检索结果-内蒙古大学图书馆

Tsinghua Science and Technology 2009年第5期14卷 646-654页

作者：殷雅俊吴继业范钦珊黄克智 Department of Engineering Mechanics School of AerospaceTsinghua University Division of Mechanics Nanjing University of Technology

This paper analyzes the geometric quantities that remain unchanged during parallel mapping （i.e., mapping from a reference curved surface to a parallel surface with identical normal direction）. The second gradient operator, the second class of integral theorems, the Gauss-curvature-based integral theorems, and the core property of parallel mapping are used to derive a series of parallel mapping invariants or geometrically conserved quantities. These include not only local mapping invariants but also global mapping invafiants found to exist both in a curved surface and along curves on the curved surface. The parallel mapping invariants are used to identify important transformations between the reference surface and parallel surfaces. These mapping invariants and transformations have potential applications in geometry, physics, biomechanics, and mechanics in which various dynamic processes occur along or between parallel surfaces.

关键词： second gradient operator second class of integral theorem parallel mapping invariants transformations

来源：评论

学校读者我要写书评

暂无评论

parallelized SLAM: Enhancing mapping and Localization Through Concurrent Processing

引用

SENSORS 2025年第2期25卷 365页

作者： Romero-Ramirez, Francisco J. Cazorla, Miguel Marin-Jimenez, Manuel J. Medina-Carnicer, Rafael Munoz-Salinas, Rafael Univ Rey Juan Carlos Dept Teoria Senal & Comunicac & Sistemas Telematic Campus Fuenlabrada Fuenlabrada 28942 Spain Univ Alicante Dept Ciencia Comp & Inteligencia Artificial Carretera San Vicente Raspeig S-N San Vicente Del Raspeig 03690 Spain Univ Cordoba Dept Informat & Anal Numer Campus Rabanales Edificio Einstein Cordoba 14071 Spain Inst Maimonides Invest Biomed IMIBIC Ave Menendez Pidal S-N Cordoba 14004 Spain

Simultaneous Localization and mapping (SLAM) systems face high computational demands, hindering their real-time implementation on low-end computers. An approach to addressing this challenge involves offline processing, i.e., a map of the environment map is created offline on a powerful computer and then passed to a low-end computer, which uses it for navigation, which involves fewer resources. However, even creating the map on a powerful computer is slow since SLAM is designed as a sequential process. This work proposes a parallel mapping method pSLAM for speeding up the offline creation of maps. In pSLAM, a video sequence is partitioned into multiple subsequences, with each processed independently, creating individual submaps. These submaps are subsequently merged to create a unified global map of the environment. Our experiments across a diverse range of scenarios demonstrate an increase in the processing speed of up to 6 times compared to that of the sequential approach while maintaining the same level of robustness. Furthermore, we conducted comparative analyses against state-of-the-art SLAM methods, namely UcoSLAM, OpenVSLAM, and ORB-SLAM3, with our method outperforming these across all of the scenarios evaluated.

关键词： SLAM lifelong mapping localization offline processing parallel mapping

来源：评论

学校读者我要写书评

暂无评论

Frame time-hopping patterns in multirate optical CDMA networks using conventional and multicode schemes

引用

IEEE TRANSACTIONS ON COMMUNICATIONS 2005年第5期53卷 863-875页

作者： Forouzan, AR Nasiri-Kenari, M Rezaee, N Univ Canterbury Dept Elect & Comp Engn Christchurch 8005 New Zealand Univ Tehran Dept Elect & Comp Engn Tehran 14399 Iran Sharif Univ Technol Dept Elect Engn Tehran 113659363 Iran

A class of generalized optical orthogonal codes (OOCs), namely, frame time-hopping (FTH) patterns with an extremely large cardinality, are studied for implementing multirate and multiservice (MR/MS) optical CDMA (OCDMA) networks. Conventional MR/MS methods, namely variable spreading rate and parallel mapping,are considered. Using FTH patterns, the problem of low OOC code cardinality in conventional MR/MS schemes is removed. Moreover, several new multicode methods, using subcode concatenation scheme, are proposed for MR/MS OCDMA. The proposed multicode schemes present flexibility for supporting MR/MS applications such as lower implementation complexity and cost, less need for service synchronization, and finally lower link power budget. Multiple-access performances of the systems are evaluated using saddle-point approximation methods considering photodetector shot-noise, dark current, and circuit thermal noise. The results show that the conventional parallel mapping outperforms the other schemes in high received powers, and the proposed multicode method, using Walsh subcode along with difference modulation, presents the best performance in low received powers for the cases considered.

