructed by integrating SCMS into both the encoder and decoder layers of the modified Xception network. Extensive experimental results show that SS-ACNet outperforms some state-of-the-art methods on four open challenge datasets: the ISPRS Vaihingen and Potsdam, Cityscapes, and PASCAL VOC 2012. [ABSTRACT FROM AUTHOR]

Semantic segmentation labels each pixel in high-resolution remote sensing (HRRS) images with a category. To tackle with the large size and complexity of HRRS images, this letter presents a novel multiscale feature aggregation lightweight network (MFALNet) for semantic segmentation. Unlike standard convolution, asymmetric depth-wise separable convolution residual (ADCR) unit is used to reduce the parameter size of the network and makes the optimized structure deeper but lightweight and less complex. The proposed network is an encoder–decoder structure, where multiscale feature aggregation is implemented in both the encoder and the decoder. The spatial self-attention block helps to capture long-range contextual information, and the gated convolution modules are further used for refining features when aggerating high- and low-level feature maps in the decoder. The proposed MFALNet has evaluated on the International Society for Photogrammetry and Remote Sensing (ISPRS) Vaihingen and Potsdam 2-D semantic labeling contest open benchmark data set, and the experimental results prove that the scheme can obtain a better tradeoff between segmentation accuracy and computational efficiency compared with the state-of-the-art semantic segmentation models. [ABSTRACT FROM AUTHOR]

In this paper, we investigate the content delivery in edge caching networks where requesters can get the interested contents from surrounding helpers by device-to-device (D2D) communications. Taking the finite service capacity and cache space of helpers into account, the problem is to maximize the number of requesters obtaining all desired contents by optimizing the selection of content sources. The hypergraph theory is first applied to analyze the cumulative content effect caused by multiple content sources, and the problem is formulated as a hypergraph based cooperation cache game with at least one pure Nash equilibrium (NE). The log-linear based cooperation caching algorithm (LBCC) is proposed to find the corresponding NE points. Simulation results verify the superiority of the proposed LBCC. [ABSTRACT FROM AUTHOR]