基于坐标注意力机制融合的反无人机系统图像识别方法

反无人机系统是识别和打击“黑飞”无人机的有效手段,图像识别无人机是反无人机系统的关键之一。针对采集的无人机样本属于小样本、提取特征不够多,识别准确率不够高的问题,提出了一种基于迁移学习、密集卷积网络和坐标注意力机制融合的反无人机系统图像识别方法。首先,运用自制设备采集了多种无人机在不同背景下的图片,建立数据样本;其次,设计针对无人机小样本识别的基于迁移学习、坐标注意力机制和密集卷积网络融合的网络TL-CA4-DenseNet-121、基于通道注意力机制融合的网络TL-SE4-DenseNet-121等网络,运用设计的网络对小样本进行识别,并进行对比,然后分别进行了基于不同位置和不同个数的坐标注意力模块和通道注意力模块的网络识别实验;最后,将识别效果最优的网络与经典卷积神经网络模型进行对比实验。实验结果表明,提出的TL-CA4-DenseNet-121网络识别效果优于其他网络,识别的平均准确率为97.93%,F1-Score为0.9826,网络训练时间为6832 s。结果表明了该网络在识别小样本无人机方面的优越性和可行性。     

低轨卫星通信系统干扰分析及策略研究

卫星通信技术日益进步,低轨卫星通信系统已然成为新的发展趋势.由于频谱资源受限,低轨卫星通信系统可能面临自身及与外部地面和卫星系统的同频干扰问题.文章指出从外部系统对低轨卫星通信系统干扰、低轨卫星通信系统对外部系统干扰以及低轨卫星通信系统内部干扰3个方面进行了系统干扰场景分析,在此基础上,根据不同的干扰情况分别给出了外部...     

采用差分星座图的SDR设备射频指纹识别

差分星座图(Differential Constellation Trace Figure,DCTF)在射频指纹(Radio Frequency Fingerprint,RFF)识别中具有良好的性能,但在低信噪比情况下会产生严重的模糊。为了提高DCTF的识别精度,设计了注意力残差卷积神经网络模型,即DCTF-Res2Net模型。该模型在网络提取特征时加入注意力模块来提高DCTF的识别精度,并将标签平滑损失函数与网络模型结合,有效缓解了DCTF中的离群点对网络模型的干扰。在实验中,根据不同的传输方式和传输场景构建了完备的DCTF的数据集,并用DCTF-Res2Net模型对其进行分类。实验结果表明,在信噪比为5 dB的情况下,与传统的残差网络相比,所提出的DCTF-Res2Net模型可以达到更高的识别精度。     

宽带卫星通信信号调制识别及信道均衡技术研究

宽带卫星通信是现代通信系统中不可或缺的一部分。文中介绍了宽带卫星通信中的信号调制识别和信道均衡技术,这些技术可以提高系统的性能和稳定性,从而更好地应对卫星通信中的多路径干扰和时变失真等问题。在实际应用中,需选择合适的技术和算法对不同的信号调制类型进行优化和调整,进一步提高系统的性能和稳定性。文中为宽带卫星通信中的信号调制识别和信道均衡技术提供了参考,并为未来卫星通信系统的发展提供了支持。     

卫星通信网络终端多载波干扰分析

近些年来,在我国经济快速发展的有力推动下,科技力量以及信息技术发展也是突飞猛进,其中尤以卫星通信网络为典型。在整个卫星通信网络系统当中,任一终端都能够实现对不同频段的载波信息进行接收。然而由于通信卫星和终端系统之间的相对运动,造成频率出现一定偏移,进而造成载波出现一定干扰。因此,文章正是基于这个角度,重点就卫星通信网络终端多载波干扰展开探讨,以供参考。     

基于同道干扰统计模型的无线自组织网络性能分析

基于无线自组织网络中节点的分布特征,结合对同道干扰的定量分析结果,给出了 一种适用于无线自组织网络的同道干扰统计模型。分析了无线自组织网络中同道干扰的统计 特性,并基于建立的同道干扰模型研究了无线自组织网络的系统性能。结果表明,同道干扰 对无线自组织网络的系统性能有严重影响,需建立合理的干扰模型才能更准确地评估自组织 网络的系统性能。     

基于MobileNet和NAM注意力机制的轻量级OpenPose网络

针对深度学习下的人体姿态识别技术,分析了OpenPose网络参数量巨大,容易产生梯度消失;MobileNet网络识别准确率较低等问题,研究了人体姿态识别、卷积神经网络以及注意力机制的工作原理。分析了OpenPose模型、ShuffleNet模型、GhostNet模型、MobileNet模型的卷积神经网络模型和NAM注意力机制的各自特点,提出了基于MobileNet网络和NAM注意力机制的人体姿态识别技术,解决了OpenPose网络参数量大运算和速率较低,MobileNet网络识别准确率较低等问题,通过实验对比识别准确率和识别速率,结果表明添加了NAM注意力机制的MobileNet网络具有识别精度高,识别速率快等特点。     

