A Reduced Search Soft-Output Detection Algorithm and Its Application to Turbo-Equalization

To decrease the complexity of MAP algorithm, reduced state or reduced search techniques can be applied. In this paper we propose a reduced search soft output detection algorithm fully based on the principle of M algorithm for turbo equalization, which is a suboptimum version of the Lee algorithm. This algorithm is called soft output M algorithm (denoted as SO M algorithm), which applies the M strategy to both the forward recursion and the extended forward recursion of the Lee algorithm. Computer simulation results show that, by properly selecting and adjusting the breadth parameter and depth parameter during the iteration of turbo equalization, this algorithm can obtain good performance and complexity trade off.

Blind Equalization Algorithm Based on MCMA and DSE-CMA

A new blind equalization algorithm based on the modified constant modulus algorithm （MCMA） and dithered signederror constant modulus algorithm （DSE-CMA） is proposed. This dithered signed-error MCMA （DSE-MCMA） can not only reduce the computational complexity, but also recover the phase rotation in the complex channel. Simulation results have verified the analysis and indicated the good property of DSE-MCMA.

Improved Algorithm of Histogram Equalization and Its Actualization

A new improved algorithm of histogram equalization was discussed and actualized by analyzing the traditional algorithm. This improved algorithm has better effect than the traditional one, especially it is used to process poor quality images.

TURBO AIDED CYCLIC PREFIX RECONSTRUCTION FOR CODED SINGLE-CARRIER SYSTEMS WITH FREQUENCY-DOMAIN EQUALIZATION (SC-FDE)

In this paper, a Turbo aided Cyclic Prefix (CP) reconstruction scheme, termed Turbo- CPR, is proposed for Single-Carrier systems with Frequency-Domain Equalization (SC-FDE) that employ insufficient CP in the transmitter. In Turbo-CPR, the decoder output is incorporated in the process of equalization, i.e. Turbo equalizer is employed. It is shown in the simulation results that Turbo-CPR not only recovers the performance loss due to insufficiency of CP, but also provides extra gains over the lower bound of performance for conventional CP reconstruction schemes.

BLIND EQUALIZATION OF MIMO SYSTEMS BASED ON ORTHOGONAL CONSTANT MODULUS ALGORITHM

This paper investigates adaptive blind source separation and equalization for Multiple Input Multiple Output (MIMO) systems. To effectively recover input signals, remove Inter-Symbol Interference (ISI) and suppress Inter-User Interference (IUI), the array input is first transformed into the signal subspace, then with the derived orthogonality between weight vectors of different input signals, a new orthogonal Constant Modulus Algorithm (CMA) is proposed. Computer simulation results illustrate the promising performance of the proposed method. Without channel identification, the proposed method can recover all the system inputs simultaneously and can be adaptive to channel changes without prior knowledge about signals.

An Orthogonal Wavelet Transform Fractionally Spaced Blind Equalization Algorithm Based on the Optimization of Genetic Algorithm

An orthogonal wavelet transform fractionally spaced blind equalization algorithm based on the optimization of genetic algorithm(WTFSE-GA) is proposed in viewof the lowconvergence rate,large steady-state mean square error and local convergence of traditional constant modulus blind equalization algorithm(CMA).The proposed algorithm can reduce the signal autocorrelation through the orthogonal wavelet transform of input signal of fractionally spaced blind equalizer,and decrease the possibility of CMA local convergence by using the global random search characteristics of genetic algorithm to optimize the equalizer weight vector.The proposed algorithm has the faster convergence rate and smaller mean square error compared with FSE and WT-FSE.The efficiency of the proposed algorithm is proved by computer simulation of underwater acoustic channels.

Blind Equalization Using a Hybrid Algorithm of Multilayer Neural Network

A novel blind equalization scheme based on multilayer neural network and Higher OrderCumulants(HOC)is proposed in the paper.The training of the neural network uses a newhybrid algorithm which has strict convex character(after a threshold)and converges muchfaster than the CMA algorithm.The inverse channel is built on the basis of the estimatedchannel and the training of neural network.The scheme can be used in nonlinear and timevarying channel and to deal with PAM or QAM signals.Simulation results Show that it per-forms well for blind equalization.

EFFICIENT CYCLIC PREFIX RECONSTRUCTION USING DECISION FEEDBACK FILTER FOR CODED SINGLE-CARRIER SYSTEMS WITH FREQUENCY-DOMAIN EQUALIZATION (SC-FDE)

In this paper, an efficient Cyclic Prefix （CP） reconstruction scheme is proposed for Single-Carrier systems with Frequency-Domain Equalization （SC-FDE） that employ insufficient length of CP at the transmitter. By utilizing a decision feedback filter to cancel the residual InterSymbol Interference （ISI） in the equalized signal, the proposed scheme can effectively lower the low bound of performance for the CP reconstruction schemes and can greatly improve the Bit Error P^te （BER） performance of SC-FDE systems. In addition, the existing methods and the proposed scheme are also optimized. It is shown in the simulation results that, when the Signal-to-Noise Ratio （SNR） exceeds a certain threshold, the proposed scheme can achieve the low bound of performance for the existing methods. Moreover, by increasing the number of iteration or through optimization, the low bound can be outperformed.

