期刊文献+
共找到2,410篇文章
< 1 2 121 >
每页显示 20 50 100
Network Coding理论的研究进展 被引量:1
1
作者 赵进 张福炎 《计算机应用与软件》 CSCD 北大核心 2008年第2期196-199,共4页
Network Coding是一种新的信息传输方式。与传统的路由相比,Network Coding可以提高网络的性能。从理论和应用两方面总结了Network Coding研究的进展,同时,分析了将Network Coding应用到实际网络中所面临的一些问题,最后提出了进一步的... Network Coding是一种新的信息传输方式。与传统的路由相比,Network Coding可以提高网络的性能。从理论和应用两方面总结了Network Coding研究的进展,同时,分析了将Network Coding应用到实际网络中所面临的一些问题,最后提出了进一步的研究方向。 展开更多
关键词 network coding 多播 网络流
在线阅读 下载PDF
Complex field network-coded cooperation based on multi-user detection in wireless networks 被引量:2
2
作者 Jing Wang Xiangyang Liu +1 位作者 Kaikai Chi Xiangmo Zhao 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2013年第2期215-221,共7页
Cooperative communication can achieve spatial diversity gains,and consequently combats signal fading due to multipath propagation in wireless networks powerfully.A novel complex field network-coded cooperation(CFNCC... Cooperative communication can achieve spatial diversity gains,and consequently combats signal fading due to multipath propagation in wireless networks powerfully.A novel complex field network-coded cooperation(CFNCC) scheme based on multi-user detection for the multiple unicast transmission is proposed.Theoretic analysis and simulation results demonstrate that,compared with the conventional cooperation(CC) scheme and network-coded cooperation(NCC) scheme,CFNCC would obtain higher network throughput and consumes less time slots.Moreover,a further investigation is made for the symbol error probability(SEP) performance of CFNCC scheme,and SEPs of CFNCC scheme are compared with those of NCC scheme in various scenarios for different signal to noise ratio(SNR) values. 展开更多
关键词 network coding complex field wireless network cooperative communication multi-user detection
在线阅读 下载PDF
D-stability and disturbance attenuation properties fornetworked control systems: switched system approach 被引量:2
3
作者 Qiuxia Chen Andong Liu 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2016年第5期1108-1114,共7页
Both D-stability and finite L2-gain properties are studiedfor a class of uncertain discrete-time systems with timevaryingnetwork-induced delays. By using coordinate transformand delay partition, the D-stability and H... Both D-stability and finite L2-gain properties are studiedfor a class of uncertain discrete-time systems with timevaryingnetwork-induced delays. By using coordinate transformand delay partition, the D-stability and H∞ performance problemsfor such networked control systems (NCSs) are equivalentlytransferred into the corresponding problems for switching systemswith arbitrary switching. Then, a sufficient condition for the existenceof the robust D-stabilizing controllers is derived in termsof linear matrix inequality (LMI), and the design method is alsopresented for the state feedback controllers which guarantee thatall the closed-loop poles remain inside the specified disk D(α,r)and the desired disturbance attenuation level. Finally, an illustrativeexample is given to demonstrate the effectiveness of the proposedresults. 展开更多
关键词 D-STABILITY networked control system (ncS) H∞performance switching system.