卫星通信侦察中的调制识别与参数估计技术

在卫星通信链路的侦察和对抗系统中,低信噪比下的调制识别和参数估计是情报获取并进一步实施干扰的前提和基础。对卫星通信链路侦察中常用的调制识别和参数估计方法进行了总结,并对各种方法的特点进行了分析,同时提出了需要进一步研究的问题。     

结合双层多头自注意力和BiLSTM-CRF的军事武器实体识别

军事武器实体识别是军事领域本体构建的一项重要任务,基于深度学习方法实现自动军事武器实体识别能够提升军事情报信息检索的效率.为提升军事武器实体识别的精确率,面向网络公开非结构化军事新闻数据,提出了一种结合双层多头自注意力机制和BiLSTM-CRF模型的武器实体识别方法.在BiLSTM-CRF模型的基础上,采用双层自注意力...     

融合注意力机制的雷达欺骗干扰域适应识别方法

针对目前雷达欺骗干扰识别中常规特征识别方法应用受限和训练高性能深度学习模型需要的大量标注样本难以高效获取的问题,该文提出一种基于对抗域适应网络的雷达欺骗干扰识别方法,以改善标签限制;并融合注意力机制残差模块进一步提升识别精度。首先,对雷达接收信号进行时频变换后,应用基于对抗网络思想的域适应技术实现从标注源域样本到未标注目标域样本的迁移识别。其次,通过所设计的空间通道注意力残差模块使网络训练聚焦于时频图全局空间特征和高响应通道,以忽略时频图像中可迁移性低的区域抑制负迁移的产生。在不同源域与目标域雷达欺骗干扰数据集上的实验结果表明了该方法的可行性和有效性。     

空地协同场景下通信干扰智能识别方法

针对现有通信干扰智能识别方法在小样本条件下识别精度低、网络模型欠拟合的问题,并形成通信干扰识别的空中与地面布设能力,该文提出一种空地协同场景下基于孪生网络的通信干扰智能识别方法。首先在空中无人机与地面设备之间构建空地协同的通信干扰认知架构,并通过提取所接收的通信干扰信号的时频图、分数阶傅里叶变换和星座图,对通信干扰信号进行智能表征,以作为网络的输入。然后搭建基于密集连接网络的网络结构,并设计双输入权值共享的孪生网络。最后,利用随机样本对孪生网络进行训练,并通过孪生单边网络构建基准通信干扰类型特征库进而实现通信干扰的智能识别。该方法通过度量两个样本之间的特征距离来判断样本的相似性,并通过相似度度量扩大了训练样本数量并训练了孪生网络模型。仿真结果表明,所提方法不但在较小数据集的条件下可有效地实现通信干扰的智能识别,而且相比现有的智能识别方法,所提方法的识别性能显著提升。     

Optimization of Cognitive Radio System Using Simulated Annealing

Cognitive radio (CR) technology has introduced a revolution in wireless communication network and it is capable to operate in a continuously varying radio frequency environment that depends on multiple parameters. In this paper, optimization of CR system has been achieved using simulated annealing (SA) Technique. SA is a stochastic global optimization technique that exploits an analogy between the way in which a metal cools and freezes into a minimum energy crystalline structure. SA has been used to meet the quality of service (QoS) that is defined by the user in terms of minimum transmit power, minimum bit error rate, maximum throughput, minimum interference and maximum spectral efficiency. The results obtained by SA are compared with the genetic algorithm (GA) results for the various QoS parameters and it has been observed that SA is outperforming GA in CR system optimization.     

基于谱图和神经网络的通信干扰模式识别方法

对通信系统受干扰的模式进行分析和模式识别,可以指导通信系统进行相应的自适应参数调整,以具有更强、更有针对性的抗干扰能力。研究宽带通信系统,利用多隐藏层的神经网络可以解决任意形式分类问题的特性,构建一种基于功率谱谱图和双隐藏层神经网络的通信干扰模式识别方法,可以对5种常见的通信干扰进行快速的模式识别。仿真结果表明,该通信干扰模式识别方法对干扰模式在不同的干噪比情况下能获得99.6%以上的平均识别概率,对除梳状谱干扰外的各种干扰模式识别准确率均达到99.7%以上,梳状谱干扰识别准确率达到98.4%以上。该方法具备较稳定的识别能力,可应用于干扰感知的流程中。     