The Fast Blind Equalization Algorithm with Global Convergence

ＴｈｅＦａｓｔＢｌｉｎｄＥｑｕａｌｉｚａｔｉｏｎＡｌｇｏｒｉｔｈｍｗｉｔｈＧｌｏｂａｌＣｏｎｖｅｒｇｅｎｃｅＬｕＪｕｎＬｉＴｏｎｇ（ＩｎｓｔｉｔｕｔｅｏｆＩｎｆｏｒｍａｔｉｏｎａｎｄＥｎｇｉｎｅｅｒｉｎｇｏｆＰＬＡ，Ｚｈｅｎｇｚｈｏｕ４５０００２）Ａ...

Fast Fading Channel Neural Equalization Using Levenberg-Marquardt Training Algorithm and Pulse Shaping Filters

Artificial Neural Network (ANN) equalizers have been successfully applied to mitigate Inter symbolic Interference (ISI) due to distortions introduced by linear or nonlinear communication channels. The ANN architecture is chosen according to the type of ISI produced by fixed, fast or slow fading channels. In this work, we propose a combination of two techniques in order to minimize ISI yield by fast fading channels, i.e., pulse shape filtering and ANN equalizer. Levenberg-Marquardt algorithm is used to update the synaptic weights of an ANN comprise only by two recurrent perceptrons. The proposed system outperformed more complex structures such as those based on Kalman filtering approach.

A Variable Metric Algorithm with Broyden Rank One Modifications for Nonlinear Equality Constraints Optimization

In this paper, a variable metric algorithm is proposed with Broyden rank one modifications for the equality constrained optimization. This method is viewed expansion in constrained optimization as the quasi-Newton method to unconstrained optimization. The theoretical analysis shows that local convergence can be induced under some suitable conditions. In the end, it is established an equivalent condition of superlinear convergence.

Single carrier frequency domain equalization and diversity combining for cooperative systems

A low-complexity single carrier frequency-domain equalizer （SC/FDE） and diversity combining method for cooperative systems with demodulate-and-forward relaying over frequency-selective channels is proposed. At the relay nodes, linear SC/FDE is adopted and normalized correlation coefficlent is mtrodueed to derive an equivalent source-to-relay-destination （S-R-D） channel that is highlighted in this study. At the destination, a joint SC/FDE and diversity combining receiver is proposed by utilizing the equivalent S-R-D channel. Simulation results demonstrate the superiority of the proposed SC/FDE scheme over the straightforward SC/FDE which ignores the decisions errors at the intermediate relay nodes.

Joint receiving mechanism based on blind equalization with variable step size for M-QAM modulation

The problem of inter symbol interference（ ISI） in wireless communication systems caused by multipath propagation when using high order modulation like M-Q AMis solved. Since the wireless receiver doesn＇t require a training sequence,a blind equalization channel is implemented in the receiver to increase the throughput of the system. To improve the performances of both the blind equalizer and the system,a joint receiving mechanismincluding variable step size（ VSS） modified constant modulus algorithms（ MC-MA） and modified decision directed modulus algorithms（ MD DMA） is proposed to ameliorate the convergence speed and mean square error（ MSE） performance and combat the phase error when using high order QAM modulation. The VSS scheme is based on the selection of step size according to the distance between the output of the equalizer and the desired output in the constellation plane. Analysis and simulations showthat the performance of the proposed VSS-MCMA-MD DMA mechanismis better than that of algorithms with a fixed step size. In addition,the MCMA-MDDMA with VSS can performthe phase recovery by itself.

面向栅格地图的区域渐进均分算法

单架无人机续航能力限制了区域全覆盖侦察,合理的区域划分是实现多无人机协同全域侦察的关键。栅格法规划侦察区域是无人机区域侦察的常用研究方法。为了解决栅格地图等量划分的问题,提出了一种面向栅格地图的区域渐进均分算法。算法由4个阶段构成。阶段1,建立区域边界确认的跳跃迭代法,根据栅格的特点制定判定条件,进行栅格特征标识。阶段2,提出一种双特征标识方法,对射线法进行改进,确定区域内部栅格。阶段3,模仿水波扩散,提出了一种邻边扩散法,实现区域初步的扩散分割。阶段4,设计补偿规则,通过邻边补偿算法,对各子区域栅格数进行数量补偿。实验证明,区域渐进均分算法相较于其他算法,具有较好的聚集性,连续性和均匀性,为多无人机协同全域侦察提供了理论保证。