在线阅读 下载PDF
TCAR: A new network coding-aware routing mechanism based on local topology detection 被引量:2
4
作者 王伟平 吴伟 +1 位作者 关庆娟 王建新 《Journal of Central South University》 SCIE EI CAS 2014年第8期3178-3185,共8页
Recent researches show that inter-session network coding could decrease the number of packets transmission and achieve higher throughput in wireless network compared with traditional forwarding mechanism. In most exis... Recent researches show that inter-session network coding could decrease the number of packets transmission and achieve higher throughput in wireless network compared with traditional forwarding mechanism. In most existing relay mechanisms based on inter-session network such as COPE, relay node demands to collect the messages from its neighbor nodes to get notice of which packets already overheard by them so as to determine whether there exists coding opportunity between or among forwarding packets. However, transmission overhead of this message collection and computing cost of opportunity determination will degrade the performance of these mechanisms. It is observed that coding opportunity at relay node is much more related with the local topology, and the opportunity of encoding three or more packets together is far less than that of encoding two packets together in wireless network with general density. Based on this, a new coding-aware routing mechanism, named TCAR, is proposed. TCAR ignores the oppommity of encoding three or more than three packets together. Each relay node maintains an encoding mapping table being established according to the result of its local topology detection, which can be used to calculate the path cost during routing setup phase, and determine that which two packets can be encoded together during the packets forwarding phase. In TCAR, instead of periodic messages collection, each relay nodes just need once local topology detection, and the encoding determination is much simpler than that of the former mechanisms. Simulation results show that compared with typical inter-session network coding mechanisms COPE and COPE-based routing, TCAR achieves 12% and 7% throughput gains, and keeps the minimum end to end delay. 展开更多
关键词 inter-session network coding coding aware routing wireless network
在线阅读 下载PDF
New algorithm for variable-rate linear broadcast network coding 被引量:1
5
作者 夏寅 张惕远 黄佳庆 《Journal of Central South University》 SCIE EI CAS 2011年第4期1193-1199,共7页
To adjust the variance of source rate in linear broadcast networks, global encoding kernels should have corresponding dimensions to instruct the decoding process. The algorithm of constructing such global encoding ker... To adjust the variance of source rate in linear broadcast networks, global encoding kernels should have corresponding dimensions to instruct the decoding process. The algorithm of constructing such global encoding kernels is to adjust heterogeneous network to possible link failures. Linear algebra, graph theory and group theory are applied to construct one series of global encoding kernels which are applicable to all source rates. The effectiveness and existence of such global encoding kernels are proved. Based on 2 information flow, the algorithm of construction is explicitly given within polynomial time O(|E| |T|.ω^2max), and the memory complexity of algorithm is O(|E|). Both time and memory complexity of this algorithm proposed can be O(ωmax) less than those of algorithms in related works. 展开更多
关键词 network coding variable-rate linear broadcast heterogeneous network code construction algorithm
在线阅读 下载PDF
NC code generation for laser assisted turn-mill of various type of clovers and square section members 被引量:1
6
作者 CHOI Jun-young LEE Choon-man 《Journal of Central South University》 SCIE EI CAS 2012年第11期3064-3068,共5页
Laser assisted turning (LAT) is one of the advanced machining technologies, which uses laser power to heat the surface of a workpiece before the material is removed. It has several advantages of low manufacturing co... Laser assisted turning (LAT) is one of the advanced machining technologies, which uses laser power to heat the surface of a workpiece before the material is removed. It has several advantages of low manufacturing costs, high productivity and high qualities to machine difficult-to-cut materials such as silicon nitride, muUite, zirconia and Ni. A large part of studies on LAT have been focused on a round bar. With increasing demands for high quality products and high performance engineering system, the researches on LAT for clover and square section members are necessary. But, these workpieces are impossible to be machined on conventional CNC lathe and to generate NC code with current CAM softwares. As a basic research for combining LAT with a tilting index table type 5-axis machining center, i.e. laser assisted turn-mill, a new method is suggested to generate NC code that can process various types of clover and square section members through development of C++ program. 展开更多
关键词 laser assisted turning (LAT) nc code clover section turn-mill square section turn-mill
在线阅读 下载PDF
Independent component analysis to physical-layer network coding over wireless fading channels
7
作者 Bing Du Xiaoming Tao 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2013年第2期196-203,共8页
A new combinational technology is proposed,which is feasible to apply physical-layer network coding(PNC) to wireless fading channels by employing the harmful interference strategically.The key step of PNC is that so... A new combinational technology is proposed,which is feasible to apply physical-layer network coding(PNC) to wireless fading channels by employing the harmful interference strategically.The key step of PNC is that sources broadcast signals simultaneously without orthogonal scheduling.Naturally,the signals overlap in the free space at the receivers.Since the signals from different sources are mutual independent,rooted on this rational assumption,an enhanced joint diagonalization separation named altering row diagonalization(ARD) algorithm is exploited to separate these signals by maximizing the cost function measuring independence among them.This ARD PNC(APNC) methodology provides an innovative way to implement signal-level network coding at the presence of interference and without any priori information about channels in fading environments.In conclusions,the proposed APNC performs well with higher bandwidth utility and lower error rate. 展开更多
关键词 physical-layer network coding (Pnc) independentcomponent analysis (ICA) joint diagonalization network coding cooperative communications.
