基于Pareto支配的MPRM电路面积与功耗优化

针对MPRM电路的面积与功耗折衷优化问题,提出一种基于多目标三值多样性粒子群MOTDPSO算法的最佳极性搜索方案。在三值多样性粒子群算法求解MPRM电路综合优化问题的基础上,对超出定义的边界范围的粒子,执行边界约束处理,并结合Pareto支配概念改进算法;然后建立基于Pareto支配的粒子与MPRM电路极性之间的参数映射关系,并结合面积与功耗估计模型以及 OR/XNOR电路混合极性转换方法,将该算法应用于MPRM电路的面积和功耗优化。最后对18个PLA格式MCNC Benchmark电路进行测试,与NSGA-II算法搜索到的结果相比,MOTDPSO算法获取的最优解的面积平均优化率为4.29%,功耗平均优化率为6.02%。     

可重配置处理器的体系结构级功耗模型与分析

按照可重配置处理器的体系结构建立并实现功耗模型;模型对处理器的电路级特性进行抽象,基于体系结构级属性和工艺参数进行静态峰值功耗估算,基于性能模拟器进行动态功耗统计,并实现三种条件时钟下的门控技术;可重配置处理器与超标量通用微处理器相比,在性能方面获得的平均加速比为3.59,而在功耗方面的平均增长率仅为1.48;通过实验还说明采用简单的CC1门控技术能有效地降低可重配置系统的功耗和硬件复杂度;该模型为可重配置处理器低功耗设计和编译器级低功耗优化研究奠定了基础。     

计数器驱动的IEEE 802.16e休眠模式研究

为了降低移动终端的能量消耗以及对基站空口资源的占用,IEEE 802.16e标准定义了休眠模式。针对计数器驱动的休眠工作模式与IEEE 802.16e标准不能较好兼容的局限,提出了改进的计数器驱动休眠算法,建立马尔可夫链模型分析其性能,推导出平均时延、平均功耗与计数器门限值等参数之间的内在关系,在此基础上给出了不同QoS要求的连接所对应的参数设置方案。     

高层次综合中基于整数线性规划模型的多目标功耗优化算法

考虑峰值周期功耗和峰值模块功耗的同时优化,并尽可能地降低电路的功耗时延乘积指标.利用参数化功能单元库中各个功能模块的具体物理信息,在多供电电压、多调度周期模式下,建立了整数线性规划模型及其相应的6组约束条件,并将高层次综合中的调度过程和功能单元的绑定过程统一起来进行了模型化.文中算法已经成功地应用到自行开发的高层次综合工具之中,算法中得到的数据结果可以直接用于下一步的布图规划.对测试用例的实验进一步说明:同时优化峰值周期功耗和峰值模块功耗可以取得更好的综合结果,并且改善功耗时延乘积项的值(平均降低了30%),提高电路的可靠性和稳定性.     

电热分析研究的现状与展望

随着IC工艺进入纳米工艺时代,集成度与工作频率的增加,伴随性能提高的是不断增加的芯片功耗.高功耗的直接后果是产生了高供电电流和高功耗密度,而过大的供电电流降低了供电网络的供电电压;过大的功耗密度升高了内核温度,反过来又会增加电路时延,降低芯片的性能.所以在芯片设计中,必须对芯片功耗、供电网络、3D热分析进行快速而精确的电热分析;同时,漏电流功耗随工作温度升高而明显增加所造成的电热耦合效应,以及纳米工艺所带来的较大工艺参数变化,都提高了电热分析的难度.文中给出了电热参量(功耗、供电电压和内核温度)分析的重要性与研究现状,展望了纳米工艺下日益显著的工艺参数变化对电热分析所带来的挑战.     

面向芯核设计的功耗层次化管理策略

在对现有的各种低功耗设计技术和动态功耗管理策略进行研究的基础上,提出了一种适合于面向芯核设计的层次化电路功耗层次化管理策略及其实现方法．通过让系统在处于空闲状态时迅速进入极低功耗模式的方法来降低功耗,不但能很好地完成复杂系统的功耗管理功能,而且具有较好的可扩展性．将此方法应用于一款数百万门级的片上系统设计中的实验结果表明,在对芯片面积和性能影响不大的情况下,在很大程度上实现了功耗管理,并大幅度地降低了系统功耗。     

低功耗低电源线噪声纳米CMOS全加器

提出一种低功耗低电源线噪声的纳米CMOS全加器。采用电源门控结构的全加器来降低纳米CMOS电路的漏电功耗,改进了传统互补CMOS全加器的求和电路,减少了所需晶体管的数目,并进一步对休眠晶体管的尺寸和全加器的晶体管尺寸进行了联合优化。用Hspice在45nmCMOS工艺下的电路仿真结果表明,改进后的全加器电路在平均功耗时延积、漏电功耗和电源线噪声等方面取得了很好的效果。     