人工智能发展和安全并重的法治探究——以人形机器人为例

基于“大数据+高算力+强算法”的人形机器人,能够实现连贯高效的人—机—环境智能交互,可服务特种领域需求,助力智能制造,成为人类生活助手和情感伴侣。尽管人形机器人存在功能安全、网络安全、个人信息安全、数据安全、国家安全、违法“代理”、侵权、伦理等多种风险,但如果过度追求安全,就会扼杀创新而阻碍发展。发展是安全的基础,安全是发展的条件,应实现人工智能发展和安全并重。通过理性实施包容审慎监管,推动人形机器人在发展中规范、在规范中发展。科学合理地开展安全风险评估,将人形机器人风险控制在可接受的水平。完善数据安全认证、个人信息保护认证等第三方声誉评价机制,并分级分类推动人形机器人标准制定。应合理配置相关主体的责任,避免人形机器人的过度拟人化,防止陷入“人形机器人陷阱”。通过完善的伦理规范约束人工智能研发者,并为人形机器人“加载”道德,使其对人和环境友好。

基于扇形链路策略的改进蚁群分簇路由协议

针对网络覆盖区域较大、节点数量较多的无线传感器网络,容易出现部分节点过早死亡等情况,提出一种基于扇形链路策略的改进蚁群分簇路由协议RACO-SL。通过加入奖惩因子,同时对精英个体采用蚁群优化算法的概率生成新的后代个体,对于普通个体,通过与随机选择的精英个体进行交叉变异操作,改进蚁群优化算法,以整个网络每次通信的能耗为优化目标选取较优的簇头节点集。为待转发簇头节点设计从可动态调节的扇形区域中选择下一跳中继节点的链路转发策略。实验结果表明,与现有协议相比,该协议在延长网络寿命、提高通信链路质量、增强网络覆盖度方面表现良好。

港口场景集装箱锁孔目标智能检测研究

为解决集装箱图像受到外界光线环境的影响造成信息对比度低、暗部细节信息不明显和图像信息难以辨认清等问题,提出一种新的图像增强算法,其重点是利用非线性变换,提高图像的暗部细节,将原本的RGB模型转化到HSV模型进行均衡化。利用实际工程中港口所提供的测试数据集,使用已训练好的模型进行广泛的实验。实验还探究了图像增强算法对卷积神经网络性能的影响,利用对比度受限的自适应直方图均衡化、伽马校正、拉普拉斯变换以及原始图像与该算法进行对比,使用10折交叉验证了该算法的精确率、召回率均大于其他的算法。采用配对T检验,比较分析采用各算法间的各项指标差异,结果表明:该算法相比其他算法的效果要好。

混合级联H桥逆变器的新型SHEPWM功率均衡控制策略

对于直流侧电压比为1∶1∶2的混合级联H桥逆变器,传统的指定谐波消除脉宽调制法(SHEPWM)存在各单元间输出功率不均衡问题,因此,提出一种双层功率均衡调制策略。首先,通过对消谐方程组中的基波幅值方程表达式进行等效拆分,使2个低压H桥单元的输出电压基波幅值之和与高压H桥单元的输出电压基波幅值相等,实现了高低压H桥单元间的功率均衡,即外层功率均衡。其次,利用逻辑运算的方法对低压单元的初始驱动信号进行逻辑重组,实现2个低压单元间的功率均衡,即内层功率均衡;然后使用改进的实数型遗传算法(IRCGA-2)求出消谐方程组的解,消除指定的谐波,得到高质量的输出电压。最后,通过仿真和实验证明了调制策略的正确性与有效性。

基于ACE与YOLOv5的电力遥感图像检测算法

针对电力遥感图像采集时存在大量浓烟以及数据样本少等问题,提出一种基于对数变换的改进型自动色彩均衡与改进后YOLOv5s模型的遥感图像去雾检测算法,旨在通过提高数据集的图像质量进而提高检测网络的检测精度。构建的改进型自动色彩均衡对电力遥感图像去雾增强,并通过图像质量和特征提取两方面进行了实验数据对比,实验结果表明改进的自动色彩均衡算法优于其它算法。其次,通过YOLOv5s检测算法对增强后的数据集进行训练,引入mosaic数据增强算法,并通过构建ghost卷积模块和NAM注意力模块降低了网络参数、提升了网络检测精度。

基于自适应模糊PID算法的锂电池组双层均衡控制

为提高电池重组时的均衡效率,在传统Buck-Boost均衡拓扑电路的基础上,设计了一种锂电池组双层均衡拓扑电路。组内采用Buck-Boost电路均衡,组间利用双向反激变压器进行均衡。均衡控制策略采用自适应模糊PID算法,以电池荷电状态(state of charge, SOC)为均衡变量,利用模糊控制算法对PID参数进行调节,缩短了均衡时间,提高了均衡效率。在Matlab/Simulink中搭建了锂电池组双层均衡拓扑电路和自适应模糊PID控制算法模型。实验结果表明:在不同工作状态下,所提出的电池组均衡拓扑及其控制策略将均衡时间效率平均提高了58.36%,验证了该方案的有效性。