在线阅读 下载PDF
Signal difference-based deadband H_∞ control approach for networked control systems with limited resources
8
作者 Yingying Liu Weiwei Che Yunkai Chu 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2015年第3期573-583,共11页
This paper investigates a signal difference-based dead- band H∞ control approach for networked control systems (NCSs) with limited resources. The effects of variable network-induced de- lays, sampling intervals and... This paper investigates a signal difference-based dead- band H∞ control approach for networked control systems (NCSs) with limited resources. The effects of variable network-induced de- lays, sampling intervals and data transmitting deadbands are con- sidered simultaneously and the model of the NCS is presented. A Lyapunov functional is adopted, which makes full use of the network characteristic information including the bounds of net- work delay (BND), the bounds of sampling interval (BSI) and the bounds of transmission deadband (BTD). In the meanwhile, the new H∞ performance analysis and controller design conditions for the NCSs are proposed, which describe the relationship of BND, BSI, BTD and the system's performance. Three examples are used to illustrate the advantages of the proposed methods. The results have shown that the proposed method not only effectively reduces the data traffic, but also guarantees the system asymptotically sta- ble and achieves the prescribed H∞ disturbance attenuation level. 展开更多
关键词 networked control system (ncS) H∞ control vari-able network-induced delay variable sampling interval transmission deadband.
在线阅读 下载PDF
桃红四物汤对大脑中动脉闭塞大鼠lncRNA表达的影响 被引量:1
9
作者 张丽娟 费长义 +6 位作者 余超 薛苏君 李雨朦 李静静 潘凌宇 段贤春 彭代银 《中国药理学通报》 CAS CSCD 北大核心 2024年第3期582-591,共10页
目的研究中药复方桃红四物汤(Tao Hong Si Wu decoction,THSWD)治疗大脑中动脉闭塞(middle cerebral artery occlusion,MCAO)大鼠长链非编码RNA(long non-coding RNA,lncRNA)的表达,并确定THSWD治疗MCAO大鼠可能的分子机制。方法从对照... 目的研究中药复方桃红四物汤(Tao Hong Si Wu decoction,THSWD)治疗大脑中动脉闭塞(middle cerebral artery occlusion,MCAO)大鼠长链非编码RNA(long non-coding RNA,lncRNA)的表达,并确定THSWD治疗MCAO大鼠可能的分子机制。方法从对照组、MCAO组和MCAO+THSWD组各获得3个大脑半球组织。采用RNA测序技术鉴定三组中的lncRNA基因表达。鉴定了THSWD调节的lncRNA基因,然后构建了THSWD调节的lncRNA-mRNA网络。