共查询到20条相似文献,搜索用时 171 毫秒
1.
2.
3.
可重构计算是未来高性能计算的发展趋势,它兼具了通用计算的灵活性和专用计算的高效性,充分利用系统资源的同时,又能发挥应用程序的效率。可重构编译是推广可重构计算的关键技术,可重构编译系统能够为传统的软件编程人员提供一个体系结构透明的开发平台,并让用户真正灵活利用可重构计算平台。 相似文献
4.
5.
在分析传统FPGA动态重构方法性能缺陷的基础上,创新性的提出了基于改进型游程编码的FPGA动态重构方法,并详细介绍了该方法的设计实现。与传统FPGA动态重构方法对比测试结果表明,基于改进型游程编码的FPGA动态重构方法不仅可以显著提高FPGA动态重构的速度,而且可以降低对程序存储器容量要求。目前,该技术已在重大工程项目中得到应用。 相似文献
6.
基于可重构核的FPGA电路设计 总被引:4,自引:0,他引:4
电路系统的自适应性、紧凑性和低成本 ,促进了在嵌入式系统中软硬件的协同设计。在线可重构FPGA不仅可以满足这一要求 ,而且在可编程专用电路系统设计的验证及可靠性等方面有着良好的应用 ,文中介绍了可重构 FPGA的实现结构及评估方法 ,提出以线性矢量表征可重构 FPGA及其可重构核的研究模型 ,以及基于可重构核的模块化设计 ,认为面向分类的专用类可重构 FPGA应当是现阶段可重构 FPGA的研究主题。 相似文献
7.
5G主要有eMBB、mMTC和uRLLC共3种应用场景,不同的场景有着不同的性能需求,需要按需分配网络资源,因此5G需要进行云化重构.本文从运营商的角度,首先提出了网络基础设施重构方案以实现软硬件解耦和基础资源管理,提出了网络功能重构方案以实现5G各项应用功能,最后提出了运营重构的解决方案,列举了5G的运营方式和网络开通方式,并举例介绍了保障能力方案和人员转型方向,为5G正式商用提供了云化解决方案. 相似文献
8.
CCSDS标准的LDPC生成矩阵具有分块循环特征,并且各种码率的生成矩阵的校验部分都可以分解为8×x的形式,提出利用CCSDS标准的LDPC的特点设计动态可重构LDPC编码器。首先提出了LDPC编码器码速率重构的4种模型,然后分析了功能重构的关键技术模块,最后对码速率重构进行了仿真,并对仿真结果和综合结果进行了分析,结果与理论分析一致。 相似文献
9.
10.
模块化综合系统的重构设计 总被引:2,自引:1,他引:1
航空电子的模块化综合是提升系统综合性能的重要途径之一,重构是模块化综合系统的关键技术.对模块化综合系统的两重构形态(故障重构和功能重构)进行了详细阐述,基于模块化综合系统的体系结构,以面向对象的系统分析方法,提出了一种重构的顶层设计、故障检测和隔离、重构逻辑的设计方法,并对重构的关键要素进行了系统性探讨. 相似文献
11.
12.
In this paper, we study the reconfiguration capability in ATM networks and its ability to reduce ATM cell losses. Digital cross-connect systems have been used in traditional reconfigurable networks to concatenate channels to avoid store- and-forward delays. We propose an implementation of the reconfiguration capability in ATM networks that restricts the transmission of cells of an O-D pair to particular channels in each physical link in a simple path from origin to destination. With this proposed implementation, digital cross-connect systems are not needed. We formulate the problem of jointly determining logical networks and the routing assignments on the logical network of a reconfigurable ATM network as a nonlinear mixed integer programming problem. We prove that the joint problem with the integrality constraints relaxed can be simplified to a convex programming problem. This provides a method to compute a lower bound on the minimum cell losses for the original problem. An algorithm is developed to solve the nonlinear mixed integer programming problem. In the computational experiments, the proposed algorithm determined a good feasible solution in a few minutes of CPU time on SUN SPARCsystem 400. In the experiments, the reconfiguration capability decreased the total cell losses by up to 80%. Problem parameters that affect the effectiveness of the reconfiguration capability in reducing ATM cell losses are identified.Supported by NSF Grants NCR 90-16348 and NCR 92-23148. 相似文献
13.
Reducing the number of visits to failure-free nodes can effectively reduce the reconstruction time of logical columns and improve the reconstruction efficiency. In this paper, we describe a new method to speed up the reconfiguration for the VLSI arrays. An efficient algorithm was proposed based on shortest path first principle for accelerating reconfiguration of VLSI processor subarrays with high power efficiency to meet the requirement of the power consumption of embedded system. The proposed algorithm greatly reduces the number of visits to the fault-free PEs for constructing a local optimal logical column and effectively reduces the construction time. Experimental results show that the proposed algorithm is capable of reducing the consumption time by 32.15% and reducing the numbers of visited PEs by 49.61% for a 128 × 128 host array with 20% falut rate. 相似文献
14.
