检索结果-内蒙古大学图书馆

Low-Complexity CSI Feedback for FDD Massive MIMO Systems via learning to optimize

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS 2025年第4期24卷 3483-3498页

作者： Ma, Yifan He, Hengtao Song, Shenghui Zhang, Jun Letaief, Khaled B. Hong Kong Univ Sci & Technol Dept Elect & Comp Engn Hong Kong Peoples R China Zhejiang Wanli Univ Coll Informat & Intelligence Engn Ningbo 315100 Peoples R China

In frequency-division duplex (FDD) massive multiple-input multiple-output (MIMO) systems, the growing number of base station antennas leads to prohibitive feedback overhead for downlink channel state information (CSI). To address this challenge, state-of-the-art (SOTA) fully data-driven deep learning (DL)-based CSI feedback schemes have been proposed. However, the high computational complexity and memory requirements of these methods hinder their practical deployment on resource-constrained devices like mobile phones. To solve the problem, we propose a model-driven DL-based CSI feedback approach by integrating the wisdom of compressive sensing and learning to optimize (L2O). Specifically, only a linear learnable projection is adopted at the encoder side to compress the CSI matrix, thereby significantly cutting down the user-side complexity and memory expenditure. On the other hand, the decoder incorporates two specially designed components, i.e., a learnable sparse transformation and an element-wise L2O reconstruction module. The former is developed to learn a sparse basis for CSI within the angular domain, which explores channel sparsity effectively. The latter shares the same long short term memory (LSTM) network across all elements of the optimization variable, eliminating the retraining cost when problem scale changes. Simulation results show that the proposed method achieves a comparable performance with the SOTA CSI feedback scheme but with much-reduced complexity, and enables multiple-rate feedback.

关键词： 6G CSI feedback learning to optimize learning to optimize massive MIMO massive MIMO model-driven deep learning model-driven deep learning model-driven deep learning

来源：评论

学校读者我要写书评

暂无评论

learning to optimize QoS-Constrained Beamforming in Multi-User Systems: A Penalty-Dual Framework

引用

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS 2024年第11期23卷 16123-16138页

作者： Li, Yang Liu, Ya-Feng Xu, Fan Shi, Qingjiang Chang, Tsung-Hui Shenzhen Res Inst Big Data Shenzhen 518172 Peoples R China Chinese Acad Sci Inst Computat Math & Sci Engn Comp Acad Math & Syst Sci State Key Lab Sci & Engn Comp Beijing 100190 Peoples R China Peng Cheng Lab Shenzhen 518055 Peoples R China Tongji Univ Sch Software Engn Shanghai 200092 Peoples R China Chinese Univ Hong Kong Sch Sci & Engn Shenzhen 518172 Peoples R China

This paper investigates a novel deep learning framework for the general nonconvex quality-of-service (QoS)-constrained beamforming design problems in multi-user systems. While existing deep learning-based approaches have shown great success for various power allocation and beamforming design problems, most of the considered problems are equipped with simple constraints (e.g., power budget constraints), which can be satisfied by a simple projection operation. However, it is still a challenge to tackle the more complicated QoS constraints, in which the beamformers and the wireless channels are commonly coupled. To fill this gap, this paper proposes an augmented Lagrangian based penalty-dual training algorithm, which trains two individual neural networks for inferring the beamformers and the corresponding Lagrange multipliers alternatingly. Furthermore, we apply the proposed penalty-dual learning framework to optimize the energy-efficient unicast beamformers and the power-minimized multicast beamformers, respectively. The neural network architectures are judiciously designed based on the solution structures of the two problems. Simulation results on the two applications demonstrate that the proposed penalty-dual approach outperforms state-of-the-art learning approaches and optimization-based algorithms in terms of the constraint violation and the computational time, respectively.