通过MCODE插件鉴定lncRNA-mRNA网络的模块。基因本体(gene ontology,GO)和京都基因与基因组百科全书数据库(kyoto encyclopedia of genes and genomes,KEGG)用于分析富集的生物功能和信号通路。鉴定了THSWD调节的lncRNA的顺式和反式调控基因。采用逆转录实时定量聚合酶链式反应(RT-qPCR)验证lncRNA。分子对接用于验证lncRNA-mRNA网络靶点和通路相关蛋白结合能力。结果在MCAO大鼠中,THSWD共调节了302个lncRNA。生物信息学分析表明,一些核心lncRNA可能在THSWD治疗MCAO大鼠中发挥重要作用,此外,我们进一步发现THSWD可能也通过lncRNA-mRNA网络以及网络富集的补体和凝血级联反应等多通路治疗MCAO大鼠。分子对接结果表明,THSWD活性化合物没食子酸和苦杏仁苷与蛋白质靶点具有一定的结合能力。结论THSWD可以通过调节lncRNA保护MCAO大鼠脑损伤,为THSWD治疗缺血性中风提供了新见解。 展开更多
关键词 桃红四物汤 大脑中动脉闭塞 长链非编码RNA lncRNA-mRNA网络 生物信息学 分子对接
在线阅读 下载PDF
CodeScore-R:用于评估代码合成功能准确性的自动化鲁棒指标
10
作者 杨光 周宇 +1 位作者 陈翔 张翔宇 《计算机研究与发展》 EI CSCD 北大核心 2024年第2期291-306,共16页
评估指标在代码合成领域中至关重要.常用的代码评估指标可以分为3种类型:基于匹配、基于语义和基于执行.其中,基于执行的Pass@k指标通过执行测试用例,能够准确判断预测代码的功能准确性.然而,该指标的计算需要大量开销,因此亟需设计一... 评估指标在代码合成领域中至关重要.常用的代码评估指标可以分为3种类型:基于匹配、基于语义和基于执行.其中,基于执行的Pass@k指标通过执行测试用例,能够准确判断预测代码的功能准确性.然而,该指标的计算需要大量开销,因此亟需设计一种自动化评估指标,在无需测试用例时仍可评估预测代码的功能准确性.此外,好的评估指标应当具有鲁棒性,即预测代码发生微小改变时,评估指标仍能保持其准确性.为此,提出了一种基于UniXcoder和对比学习的自动化鲁棒指标CodeScore-R,用于评估代码合成的功能准确性. CodeScore-R采用草图化处理、语法等价转换和变异测试等技术手段,有效减轻了标识符、语法结构和运算符对评估结果的干扰.实验结果表明,在Java和Python语言上的代码生成和迁移任务中,CodeScore-R的表现优于其他无需测试用例的评估指标,且更接近Pass@k指标,并具有更强的鲁棒性. 展开更多
关键词 代码合成评估指标 功能准确性 鲁棒性 代码合成 神经网络
在线阅读 下载PDF
Design and analysis of a network coding algorithm for ad hoc networks
11
作者 王远 徐华 贾培发 《Journal of Central South University》 SCIE EI CAS CSCD 2015年第4期1358-1365,共8页
Network coding is proved to have advantages in both wireline and wireless networks. Especially, appropriate network coding schemes are programmed for underlined networks. Considering the feature of strong node mobilit... Network coding is proved to have advantages in both wireline and wireless networks. Especially, appropriate network coding schemes are programmed for underlined networks. Considering the feature of strong node mobility in aviation communication networks, a hop-by-hop network coding algorithm based on ad hoc networks was proposed. Compared with COPE-like network coding algorithms, the proposed algorithm does not require overhearing from other nodes, which meets confidentiality requirements of aviation communication networks. Meanwhile, it does save resource consumption and promise less processing delay. To