Photonic integrated neuro-synaptic core for convolutional spiking neural network

Neuromorphic photonic computing has emerged as a competitive computing paradigm to overcome the bottlenecks of the von-Neumann architecture.Linear weighting and nonlinear spike activation are two fundamental functions of a photonic spiking neural network(PSNN).However,they are separately implemented with different photonic materials and devices,hindering the large-scale integration of PSNN.Here,we propose,fabricate and experimentally demonstrate a photonic neuro-synaptic chip enabling the simultaneous implementation of linear weighting and nonlinear spike activation based on a distributed feedback(DFB)laser with a saturable absorber(DFB-SA).A prototypical system is experimentally constructed to demonstrate the parallel weighted function and nonlinear spike activation.Furthermore,a fourchannel DFB-SA laser array is fabricated for realizing matrix convolution of a spiking convolutional neural network,achieving a recognition accuracy of 87%for the MNIST dataset.The fabricated neuro-synaptic chip offers a fundamental building block to construct the large-scale integrated PSNN chip.

Structural knowledge-driven meta-learning for task offloading in vehicular networks with integrated communications,sensing and computing

Task offloading is a potential solution to satisfy the strict requirements of computation-intensive and latency-sensitive vehicular applications due to the limited onboard computing resources.However,the overwhelming upload traffic may lead to unacceptable uploading time.To tackle this issue,for tasks taking environmental data as input,the data perceived by roadside units(RSU)equipped with several sensors can be directly exploited for computation,resulting in a novel task offloading paradigm with integrated communications,sensing and computing(I-CSC).With this paradigm,vehicles can select to upload their sensed data to RSUs or transmit computing instructions to RSUs during the offloading.By optimizing the computation mode and network resources,in this paper,we investigate an I-CSC-based task offloading problem to reduce the cost caused by resource consumption while guaranteeing the latency of each task.Although this nonconvex problem can be handled by the alternating minimization(AM)algorithm that alternatively minimizes the divided four sub-problems,it leads to high computational complexity and local optimal solution.To tackle this challenge,we propose a creative structural knowledge-driven meta-learning(SKDML)method,involving both the model-based AM algorithm and neural networks.Specifically,borrowing the iterative structure of the AM algorithm,also referred to as structural knowledge,the proposed SKDML adopts long short-term memory(LSTM)networkbased meta-learning to learn an adaptive optimizer for updating variables in each sub-problem,instead of the handcrafted counterpart in the AM algorithm.Furthermore,to pull out the solution from the local optimum,our proposed SKDML updates parameters in LSTM with the global loss function.Simulation results demonstrate that our method outperforms both the AM algorithm and the meta-learning without structural knowledge in terms of both the online processing time and the network performance.

Chip-Based High-Dimensional Optical Neural Network

Parallel multi-thread processing in advanced intelligent processors is the core to realize high-speed and high-capacity signal processing systems.Optical neural network(ONN)has the native advantages of high parallelization,large bandwidth,and low power consumption to meet the demand of big data.Here,we demonstrate the dual-layer ONN with Mach-Zehnder interferometer(MZI)network and nonlinear layer,while the nonlinear activation function is achieved by optical-electronic signal conversion.Two frequency components from the microcomb source carrying digit datasets are simultaneously imposed and intelligently recognized through the ONN.We successfully achieve the digit classification of different frequency components by demultiplexing the output signal and testing power distribution.Efficient parallelization feasibility with wavelength division multiplexing is demonstrated in our high-dimensional ONN.This work provides a high-performance architecture for future parallel high-capacity optical analog computing.

An overview of multi-antenna technologies for space-ground integrated networks

Multi-antenna technologies have already achieved a series of great successes in the development of information networks. For future space-ground integrated networks(SGINs), the traditional various kinds of separated information networks will converge to a whole fully connected information network to provide more flexible and reliable services on a world scale. Regarding their great successes in existing systems, multiantenna technologies will be of critical importance for the realization of SGINs and multi-antenna technologies are definitely one of the most important enabling technologies for future converged SGINs. In this article, a comprehensive overview on multi-antenna technologies is given. We first investigate multi-antenna technologies from a theoretical viewpoint. It is shown that we can understand multi-antenna technologies in a general and unified point of view. This fact has two-fold meanings. First, the research on multi-antennas can help us understand the relationships between different technologies e.g., OFDMA, CDMA, etc. On the other hand,multi-antenna technologies are easy to integrate into various information systems. Following that, we discuss in depth the potentials and challenges of the multi-antenna technologies on different platforms and in different applications case by case. More specifically, we investigate spaceborne multi-antenna technologies, airborne multi-antenna technologies, shipborne multi-antenna technologies, etc. Moreover, the combinations of multiantenna technologies with other advanced wireless technologies e.g., physical layer network coding, cooperative communication, etc., are also elaborated.