关键词： frame time-hopping (FTH) multicode multirate (MR) multiservice (MS) optical CDMA optical orthogonal codes (OOCs) parallel mapping subcode concatenation variable spreading rate Walsh code

来源：评论

学校读者我要写书评

暂无评论

parallel implementation of self-organizing map on the partial tree shape neurocomputer

引用

NEURAL PROCESSING LETTERS 2000年第2期12卷 171-182页

作者： Kolinummi, P Pulkkinen, P Hämäläinen, T Saarinen, J Tampere Univ Technol Signal Proc Lab FIN-33720 Tampere Finland

A parallel mapping of self-organizing map (SOM) algorithm is presented for a partial tree shape neurocomputer (PARNEU). PARNEU is a general purpose parallel neurocomputer that is designed for soft computing applications. Practical scalability and a reconfigurable partial tree network are the main architectural features. The presented neuron parallel mapping of SOM with on-line learning illustrates a parallel winner neuron search and a coordinate transfer that are performed in the partial tree network. Phase times are measured to analyse speedup and scalability of the mapping. The performance of the learning phase in SOM with a four processor PARNEU configuration is about 26 MCUPS and the recall phase performs 30 MCPS. Compared to other mappings done for general purpose neurocomputers, PARNEU's performance is very good.

关键词： computer architecture neurocomputer parallel hardware implementation parallel mapping self-organizing map

来源：评论

学校读者我要写书评

暂无评论

Scalable realization of Sparse Distributed Memory on a multiprocessor system 9

Scalable realization of Sparse Distributed Memory on a multi...

引用

9th International Conference on Intelligent Systems

作者： Kolinummi, P Tikkanen, KN Hämäläinen, T Saarinen, J Tampere Univ Technol Digital & Comp Syst Lab FIN-33101 Tampere Finland

ISBN: (纸本)1880843331

parallel mappings of Sparse Distributed Memory (SDM) are presented for a partial tree shape neurocomputer. The implemented partial tree shape neurocomputer called PARNEU is a scalable co-processor system to speed up soft computing applications. The used SDM algorithm benefits from the improved reading method and mask based activation. Two different mapping strategies are presented and analysed in detail The effect of SDM size is analysed and the system scalability is verified with speedup figures of different SDM networks. Almost linear speedup is achieved with reasonable size SDM algorithm.

关键词： Sparse Distributed Memory neurocomputer parallel mapping

来源：评论

学校读者我要写书评

暂无评论

A Fast Inter-view Mode Selection Algorithm Based on Video Array Processor

A Fast Inter-view Mode Selection Algorithm Based on Video Ar...

引用

Annual Summit and Conference of the Asia-Pacific-Signal-and-Information-Processing-Association (APSIPA ASC)

作者： Wang, Yu Li, Xueting Zhu, Yun He, Feilong Xian Univ Posts & Telecommun Xian Peoples R China

ISBN: (纸本)9781728132488

In the encoding depth map of 3D-HEVC, a fast block search algorithm is proposed to solve the problem of fixed block pattern of inter-view prediction algorithm. Variable block sizes mode is used to select prediction mode by comparing depth values, and then the algorithm is parallelly mapped to a video array processor. Compared with HTM16.2 software implementation, the average encoding time of the proposed algorithm is reduced by more than 100 times. In addition, the hardware resource usage is reduce by 27.21% with similar processing speeds.

关键词： depth map inter-view prediction fast search algorithm variable block parallel mapping

来源：评论

学校读者我要写书评

暂无评论

An Implementation Method of HEVC Dataflow Graph Based on Reconfigurable Processer

An Implementation Method of HEVC Dataflow Graph Based on Rec...

引用

Asia-Pacific-Signal-and-Information-Processing-Association Annual Summit and Conference (APSIPA ASC)

作者： Zhu, Yun Hu, Chuanzhan Jiang, Lin Shen, Xubang Xian Univ Posts & Telecommun Xian Peoples R China Xian Univ Sci & Technol Xian Peoples R China Xian Microelect Technol Res Inst Xian Peoples R China

ISBN: (纸本)9789881476890

Aiming at the problems of low resource utilization and long algorithm execution time caused by unreasonable task division when complex applications in high-efficiency video coding are mapped on a reconfigurable processor. This paper proposes a method for implementing a dataflow graph (DFG) of HEVC algorithm based on a reconfigurable processor. According to analyzed the data dependence and parallelism between the algorithms, the loop optimization is performed on the HEVC part with more loops and the algorithm DFG is constructed. This paper mainly analyzes the mapping of smaller-scale DFG to different processing units in the array processor. The experimental results show that compared with the serial scheme, the speedup ratio can reach up to 7.4x, the processing element utilization can be increased by up to 80%, and the execution time can be reduced by up to 86%.