analyze the performance of the network coding algorithm in scalable networks with different traffic models, a typical network was built in a network simulator, through which receiving accuracy rate and receiving delay were both examined.The simulation results indicate that, by virtue of network coding, the proposed algorithm works well and improves performance significantly. More specifically, it has better performance in enhancing receiving accuracy rate and reducing receiving delay, as compared with any of the traditional networks without coding. It was applied to both symmetric and asymmetric traffic flows and, in particular, it achieves much better performance when the network scale becomes larger. Therefore, this algorithm has great potentials in large-scale multi-hop aviation communication networks. 展开更多
关键词 ad hoc network network coding algorithm aviation communication network
在线阅读 下载PDF
Analysis and Design of Wireless Networked Control System Utilizing Adaptive Coded ModulationAnalysis and Design of Wireless Networked Control System Utilizing Adaptive Coded Modulation 被引量:4
12
作者 YANG Li GUAN Xin-Ping LONG Cheng-Nian LUO Xiao-Yuan 《自动化学报》 EI CSCD 北大核心 2009年第7期911-918,共8页
关键词 无线网络控制系统 计算机技术 自动化系统 网络条件
在线阅读 下载PDF
AVariablebitrateVideoCodingSchemeBasedonMorphologicalOperationsforATMNetworks
13
作者 ChenHuifang DaiWenqi 《通信学报》 EI CSCD 北大核心 1998年第5期84-89,共6页
AVariablebitrateVideoCodingSchemeBasedonMorphologicalOperationsforATMNetworksChenHuifangDaiWenqiQiuPeiliang(... AVariablebitrateVideoCodingSchemeBasedonMorphologicalOperationsforATMNetworksChenHuifangDaiWenqiQiuPeiliang(InstituteofInfo... 展开更多
关键词 视频编码 ATM网络 异步传输网 多位率 数学形态操作
在线阅读 下载PDF
基于差异性隔离和复用的网络切片无线资源分配方案 被引量:2
14
作者 孙君 霭振宇 《通信学报》 北大核心 2025年第3期109-121,共13页
为研究网络切片无线资源复用,同时考虑复用、隔离和优先级三者之间的权衡问题,提出了一种基于差异性隔离和复用的网络切片无线资源分配方案。在现有文献成果基础上,重新定义复用增益和隔离因子2个参数,以复用增益和隔离因子构建加权和函... 为研究网络切片无线资源复用,同时考虑复用、隔离和优先级三者之间的权衡问题,提出了一种基于差异性隔离和复用的网络切片无线资源分配方案。在现有文献成果基础上,重新定义复用增益和隔离因子2个参数,以复用增益和隔离因子构建加权和函数,并引入切片优先级。为求解优化问题设计了复用隔离优先级无线接入(MIPWA)算法,该算法基于改进的遗传算法(GA),引入矩阵编码、轮盘赌选择和最优保留方法来解决问题。结果表明,MIPWA算法使切片1、切片2和切片3的隔离性能分别提高了66.37%、52.73%和21.16%,复用增益仅损失了5.82%、3.86%和3.50%。与仅考虑隔离的算法相比,复用增益分别提高了65.35%、52.74%和22.81%,隔离增益仅损失了2.85%、3.85%和1.86%。以复用增益为例,3个切片下MIPWA算法的优化结果要比传统GA分别高出5.07%、1.81%和1.4%。 展开更多