In order to improve the suppression capability of parametric perturbation and energy efficiency (EE) of heterogeneous networks (HetNets),a robust resource allocation algorithm was proposed to maximize system EE for reducing cross-tier interference power in non-orthogonal multiple access (NOMA) based HetNets.Firstly,the resource optimization problem was formulated as a mixed integer and nonlinear programming one under the constraints of the interference power of macrocell users,maximum transmit power of small cell base station (BS),resource block assignment and the quality of service (QoS) requirement of each small cell user.Then,based on ellipsoid bounded channel uncertainty models,the original problem was converted into the equivalent convex optimization problem by using the convex relaxation method,Dinkelbach method and the successive convex approximation (SCA) method.The analytical solutions were obtained by using the Lagrangian dual approach.Simulation results verifiy that the proposed algorithm had better EE and robustness by comparing it with the existing algorithm under perfect channel state information. 相似文献
15.
Weiguo Ju Shanguo Huang ZhenZhen Xu Jie Zhang Wanyi Gu 《International Journal of Communication Systems》2015,28(2):358-373
Flexgrid optical networking is an attractive solution for efficiently matching allocated bandwidth with link demand but suffers from inevitable spectrum fragmentation. Spectrum fragmentation impairs network performance and results in high blocking rate and low spectrum utilization efficiency. Therefore, an optimization mechanism handling spectrum fragmentation is of vital importance in flexgrid optical networks. In this paper, we propose a genetic algorithm for solving the spectrum fragmentation problem with the objective of compacting occupation of the spectrum in flexgrid optical networks. A string of lightpaths are coded as the chromosome. The spectrum fusion degree and fragment fusion degree are introduced as the fitness functions to conduct the evolution in genetic algorithm, which can also be used to assess the degree of spectrum fragmentation in the network. As a result, the genetic algorithm provides a lightpath reconfiguration map, which identifies the candidate lightpaths to be reallocated, their reconfiguration sequence, and new locations. The proposed algorithm is compared with commonly used approaches under different network conditions. Simulation results demonstrate the ability of the proposed algorithm to efficiently solve the problem of spectrum defragmentation in flexgrid optical networks. Copyright © 2013 John Wiley & Sons, Ltd. 相似文献
16.
针对动态可重构处理器的配置信息加载延时,提出了一种基于神经网络的可扩展的重构指令预取机制.增加感受器的历史指令信息,并结合感受器权重构建新型的感受器模型,通过权重与历史指令信息的协同训练学习重构指令调用规律.在处理器运行过程中,提前完成对后续重构指令的预测及配置信息的预取,隐藏指令重构成本.进一步提出了本方法的可扩展实现框架,神经网络的学习结果作为重构指令的关联信息,被移至内存并分布式存储.在重构指令预取时,完成对神经网络学习信息的加载.实验结果表明,该方法对重构指令的预测准确率达91%,综合性能平均提升40%. 相似文献
17.
均衡虚拟网构建算法研究 总被引:2,自引:1,他引:1
如何在物理资源有限的条件下构建出满足用户需求的虚拟网是一个热点问题。该文针对虚拟网(Virtual Network,VN)的构建方法进行了论述,分析了虚拟网构建的数学模型,给出了构建策略的若干重要原则,为提高构网成功率,分别提出基于链路负载均衡度和节点负载均衡度的虚拟网构建算法,在此基础上提出了考虑均衡的自适应虚拟网构建算法(Balanced Adaptive VN Construction Algorithm, BACA)以及链路失效后的处理方法和迁移策略。最后,通过仿真实验,从虚拟网构建需求接收率和构建后整个网络的链路均衡度以及节点均衡度3个方面验证了BACA的优越性。 相似文献
18.
多连接技术允许用户同时建立和保持与多个小区/接入点的连接,通过网络元素之间的协调在吞吐量和可靠性方面大幅提高网络性能。针对毫米波通信中超高频段的链路中断问题,研究了多连接基于链路配置的调度算法,以提高链路调度效率,降低复杂度。首先,在系统模型中采用链路配置作为优化变量;其次,设计了多连接比例公平的调度准则;最后,提出一种基于列生成算法的链路配置调度优化算法,利用Dantzig-Wolfe分解将原问题分解为限制主问题和定价问题,并结合分支定界方法获得最优解。仿真结果表明,所提算法能够在数值上逼近全局最优,并且比现有的毫米波蜂窝网络链路调度方案增益平均提高40%以上。 相似文献
19.
20.
为了提高现场可编程门阵列(FPGA)的资源利用率,在介绍FPGA重构技术的原理和分类的基础上,讨论了Virtex-4系列FPGA的配置原理和动态重构的方法,并设计出数字信号处理器(DSP)配置FPGA的硬件方案来实现可重构系统。FPGA采用SelectMAP配置方式,实现配置逻辑的快速重构和局部动态重构,最后根据Virtex-4的配置流程和时序关系,给出了可重构系统配置的软件流程。经实验测试,该系统稳定可靠,可在1 s内完成5 Mbyte配置程序的动态重构。 相似文献