Deep Learning Applied to Computational Mechanics:A Comprehensive Review,State of the Art,and the Classics

Three recent breakthroughs due to AI in arts and science serve as motivation:An award winning digital image,protein folding,fast matrix multiplication.Many recent developments in artificial neural networks,particularly deep learning(DL),applied and relevant to computational mechanics(solid,fluids,finite-element technology)are reviewed in detail.Both hybrid and pure machine learning(ML)methods are discussed.Hybrid methods combine traditional PDE discretizations with ML methods either(1)to help model complex nonlinear constitutive relations,(2)to nonlinearly reduce the model order for efficient simulation(turbulence),or(3)to accelerate the simulation by predicting certain components in the traditional integration methods.Here,methods(1)and(2)relied on Long-Short-Term Memory(LSTM)architecture,with method(3)relying on convolutional neural networks.Pure ML methods to solve(nonlinear)PDEs are represented by Physics-Informed Neural network(PINN)methods,which could be combined with attention mechanism to address discontinuous solutions.Both LSTM and attention architectures,together with modern and generalized classic optimizers to include stochasticity for DL networks,are extensively reviewed.Kernel machines,including Gaussian processes,are provided to sufficient depth for more advanced works such as shallow networks with infinite width.Not only addressing experts,readers are assumed familiar with computational mechanics,but not with DL,whose concepts and applications are built up from the basics,aiming at bringing first-time learners quickly to the forefront of research.History and limitations of AI are recounted and discussed,with particular attention at pointing out misstatements or misconceptions of the classics,even in well-known references.Positioning and pointing control of a large-deformable beam is given as an example.

星地融合边云协同网络下的资源调度研究

通过边缘计算和云计算的优势互补,面向6G的星地融合网络能够实现资源的弹性分配和协同利用,满足数据处理的实时性和智能化要求。为了提升资源服务能力和用户体验质量,提出了星地融合边云协同网络架构。首先,阐述了三种不同边云协同模式下的资源调度方案,从时延、能耗和多目标优化的角度分析了不同场景下的资源调度策略。然后,对比了现有资源调度求解模型和算法的优势和局限性。最后,对基于边云协同的星地融合网络中的资源调度研究进行了总结与展望。

面向智算融合网络的自主防御范式研究

随着数字经济时代算力供给模式的变革,以算力为核心的新型网络基础设施已成为实现算力资源共享、支撑数字经济转型的重要动力.在算力网络中,多元异构用户终端通过多种方式高频接入网络以随时随地获取算力服务,网络的开放性和动态性增大,算力网络将面临更严峻的安全挑战.然而,基于传统网络的安全防御模式通常针对具体安全问题静态式增补安全防护组件,无法主动适配用户需求灵活调整防御策略,难以应对算力网络中的安全风险.因此,本文面向新型算力网络安全需求,将安全功能作为网络内部属性,基于智算融合网络提出一种多维协同自主防御范式.结合智算融合网络“三层”“三域”的设计思想,在“三层”中,以广义服务层定义安全固有服务,以映射适配层智慧适配安全功能,以融合组件层执行安全策略;在“三域”中,以实体域先导资源适配,以知识域驱动安全服务流程,以感控域实施具体安全技术,构建“检测”“溯源”“防御”三维一体的完整基础管控流程,其中安全策略与技术可根据场景扩展性与业务安全性进行灵活调整.最终,通过仿真实验对所提范式有效性进行了验证,为未来智算融合安全的进一步研究和应用提供参考.

光子神经网络研究进展

在数据海量化、信息化的时代,电子计算机处理系统所面临的算力和能耗等性能要求愈发严苛,传统冯·诺依曼架构存在“内存墙”和“功耗墙”瓶颈,加之摩尔定律放缓甚至失效,使得电子芯片在计算速度和功耗方面遇到极大挑战,利用光计算替代传统电子计算将是解决当前算力与功耗问题的极具潜力的途径之一。本文系统地梳理了片上集成和自由空间的光子神经网络架构与算法方面的研究进展,详细介绍了典型的研究工作,然后讨论并对比了这两种光子神经网络的优劣势,以及光子神经网络的训练策略等。最后探讨了光子神经网络面临的挑战,并对其未来发展进行了前瞻性的展望。