关键词： Quality of service Array signal processing Neural networks Wireless communication Training Resource management Unicast Beamforming design learning to optimize nonconvex optimization penalty-dual quality-of-service constraints

来源：评论

学校读者我要写书评

暂无评论

learning to optimize: A Primer and A Benchmark

引用

JOURNAL OF MACHINE learning RESEARCH 2022年第1期23卷 1-59页

作者： Chen, Tianlong Chen, Xiaohan Chen, Wuyang Wang, Zhangyang Heaton, Howard Liu, Jialin Yin, Wotao Engn Univ Texas Austin Dept Elect & Comp Austin TX 78712 USA Typal LLC Typal Res Los Angeles CA 90064 USA Damo Acad Decis Intelligence Lab Alibaba US Bellevue WA 98004 USA

learning to optimize (L2O) is an emerging approach that leverages machine learning to develop optimization methods, aiming at reducing the laborious iterations of hand engi-neering. It automates the design of an optimization method based on its performance on a set of training problems. This data-driven procedure generates methods that can efficiently solve problems similar to those in training. In sharp contrast, the typical and traditional designs of optimization methods are theory-driven, so they obtain performance guarantees over the classes of problems specified by the theory. The difference makes L2O suitable for repeatedly solving a particular optimization problem over a specific distribution of data, while it typically fails on out-of-distribution problems. The practicality of L2O depends on the type of target optimization, the chosen architecture of the method to learn, and the training procedure. This new paradigm has motivated a community of researchers to explore L2O and report their *** article is poised to be the first comprehensive survey and benchmark of L2O for continuous optimization. We set up taxonomies, categorize existing works and research directions, present insights, and identify open challenges. We benchmarked many existing L2O approaches on a few representative optimization problems. For reproducible research and fair benchmarking purposes, we released our software implementation and data in the package Open-L2O at https://***/VITA-Group/Open-L2O.

关键词： learning to optimize meta learning optimization algorithm unrolling

来源：评论

学校读者我要写书评

暂无评论

learning to optimize:A tutorial for continuous and mixed-integer optimization

引用

Science China Mathematics 2024年第6期67卷 1191-1262页

作者： Xiaohan Chen Jialin Liu Wotao Yin Decision Intelligence Lab Alibaba DAMO AcademyBellevueWA98004USA

learning to optimize(L2O)stands at the intersection of traditional optimization and machine learning,utilizing the capabilities of machine learning to enhance conventional optimization *** real-world optimization problems frequently share common structures,L2O provides a tool to exploit these structures for better or faster *** tutorial dives deep into L2O techniques,introducing how to accelerate optimization algorithms,promptly estimate the solutions,or even reshape the optimization problem itself,making it more adaptive to real-world *** considering the prerequisites for successful applications of L2O and the structure of the optimization problems at hand,this tutorial provides a comprehensive guide for practitioners and researchers alike.

关键词： AI for mathematics(AI4Math) learning to optimize algorithm unrolling plug-and-play methods differentiable programming machine learning for combinatorial optimization(ML4CO)

来源：评论

学校读者我要写书评

暂无评论

learning to optimize Distributed Optimization: ADMM-based DC-OPF Case Study

Learning to Optimize Distributed Optimization: ADMM-based DC...

引用

IEEE-Power-and-Energy-Society General Meeting (PESGM)

作者： Li, Meiyi Kolouri, Soheil Moharnmadil, Javad Univ Texas Austin Dept Civil Architectural & Environm Engn Austin TX 78712 USA Vanderbilt Univ Dept Comp Sci Nashville TN USA

ISBN: (纸本)9781665464413

The decision-making paradigms of future energy systems are increasingly becoming decentralized and multi-entity/agent. The Alternating Direction Method of Multipliers (ADMM) has been widely used to address the computational needs of decentralized decision-making problems, e.g., optimal power flow (OPE). In this paper, we propose a novel data-driven method to accelerate the convergence of ADMM for decentralized DC-OPF, where our optimizer will learn the iterative behavior of agents to produce a high-quality feasible solution. The proposed method utilizes the gauge maps to enforce feasibility with respect to agents' local constraints while iteratively penalizing violations of the shared constraints. We used the IEEE 57-bus system to showcase the performance of the proposed method. Our experimental results demonstrate significant run-time reduction for using ADMM to solve the DC-OPF problem.