高层次综合与布图规划相结合的方法与技术

介绍了高层次综合与布图规划相结合的基本方法与技术及其研究进展．该方法主要解决集成电路制造工艺的持续发展给集成电路电路设计所带来的2个问题：集成电路本身的集成复杂度使得集成电路的设计工作必须向更高的抽象层次前进;集成电路特征尺寸的缩小导致物理寄生效应已经在电路的时延、功耗等指标上成为主导因素,这要求在更高的抽象层次关注物理参数的影响．对本领域有代表性的算法进行了系统的描述,并且对这些算法的基本思路进行了分析和总结．     

一种基于最大偏差的AES功耗分析攻击方法

针对高级加密标准(AES)密码算法的电路实现,提出了一种改进的功耗攻击方法.该方法的基本思想是选取2次不同明文输入下的汉明重量差为改进功耗模型,通过选择明文能够最大可能性地增大功耗偏差,从而恢复出密钥.采用UMC 0.25 μm 1.8 v标准CMOS工艺库,利用Synopsys公司的EDA工具得到AES电路加密过程的功耗仿真曲线,建立起功耗攻击平台,并在此平台上进行多种功耗攻击方法的分析和比较.实验结果表明,与普通的差分功耗分析(DPA)和相关功耗分析(CPA)攻击方法比较,提出的改进攻击方法能够以适当的功耗测量次数,以及更小的计算复杂度实现DPA攻击.     

用神经网络求解性能驱动的电路划分问题

文中考虑一种以连线代价最小目标的,以面积和时延为约束的,划分块与划分块之间有确定和拓扑关系的电路划分问题,提出了一个性能驱动电路划分的均场退火算法。算法通过换拉矩阵把问题映射为神经网络,并建立了包含优化目标项,面积约束项和时延约束项的能量函数,再用均场退火方程失代求解。     

Message length adaptive LDPC codes

LDPC codes achieve better performance and lower decoding complexity than turbo codes, with a major drawback of high encoding complexity. The encoder generator matrix is derived from the inverse of portion of parity check matrix. If the message length is changed, the structure of parity check matrix is modified and hence, the generator matrix must be re-computed. This increases the encoding complexity as the computation of matrix inverse is time and resource consuming operation. In this paper, we consider the encoding problem for LDPC codes as the complexity of encoding is essentially quadratic with respect to the block length. Using an efficient encoding method proposed by Richardson and Urbanke, we propose a systematic procedure to construct parity check matrix and generator matrix such that with change in message length, the re-computation for constructing generator matrix is avoided. The presented design uses fixed sub-matrices to construct a semi-random parity check matrix. The resultant design will reduce the pre-computation time of converting parity check matrix to generator matrix. The reported encoder reduces encoding time without the loss of coding gain and Bit Error Rate (BER) performance.     

基于RLS的改进型盲多用户检测算法

针对传统的关于可变约束MOE盲多用户检测算法需直接进行特征值分解计算和不能实现二次约束限制的缺点,提出了一种基于RLS的改进型MOE盲多用户检测算法.该算法避免了矩阵求逆和大量的特征值分解运算,计算复杂度大大降低.仿真结果表明,该算法的误码率性能和收敛性均优于直接利用特征值分解的传统算法,且程序运行时间大大缩小.     

Model complexity reduction of chemical reaction networks using mixed-integer quadratic programming

The model complexity reduction problem of large chemical reaction networks under isobaric and isothermal conditions is considered. With a given detailed kinetic mechanism and measured data of the key species over a finite time horizon, the complexity reduction is formulated in the form of a mixed-integer quadratic optimization problem where the objective function is derived from the parametric sensitivity matrix. The proposed method sequentially eliminates reactions from the mechanism and simultaneously tunes the remaining parameters until the pre-specified tolerance limit in the species concentration space is reached. The computational efficiency and numerical stability of the optimization are improved by a pre-reduction step followed by suitable scaling and initial conditioning of the Hessian involved. The proposed complexity reduction method is illustrated using three well-known case studies taken from reaction kinetics literature.     