关键词： HEVC Loop optimization DFG parallel mapping

来源：评论

学校读者我要写书评

暂无评论

Implementation of Packet Classification Algorithm in NetworkProcessor based Router to enhance Multimedia Applications

Implementation of Packet Classification Algorithm in Network...

引用

2nd International Conference on Internet Multimedia Services Architecture and Application

作者： Avudaiammal, R. SivaSubramanian, R. Pandian, R. Seethalakshmi, P. Anna Univ Madras 600025 Tamil Nadu India Vinayaka Missions Univ Res Scholar Salem Tamil Nadu India

ISBN: (纸本)9781424426843

The objective of this paper is to implement the Tuple Pruning Search(TPS) based packet classification algorithm in two different mappings namely parallel mapping and Pipelined mapping for Network Processor(NP) based router to support multimedia applications. The performance is evaluated using Intel IXP2400 NP Simulator. The results demonstrate that, TPS has better classification rate and lesser memory occupancy than Tuple Space Search algorithm. parallel mapping of TPS based algorithm achieves better performance than the Pipelined mapping and is well suited to achieve high speed packet classification to support multimedia applications.

关键词： Multimedia QoS Packet Classification TPS Network Processor IXP 2400 parallel mapping Pipelined mapping

来源：评论

学校读者我要写书评

暂无评论

A High-Performance Area-Efficient AES Cipher on a Many-Core Platform

A High-Performance Area-Efficient AES Cipher on a Many-Core ...

引用

45th Asilomar Conference on Signals, Systems and Computers (ASILOMAR)

作者： Liu, Bin Baas, Bevan M. Univ Calif Davis Dept Elect & Comp Engn Davis CA 95616 USA

ISBN: (纸本)9781467303231

This paper presents the design and software implementation of a high-performance area-efficient Advanced Encryption Standard (AES) cipher on a many-core platform. A preliminary cipher design is partitioned and mapped to an array of 70 small processors, and offers a throughput of 16.625 clock cycles per byte. The usage of instruction and data memory, and the workload of each processor are characterized for further optimization. Through workload balancing and processor fusion, the throughput of the cipher is increased by 43% to 9.5 clock cycles per byte, while the number of processors utilized is reduced to 59, which is only 10.03 mm2 in a 65 nm fine-grained many-core system. In comparison with published AES implementations on general purpose processors, our design has 3.6-10.7 times higher throughput per area. Moreover, the presented design shows 1.5 times higher throughput than the TI DSP C6201 and 3.4 times higher throughput per area than the GeForce 8800 GTX.

关键词： Advanced Encryption Standard (AES) parallel mapping many-core processor Asynchronous Array of Simple Processors (AsAP)

来源：评论

学校读者我要写书评

暂无评论

Adaptive parallelism mapping in dynamic environments using machine learning

Adaptive parallelism mapping in dynamic environments using m...

引用

作者： Emani, Murali Krishna University of Edinburgh

学位级别：doctor

Modern day hardware platforms are parallel and diverse, ranging from mobiles to data centers. Mainstream parallel applications execute in the same system competing for resources. This resource contention may lead to a drastic degradation in a program’s performance. In addition, the execution environment composed of workloads and hardware resources, is dynamic and unpredictable. Efficient matching of program parallelism to machine parallelism under uncertainty is hard. The mapping policies that determine the optimal allocation of work to threads should anticipate these variations. This thesis proposes solutions to the mapping of parallel programs in dynamic environments. It employs predictive modelling techniques to determine the best degree of parallelism. Firstly, this thesis proposes a machine learning-based model to determine the optimal thread number for a target program co-executing with varying workloads. For this purpose, this offline trained model uses static code features and dynamic runtime information as input. Next, this thesis proposes a novel solution to monitor the proposed offline model and adjust its decisions in response to the environment changes. It develops a second predictive model for determining how the future environment should be, if the current thread prediction was optimal. Depending on how close this prediction was to the actual environment, the predicted thread numbers are adjusted. Furthermore, considering the multitude of potential execution scenarios where no single policy is best suited in all cases, this work proposes an approach based on the idea of mixture of experts. It considers a number of offline experts or mapping policies, each specialized for a given scenario, and learns online the best expert that is optimal for the current execution. When evaluated on highly dynamic executions, these solutions are proven to surpass default, state-of-art adaptive and analytic approaches

关键词： parallel mapping machine learning

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：