关键词： Alternating Direction Method of Multipliers (ADMM) DC optimal power flow neural network machine learning learning to optimize

来源：评论

学校读者我要写书评

暂无评论

L2O-ILT: learning to optimize Inverse Lithography Techniques

引用

IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS 2024年第3期43卷 944-955页

作者： Zhu, Binwu Zheng, Su Yu, Ziyang Chen, Guojin Ma, Yuzhe Yang, Fan Yu, Bei Wong, Martin D. F. Chinese Univ Hong Kong Dept Comp Sci & Engn Hong Kong Peoples R China Hong Kong Univ Sci & Technol Guangzhou Microelect Thrust Guangzhou 511453 Peoples R China Fudan Univ Microelect Dept State Key Lab AS & Syst Shanghai 200437 Peoples R China

Inverse lithography technique (ILT) is one of the most widely used resolution enhancement techniques (RETs) to compensate for the diffraction effect in the lithography process. However, ILT suffers from runtime overhead issues with the shrinking size of technology nodes. In this article, our proposedL2O-ILT framework unrolls the iterative ILT optimization algo-rithm into a learnable neural network with high interpretability, which can generate a high-quality initial mask for fast refinement. Experimental results demonstrate that our method achieves better performance on both mask printability and runtime than the previous methods.

关键词： Lithography Optimization Optical imaging Optical diffraction Computational modeling Kernel Adaptive optics Design for manufacture mask optimization learning to optimize

来源：评论

学校读者我要写书评

暂无评论

Towards Robust learning to optimize with Theoretical Guarantees

Towards Robust Learning to Optimize with Theoretical Guarant...

引用

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Song, Qingyu Lin, Wei Wang, Juncheng Xu, Hong CUHK Hong Kong Peoples R China HKBU Hong Kong Peoples R China

ISBN: (纸本)9798350353006

learning to optimize (L2O) is an emerging technique to solve mathematical optimization problems with learning-based methods. Although with great success in many real-world scenarios such as wireless communications, computer networks, and electronic design, existing L2O works lack theoretical demonstration of their performance and robustness in out-of-distribution (OOD) scenarios. We address this gap by providing comprehensive proofs. First, we prove a sufficient condition for a robust L2O model with homogeneous convergence rates over all In-Distribution (InD) instances. We assume an L2O model achieves robustness for an InD scenario. Based on our proposed methodology of aligning OOD problems to InD problems, we also demonstrate that the L2O model's convergence rate in OOD scenarios will deteriorate by an equation of the L2O model's input features. Moreover, we propose an L2O model with a concise gradient-only feature construction and a novel gradient-based history modeling method. Numerical simulation demonstrates that our proposed model outperforms the state-of-the-art baseline in both InD and OOD scenarios and achieves up to 10 convergence speedup. The code of our method can be found from https://***/NetX-lab/GoMathL2O-Official.

关键词： learning to optimize Out-of-distribution generalization

来源：评论

学校读者我要写书评

暂无评论

learning to optimize: a primer and a benchmark

The Journal of Machine Learning Research

引用

The Journal of Machine learning Research 2022年第1期23卷 8562-8620页

作者： Tianlong Chen Xiaohan Chen Wuyang Chen Zhangyang Wang Howard Heaton Jialin Liu Wotao Yin Department of Electrical and Computer and Engineering The University of Texas at Austin Austin TX Typal Research Typal LLC Los Angeles CA Alibaba US Damo Academy Decision Intelligence Lab Bellevue WA

learning to optimize (L2O) is an emerging approach that leverages machine learning to develop optimization methods, aiming at reducing the laborious iterations of hand engineering. It automates the design of an optimization method based on its performance on a set of training problems. This data-driven procedure generates methods that can efficiently solve problems similar to those in training. In sharp contrast, the typical and traditional designs of optimization methods are theory-driven, so they obtain performance guarantees over the classes of problems speci_ed by the theory. The difference makes L2O suitable for repeatedly solving a particular optimization problem over a specific distribution of data, while it typically fails on out-of-distribution problems. The practicality of L2O depends on the type of target optimization, the chosen architecture of the method to learn, and the training procedure. This new paradigm has motivated a community of researchers to explore L2O and report their *** article is poised to be the first comprehensive survey and benchmark of L2O for continuous optimization. We set up taxonomies, categorize existing works and research directions, present insights, and identify open challenges. We benchmarked many existing L2O approaches on a few representative optimization problems. For reproducible research and fair benchmarking purposes, we released our software implementation and data in the package Open-L2O.