关键词 网络切片 资源分配 无线资源隔离 无线资源复用 矩阵编码
在线阅读 下载PDF
考虑裂纹分形维数的平行黏结模型细观参数标定的神经网络模型
15
作者 龚囱 戚燕顺 +4 位作者 缪浩杰 肖琦 熊良锋 曾鹏 赵奎 《岩土力学》 北大核心 2025年第1期327-336,共10页
针对试错法在平行黏结模型细观参数标定过程中存在繁琐耗时,且无法定量评价数值模拟与室内试验的裂纹匹配程度等局限性,统计并分析了近10年平行黏结模型细观参数取值范围,采用盒计数法获取了数值模拟试验、室内试验所得破坏后岩石表面... 针对试错法在平行黏结模型细观参数标定过程中存在繁琐耗时,且无法定量评价数值模拟与室内试验的裂纹匹配程度等局限性,统计并分析了近10年平行黏结模型细观参数取值范围,采用盒计数法获取了数值模拟试验、室内试验所得破坏后岩石表面裂纹分形维数。在此基础上,建立了以宏观弹性模量、宏观泊松比、峰值强度和裂纹分形维数等4个参数为输入层,黏结弹性模量、黏结法向与切向刚度比、黏结内聚力、黏结内摩擦角、黏结抗拉强度和摩擦系数等6个细观参数为输出层的神经网络模型,对比分析了考虑与不考虑裂纹分形维数时平行黏结模型细观参数标定效果。研究结果表明:(1)所建立的神经网络模型具有较好的收敛速度、预测精度与泛化性能,测试集输出数据与期望值误差约为3.34%。(2)将裂纹分形维数纳入神经网络模型后,数值模拟所得弹性模量、峰值应力与泊松比等宏观参数与室内试验结果的误差小于3.00%,优于不考虑裂纹分形维数标定结果。(3)该方法可定量保障数值模拟所得裂纹不规则性与室内试验结果的一致性,其在一定程度上可视为对现有神经网络模型细观参数标定结果的修正。研究成果可为提高平行黏结模型细观参数标定效果提供新思路。 展开更多
关键词 分形维数 颗粒流 平行黏结模型 参数标定 神经网络
在线阅读 下载PDF
面向高分辨率图像传输的CNN网络编码方案研究
16
作者 刘娜 杨颜博 +2 位作者 张嘉伟 李宝山 马建峰 《西安电子科技大学学报》 北大核心 2025年第2期225-238,共14页
网络编码技术可以有效提升网络的吞吐率,然而,传统网络编码的编解码复杂度高且难以自适应环境噪声等动态因素的影响而容易导致解码失真,近年来有研究者引入神经网络以优化网络编码过程,但在高分辨率图像传输任务中,现有的神经网络编码... 网络编码技术可以有效提升网络的吞吐率,然而,传统网络编码的编解码复杂度高且难以自适应环境噪声等动态因素的影响而容易导致解码失真,近年来有研究者引入神经网络以优化网络编码过程,但在高分辨率图像传输任务中,现有的神经网络编码方案对高维度空间信息的捕捉能力不足,带来较大的通信及计算开销。为此,文中提出采用二维卷积神经网络(CNN)对各网络节点的编解码器进行参数化设计的联合源的深度学习网络编码方案,通过CNN捕捉深层空间结构信息并降低网络节点的计算复杂度。在信源节点,通过卷积层运算实现对传输数据的降维处理,提升数据的传输速率;在中间节点,接收来自两个信源的数据并通过CNN编码压缩至单个信道传输;在信宿节点,对接收到的数据利用CNN进行升维解码而恢复出原始图像。实验表明,在不同信道带宽占用比和信道噪声水平下,该方案在峰值信噪比和结构相似度上展现出优良的解码性能。 展开更多
关键词 网络编码 深度学习 卷积神经网络 高分辨率图像 图像通信
在线阅读 下载PDF
水声网络基于优先级与可Zigzag解码的在线喷泉码
17
作者 杜秀娟 王玉杰 +1 位作者 柳秀秀 赵建 《计算机应用研究》 北大核心 2025年第3期895-902,共8页
水声网络(underwater acoustic network,UAN)具有长传播时延、高误码率、半双工通信等特性,这些特性严重影响了UAN中数据的可靠传输。而在线喷泉码具有在线控制、编解码复杂度低、码率自适应等诸多优势,在线喷泉码适合于保障UAN中数据... 水声网络(underwater acoustic network,UAN)具有长传播时延、高误码率、半双工通信等特性,这些特性严重影响了UAN中数据的可靠传输。而在线喷泉码具有在线控制、编解码复杂度低、码率自适应等诸多优势,在线喷泉码适合于保障UAN中数据的可靠传输。针对递归与限制反馈的在线喷泉码(recursive OFC with limited feedback,ROFC-LF)存在不理想覆盖和4元环问题导致略高的开销和频繁的反馈,提出适用于UAN的基于优先级与可Zigzag解码的ROFC-LF(priority-based and zigzag-decodable ROFC-LF,P-ZROFC-LF)。P-ZROFC-LF在建立阶段选取具有最高优先级的原始包进行编码直至所有原始包均参与编码。同时,引入可Zigzag解码编码,将无用编码包进行移位异或转换为有用编码包来提高解码性能。通过随机图理论,分析P-ZROFC-LF所需编码包数与原始包数之间的关系。理论分析与仿真结果表明,与大部分在线喷泉码相比,P-ZROFC-LF显著提高了反馈和开销性能。其中P-ZROFC-LF相比于ROFC-LF的反馈和开销分别减少了18%和0.0176,更适用于UAN。 展开更多