基于深度确定性策略梯度的星地融合网络可拆分任务卸载算法

为解决低轨卫星网络中星地链路任务卸载时延长的问题,提出了一种基于深度确定性策略梯度(DDPG)的星地融合网络可拆分任务卸载算法。针对不同地区用户建立了星地融合网络的多接入边缘计算结构模型,通过应用多智能体DDPG算法,将系统总服务时延最小化的目标转化为智能体奖励收益最大化。在满足子任务卸载约束、服务时延约束等任务卸载约束条件下,优化用户任务拆分比例。仿真结果表明,所提算法在用户服务时延和受益用户数量等方面优于基线算法。

基于计算机视觉和DNN的运动姿态检测算法

针对传统人体运动姿态检测算法存在鲁棒性较差及准确率偏低的问题,文中基于改进的深度图卷积网络提出了一种运动姿态检测算法。该算法将图卷积网络的时域和空间域模型相结合,提升了模型的感受野,并从时、空两个维度提取人体特征点的数据,再利用残差网络将时域和空间域卷积相连接,进而改善了模型拟合能力较差的不足。同时,对于模型因卷积核固定而导致无法适应多种类数据的缺陷,使用多头注意力机制来增强其自适应能力。在实验测试中,所提出的模型改进项相较原算法性能有一定提升,且与最优算法相比,该模型的误差指标降低了1.14 mm,准确率则提升了1.3%,证明了所提方法的有效性及优越性。

面向微控制器的卷积神经网络加速器设计

针对目前嵌入式微控制器的性能难以满足实时图像识别任务的问题,提出一种适用于微控制器的卷积神经网络加速器。该加速器在卷积层设计了无阻塞的行并行乘法-加法树结构,获得了更高的硬件利用率;为了满足行并行的数据吞吐量,设计了卷积专用SRAM存储器。加速器将池化和激活单元融入数据通路,有效减少数据重复存取带来的时间开销。FPGA原型验证表明加速器的性能达到92.2 GOPS@100 MHz;基于TSMC 130 nm工艺节点进行逻辑综合,加速器的动态功耗为33 mW,面积为90 764.2μm^(2),能效比高达2 793 GOPS/W,比FPGA加速器方案提高了约100倍。该加速器低功耗、低成本的特性,有利于实现嵌入式系统在目标检测、人脸识别等机器视觉领域的广泛应用。

面向6G的超可靠低延迟通信关键技术:研究进展与展望

URLLC作为5G和6G蜂窝网络的主要通信服务之一,对于支持各种新兴的关键任务应用至关重要。首先,介绍了6G的背景和URLLC的重要性,随后回顾了URLLC的基础知识、6G URLLC的应用需求及其面临的主要挑战。接着,重点讨论了6G URLLC的几项关键技术及其研究机遇,包括实现全覆盖的空天地一体化网络、高频段通信、智能计算、可重构智能表面,以及人工智能驱动的网络优化。最后,展望了未来通信感知计算一体化、元宇宙与数字孪生支持的无线网络演化及分布式框架的演进。这些技术的发展将显著提升6GURLLC的性能和可靠性,为未来新兴应用提供坚实的基础。

基于移动边缘计算的空天地一体化网络架构

下一代无线通信技术将通过支持智能交通、智能医疗、虚拟现实/增强等服务来提高用户的体验质量。这些新兴的服务通常都是计算密集型和时延敏感型的,必须满足严格的时延、能耗和可靠性要求,而基于云计算的服务难以满足这些要求。为了解决以上问题,提出移动边缘计算(Mobile Edge Computing,MEC)技术,解决将卸载请求回传到云计算中心所具有的高时延和高带宽消耗的问题。空天地一体化网络(Space-Air-Ground Integrated Network,SAGIN)作为6G的重要研究方向,可以弥补全球范围内的巨大覆盖缺口,受到广泛关注。通过将MEC技术、联邦学习技术和人工智能技术引入SAGIN,高效管理网络中海量、异构的资源,构建低时延、低能耗、高可靠性的SAGIN,以支持新兴的各种服务。

IP承载网络技术演进方向研究

互联网业务的快速发展对IP承载网络提出了确定性、高可靠性、大带宽、弹性连接等新要求,而运营商业务本身的发展对IP承载网络提出了降本增效、差异化承载、安全可信的需求。基于业务演进对网络的需求、IP新技术的发展,对IP承载网络演进方向开展讨论,旨在打造高质量、差异化承载底座,满足多样化业务需求,提升网络价值,降低网络运营成本,筑牢网络安全底座,保障网络安全。