关键词： learning to optimize meta learning optimization algorithm unrolling

来源：评论

学校读者我要写书评

暂无评论

EDformer family: End-to-end multi-task load forecasting frameworks for day-ahead economic dispatch☆

引用

APPLIED ENERGY 2025年 383卷

作者： Tian, Zhirui Liu, Weican Zhang, Jiahao Sun, Wenpu Wu, Chenye Chinese Univ Hong Kong Sch Sci & Engn Shenzhen Guangdong Peoples R China

The highly penetrated renewable energy resources have significantly increased the uncertainty faced by the power system. Accurate day-ahead economic dispatch (ED) is crucial for managing this uncertainty and ensuring the efficient use of energy resources. To guarantee the planned dispatch schedules align closely with actual operations, most existing methods focus on improving the accuracy of time-series load forecasting. However, minimizing the commonly adopted forecasting accuracy metrics, e.g., the mean squared error (MSE), cannot directly correspond to the fidelity of day-ahead ED. This misalignment for load forecasting limits the practical effectiveness of the existing methods. To this end, based on the multi-task learning technique and end-to-end learning framework, we propose the EDformer family which contains two learning frameworks, EDformer and EEDformer. Specifically, the proposed frameworks parallel multiple load forecasting tasks based on customized encoder-decoder structures and address the inconsistency between the forecasting accuracy metrics and actual ED costs through a customized loss function. Furthermore, the proposed frameworks adopt the learning to optimize (L2O) technique, utilizing fully connected neural networks based on data transposition strategies to learn the mapping relationships, replacing conventional solvers and thereby accelerating the training process. EDformer, through independent parameter training, achieves high forecasting accuracy while requiring high computational resources. EEDformer, on the other hand, adopts parameter- sharing techniques, slightly sacrificing forecasting accuracy to significantly improve computational efficiency, making the proposed EDformer family adaptable to different practical requirements. Validation using real data from Australia on the IEEE 9-Bus and 39-Bus systems proves the superiority of the EDformer family in enhancing day-ahead dispatch effectiveness.

关键词： Day-ahead economic dispatch Load forecast End-to-end learning Multi-task learning learning to optimize

来源：评论

学校读者我要写书评

暂无评论

Multi-objective optimization-assisted single-objective differential evolution by reinforcement learning

引用

SWARM AND EVOLUTIONARY COMPUTATION 2025年 94卷

作者： Zhang, Haotian Guan, Xiaohong Wang, Yixin Nan, Nan Xi An Jiao Tong Univ Frontier Inst Sci & Technol Ctr Art & Sci & Presentat & Commun Xian 710049 Peoples R China Xi An Jiao Tong Univ Fac Elect & Informat Engn MOE KLINNS Lab Xian 710049 Peoples R China Tsinghua Univ Ctr Intelligent & Networked Syst Dept Automat Beijing 100084 Peoples R China

"learning to optimize"design systems for evolutionary algorithm (EA) automatic design have become a trend, especially for differential evolution (DE). "learning to optimize"design systems for EAs have two main parts: an excellent "backbone"algorithm with learnable components, and a learning scheme to determine the components of the "backbone"algorithm. A good "backbone"algorithm is of great importance for the algorithm design, because it determines the algorithm design space and potential. The learning scheme determines whether we can realize the potential or not. Existing studies generally choose one developed EA as the "backbone"algorithm, which constrains the potential of the design system because the "backbone"algorithm is relatively simple. To solve the problem and design a good EA, in this paper, we first propose a three-stage hybrid DE framework for single objective optimization, called SMS-DE, which implements single- objective DE, multi-objective DE, and single-objective DE sequentially. The multi-objective DE aims to enhance exploration ability. Second, we apply the framework to two advanced DEs, JADE and LSHADE, which results in two new algorithms: SMS-JADE and SMS-LSHADE. Third, the newly proposed algorithm, SMS-LSHADE, is considered the "backbone"algorithm, and the reinforcement learning method (Q-learning) is used to control the parameter for allocating computational resources to each stage, which results in another algorithm called QSMS-LSHADE. Experimental results on the CEC 2018 test suite show that SMS-DE, SMS-JADE, and SMSLSHADE can perform significantly better than their counterparts and that SMS-QLSHADE performs the best among many developed DEs.

关键词： Evolutionary algorithm Multi-objective optimization Single-objective optimization learning to optimize

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：