关键词 水声网络 在线喷泉码 可Zigzag解码 反馈 开销
在线阅读 下载PDF
相关噪声下基于深度学习的LDPC码码率半盲识别算法
18
作者 袁磊 杨艳娟 +1 位作者 郭毅 戴鹏 《系统工程与电子技术》 北大核心 2025年第4期1335-1345,共11页
为了正确识别相关噪声下采用低密度奇偶校验码和高阶调制的无线通信系统的信道编码参数,在已知候选码率集合和相应奇偶校验矩阵的假定下,提出两种基于深度学习的码率半盲识别算法。所提神经网络由降噪子网络和码率识别子网络构成,降噪... 为了正确识别相关噪声下采用低密度奇偶校验码和高阶调制的无线通信系统的信道编码参数,在已知候选码率集合和相应奇偶校验矩阵的假定下,提出两种基于深度学习的码率半盲识别算法。所提神经网络由降噪子网络和码率识别子网络构成,降噪子网络设计实数降噪子网络和复数降噪子网络。相比于实数降噪子网络,复数降噪子网络以高复杂度为代价,获得更好的处理复信号的能力。进一步,为了降低复数降噪子网络的复杂度,提出一种基于网络剪枝技术的网络压缩算法。仿真实验结果表明,通过使用联合优化降噪损失函数和码率识别损失函数的多任务学习策略:一方面,在相关噪声下提出的神经网络比传统算法具有更好的识别性能;另一方面,当利用网络压缩算法将基于复数降噪子网络识别算法的复杂度降低到与基于实数降噪的子网络识别算法的复杂度相近时,其性能仍优于基于实数降噪子网络的识别算法。 展开更多
关键词 相关噪声 信道编码盲识别 低密度奇偶校验码 深度学习 复数网络
在线阅读 下载PDF
低密度奇偶校验码正则化神经网络归一化最小和译码算法
19
作者 周华 周鸣 张立康 《电子与信息学报》 北大核心 2025年第5期1486-1493,共8页
低密度奇偶校验(LDPC)码基于神经网络的归一化最小和(NNMS)译码算法按照网络中权重的共享方式可分为不共享(NNMS)、全共享(SNNMS)、部分共享(VC-SNNMS和CV-SNNMS)等。该文针对LDPC码在使用NNMS,VC-SNNMS和CV-SNNMS译码时因高复杂度导致... 低密度奇偶校验(LDPC)码基于神经网络的归一化最小和(NNMS)译码算法按照网络中权重的共享方式可分为不共享(NNMS)、全共享(SNNMS)、部分共享(VC-SNNMS和CV-SNNMS)等。该文针对LDPC码在使用NNMS,VC-SNNMS和CV-SNNMS译码时因高复杂度导致的过拟合问题,引入正则化(Regularization)优化了神经网络中边信息的权重训练,抑制了基于神经网络译码的过拟合问题,分别得到RNNMS,RVC-SNNMS和RCVSNNMS算法。仿真结果表明:采用共享权重可以减轻神经网络训练负担,降低LDPC码基于神经网络译码的误比特率(BER);正则化能有效缓解过拟合现象提升神经网络的译码性能。针对码长为576,码率为0.75的LDPC码,当误码率BER=10-6时,RNNMS,RVC-SNNMS和RCV-SNNMS算法相较于NNMS,VC-SNNMS和CV-SNNMS算法分别得到了0.18 dB,0.22 dB和0.27 dB的信噪比(SNR)增益,其中最佳的RVC-SNNMS算法相较于BP算法、NNMS算法和SNNMS算法,分别获得了0.55 dB,0.51 dB和0.22 dB的信噪比增益。 展开更多
关键词 低密度奇偶校验码 神经网络 归一化最小和译码 过拟合 正则化
在线阅读 下载PDF
新型多幅度编码在再调制无源光网络中的应用 被引量:1
20
作者 邱文昊 卢旸 +3 位作者 成亮 姚陈 翟彦蓉 毕美华 《浙江大学学报(工学版)》 北大核心 2025年第2期433-441,共9页
为了在使用再调制技术的波分复用无源光网络(WDM PON)中实现透明光网络单元(colorless ONU),提出2种新型的多幅度编码方式(PAM补码和可变幅线路编码).新的编码方式结合了频谱整形编码和脉冲幅度调制(PAM).频谱整形编码能够抑制下行信号... 为了在使用再调制技术的波分复用无源光网络(WDM PON)中实现透明光网络单元(colorless ONU),提出2种新型的多幅度编码方式(PAM补码和可变幅线路编码).新的编码方式结合了频谱整形编码和脉冲幅度调制(PAM).频谱整形编码能够抑制下行信号的低频分量,减少下行信号对上行再调制信号的串扰.高阶的PAM能够提高信号的编码效率,采用高阶PAM的下行信号可以被传统的二进制接收机接收.通过仿真验证了所提编码方式在再调制WDM PON中应用的可行性,测试了速率为625 Mb/s和1.25 Gb/s的上行信号再调制在速率为10Gb/s的下行信号上的性能.基于新型多幅度编码的再调制WDM PON方案能够减少上行再调制信号遭受的串扰,提高下行信号的编码效率. 展开更多
关键词 无源光网络 透明光网络单元 再调制 多幅度编码 编码效率
在线阅读 下载PDF
上一页 1 2 121 下一页 到第
使用帮助 返回顶部