Designing parity check matrix to achieve linear encoding time in LDPC codes

Low-density parity-check (LDPC) codes have become the part of various communication standards due to their excellent error correcting performance. Existing methods require matrix inverse computation for obtaining a systematic generator matrix from parity check matrix. With the change in code rate or code length the process is repeated and hence, a large number of pre-processing computations time and resources are required. In the existing methods, the complexity of encoding is essentially quadratic with respect to the block length. In this paper, it is shown that the parity check matrix can be constructed using patterned sub-matrix structure such that the matrix inverse operation is replaced by matrix multiplication of sparse matrices. The sparseness of matrices is then utilized to obtain efficient encoders which can achieve encoding in real time with reduced pre-computation complexity. Hardware implementation of encoder and simulation results show that the proposed encoder achieves throughput in excess of 1 Gbps with the same error correcting performance as the conventional designs.     

Relaxed multi-view clustering in latent embedding space

Although many multi-view clustering approaches have been developed recently, one common shortcoming of most of them is that they generally rely on the original feature space or consider the two components of the similarity-based clustering separately (i.e., similarity matrix construction and cluster indicator matrix calculation), which may negatively affect the clustering performance. To tackle this shortcoming, in this paper, we propose a new method termed Multi-view Clustering in Latent Embedding Space (MCLES), which jointly recovers a comprehensive latent embedding space, a robust global similarity matrix and an accurate cluster indicator matrix in a unified optimization framework. In this framework, each variable boosts each other in an interplay manner to achieve the optimal solution. To avoid the optimization problem of quadratic programming, we further propose to relax the constraint of the global similarity matrix, based on which an improved version termed Relaxed Multi-view Clustering in Latent Embedding Space (R-MCLES) is proposed. Compared with MCLES, R-MCLES achieves lower computational complexity with more correlations between pairs of data points. Extensive experiments conducted on both image and document datasets have demonstrated the superiority of the proposed methods when compared with the state-of-the-art.     

一种面向H.264/AVC的新型宏块级码率控制算法

针对H.264/AVC码率控制模型的不足，提出了平均绝对变换量化误差（MATQD）的概念，以及基于MATQD的加权预测模型和新的二次R-D（Rate-Distortion）模型，然后给出完整的宏块级码率控制算法，其中，在预测宏块纹理码字分配时，提出了用MATQDratio计算宏块目标码字以及一个头码字的加权预测模型。实验结果表明，与JM8.5下的标准码率控制算法JVT-G012相比，本文提出的算法的码率控制精度更高，输出码流更加平稳，图像的平均质量更好。     

Low-complexity approximate iterative LMMSE detection for large-scale MIMO systems

This paper deals with iterative detection for uplink large-scale MIMO systems. The well-known iterative linear minimum mean squared error (LMMSE) detector requires quadratic complexity (per symbol per iteration) with the number of antennas, which may be a concern in large-scale MIMO. In this work, we develop approximate iterative LMMSE detectors based on transformed system models where the transformation matrices are obtained through channel matrix decompositions. It is shown that, with quasi-linear complexity (per symbol per iteration), the proposed detectors can achieve almost the same performance as the conventional LMMSE detector. It is worth mentioning that the linear transformations are also useful to reduce the complexity of downlink precoding, so the relevant computational complexity can be shared by both uplink and downlink.     

Quadratic polynomial guided fuzzy C-means and dual attention mechanism for medical image segmentation

Medical image segmentation is the most complex and important task in the field of medical image processing and analysis, as it is linked to disease diagnosis accuracy. However, due to the medical image's high complexity and noise, segmentation performance is limited. We propose a novel quadratic polynomial guided fuzzy C-means and dual attention mechanism composite network model architecture to address the aforementioned issues (QPFC-DA). It has mechanisms for channel and spatial edge attention, which guide the content and edge segmentation branches, respectively. The bi-directional long short-term memory network was added after the two content segmentation branches to better integrate multi-scale features and prevent the loss of important features. Furthermore, the fuzzy C-means algorithm guided by the quadratic polynomial can better distinguish the image's weak edge regions and has a degree of noise resistance, resulting in a membership matrix with less ambiguity and a more reliable segmentation result. We also conducted comparison and ablation experiments on three medical data sets. The experimental results show that this method is superior to several other well-known methods.     

Global exponential stability of discrete-time recurrent neural network for solving quadratic programming problems subject to linear constraints

In this paper, a discrete-time recurrent neural network with global exponential stability is proposed for solving linear constrained quadratic programming problems. Compared with the existing neural networks for quadratic programming, the proposed neural network in this paper has lower model complexity with only one-layer structure. Moreover, the global exponential stability of the neural network can be guaranteed under some mild conditions. Simulation results with some applications show the performance and characteristic of the proposed neural network.     

稳态大系统中采用局部反馈的价格法协调

本文对于稳态大系统递阶控制的校正机制,给出一个新的局部反馈方法。新的迭代法需要根据理论分析或试凑法来选择对角线阵K_c的系数,以保证迭代的稳定性。数字实例表明此法能提高系统的目标函数值。