云边端融合的算网架构在智慧园区中的应用研究

随着全社会数字化变革的持续推进,利用云计算、边缘计算、人工智能、大数据、物联网等技术为各类园区赋能,打造新型智慧园区已成为趋势。然而,各类园区普遍占地规模大、入驻单位众多、持续建设周期长,在信息化系统建设领域存在重复建设、“信息孤岛”、网络架构复杂、算力不足且分配不均、缺乏统一算力调度体系等问题。同时,多种高算力需求技术应用越发普遍,算力和网络融合调度不足或成为制约园区数字化转型效果的基石性问题。基于此,从园区算网现状、云边端融合的算网架构、算网融合赋能智慧园区的具体案例等维度研究云边端融合的算网架构在智慧园区中的应用。

Distributed continuation power flow method for integrated transmission and active distribution network

As the integration of distributed generations(DGs)transforms the traditional distribution network into the active distribution network,voltage stability assessments(VSA)of transmission grid and distribution grid are not suitable to be studied separately.This paper presents a distributed continuation power flow method for VSA of global transmission and distribution grid.Two different parameterization schemes are adopted to guarantee the coherence of load growth in transmission and distribution grids.In the correction step,the boundary bus voltage,load parameter and equivalent power are communicated between the transmission and distribution control centers to realize the distributed computation of load margin.The optimal multiplier technique is used to improve the convergence of the proposed method.The three-phase unbalanced characteristic of distribution networks and the reactive capability limits of DGs are considered.Simulation results on two integrated transmission and distribution test systems show that the proposed method is effective.

空天地一体化场景中的6G通感算融合与数字孪生技术

随着接入需求的增加以及网络规模的扩大,需要在6G网络将通信、感知和计算3类技术融为一体,共同促进整体系统性能的提升,并利用数字孪生技术增强系统的实时维护和建模能力。分析了5G的局限性,研究了通感算融合技术在空天地一体化网络中的应用,以及数字孪生技术在构建未来“数字孪生、智慧泛在”世界时发挥的功能。提出了一种基于通感算融合技术的数字孪生网络实现方案,将空天地一体化网络映射至数字世界,实现反作用于物理世界,并展望了6G在通感算深度融合、全域覆盖方面的需求与优势,以期达到“一念天地,万物随心”的愿景。

通感算融合赋能的低轨卫星星座网络架构与关键技术

为了进一步支撑空天地一体化多维协同通信、精准遥感和高效泛在计算的协同演进,低轨卫星星座网络亟需发展通信、感知、计算融合(ISCC)理论与关键技术。基于此,首先概述了低轨卫星星座网络的研究背景和通感算融合网络研究现状;进一步地,聚焦于通信、感知和计算的深度融合,深入分析了通感算融合赋能的低轨卫星星座网络架构与关键技术;最后探讨了通感算赋能的低轨卫星星座网络面临的未来挑战与发展前景。本研究有望为未来实现空天地一体化的多维协同通信、精准遥感和高效泛在计算奠定基础。

通信-感知-计算-存储深度融合下的无线资源管控

为满足多样化应用与极致性能需求,6G网络将提供通信、感知、计算与存储的融合服务,实现无缝泛在连接、全域精准感知的愿景。围绕通感算存深度融合下无线资源管控研究,首先简要介绍了通感算存融合网络的产生背景和无线资源管控研究现状,并对通感算存深度融合的灵活组网架构构建与关键技术进行了阐述。进一步地,为实现通感算存深度融合下的多域资源高效管控,对移动边缘计算网络和通感一体化网络的无线资源管控关键技术进行了分析讨论,并在此基础上探讨了通信、感知、计算与存储融合下的无线资源管控相关技术难点。

面向存算架构的神经网络数字系统设计

随着深度学习与神经网络的不断发展,庞大的计算量使得传统的冯·诺依曼架构设备面临“存储墙”等问题,因此“存内计算(Compute-In-Memory,CIM)”成为满足神经网络高时效需求和高运算量要求的主流设计方向。针对高密度数据的高性能计算提供高速且节能的解决方案,设计了一款神经网络加速器。首先,完成了对ResNet14神经网络的量化,依据其结构设计了一种面向存内计算的数字系统。而后,为了增强该系统的多网络适配性,提出了一种兼容性架构构想,使该数字系统可适配ResNet18或其他卷积神经网络的部分卷积层。最后,将该系统加载到FPGA上进行验证。在10 MHz的时钟频率下,以Cifar-10和MNIST数据集进行目标分类任务,分别得到60 FPS下84.17%和98.79%的准确率,具有更小的数据位宽和相近的准确率。