Autonomous Schemes for Inter-cell Interference Coordination in the Downlink of LTE Systems

Inter-cell interference (ICI) is one of the key challenges that limit the performance of Long Term Evolution and Long Term Evolution Advanced cellular systems. One approach to deal with ICI is through interference avoidance. Unlike static avoidance techniques where a-priori frequency planning and/or explicit inter-cell coordination is used, dynamic avoidance techniques rely on adapting its frequency planning and allocation based on the current state of the network. However, this improvement in performance comes with the cost of an increased complexity due to the coordination and alignments needed among the base stations (eNB) to manage and allocate channels among the users. Accordingly, autonomous ICI coordination techniques are receiving much interest among the various interference avoidance techniques. In this paper, we propose an autonomous self-adaptive scheme (SA, for short) for radio resource management and interference coordination. We then extend the proposed scheme to become self-adaptive power-aware (SAPA) in order to optimize and reduce the transmission power of the eNBs. A key feature in the proposed schemes is that all computations are independent of the number of users and cells in the network. This allows the proposed schemes to adapt to networks of any size and with an arbitrary number of users. Extensive simulation confirms that the proposed SA scheme ensures efficient frequency reuse patterns that lead to significant performance improvements in the throughput of the edge users without affecting other users. Moreover, the SAPA scheme achieves significant improvement in the power efficiency, while maintaining the throughput enhancements achieved by the SA scheme for both center and edge users.     

基于奇异值分解与神经网络的干扰识别

无线通信中的抗干扰技术对通信的稳定性和安全性都具有重要意义,干扰识别作为抗干扰技术的重要环节一直是研究的热点。该文提出一种基于奇异值分解与神经网络的干扰识别方法,该方法只计算信号矩阵的奇异值即完成特征提取,与传统方法相比节省了多个谱特性的计算量。仿真结果表明:基于奇异值分解与神经网络的干扰识别方法与传统方法相比在干信比为0 dB左右的条件下识别准确率有10%～25%的提高。     

On the performance of two‐user full‐duplex network‐coded cooperation

In this paper, we propose two schemes based on a full‐duplex network‐coded cooperative communication (FD‐NCC) strategy, namely, full‐duplex dynamic network coding (FD‐DNC) and full‐duplex generalized dynamic network coding (FD‐GDNC). The use of full‐duplex communication aims at improving the spectrum efficiency of a two‐user network where the users cooperatively transmit their independent information to a common destination. In the proposed FD‐NCC schemes, the self‐interference imposed by full‐duplexing is modeled as a fading channel, whose harmful effect can be partially mitigated by interference cancellation techniques. Nevertheless, our results show that, even in the presence of self‐interference, the proposed FD‐NCC schemes can outperform (in terms of outage probability) the equivalent half‐duplex network‐coded cooperative (HD‐NCC) schemes, as well as traditional cooperation techniques. Moreover, the ?‐outage capacity, that is, the maximum information rate achieved by the users given a target outage probability, is evaluated. Finally, we examine the use of multiple antennas at the destination node, which increases the advantage of the FD‐NCC (in terms of the diversity‐multiplexing trade‐off and ?‐outage capacity).     

基于深度残差网络的特定协议信号识别

针对短波信道下信号截获质量差,信道环境复杂以及单一特征识别率低等问题,提出了基于深度残差网络的信号特征自动提取算法,设计了一种具有自适应学习能力的短波特定通信协议识别模型.通过对具有特殊结构的协议信号的时频视觉差异进行理论推导,将信号的时频能量转换成灰度图像,并用于对所构建的深度残差网络进行训练.该方法克服了传统方法对信号质量要求高、先验信息需求多等缺陷,可直接对中频接收信号进行处理,适合实际工程应用.实验表明,当深度残差网络达到稳态时,识别准确率高,在低信噪比、多径衰落、多普勒频偏以及信号被强干扰所遮挡的情况下,依旧能准确识别协议类别.     

Proposal of a Fixed Communication System Sharing the Bandwidth of an Existing Personal Communication Network

This paper deals with the proposal and performance evaluation of a fixed microwave communication (FMC) system that shares the same bandwidth with a wireless personal communication network. In particular, the focus here is on the case of an existing universal mobile telecommunication system (UMTS) network for which an FMC system is considered to provide a wireless connection between a remote base station not having the possibility of a wired connection to the core network and a base station having this possibility (a core network access point). This facility is of special interest in a dense urban environment or whenever a wired connection is prohibitive for service providers due to its implementation cost or lack of frequency spectrum. The mutual interference effects between the FMC systems and the existing UMTSs are investigated by focusing on typical application scenarios. A receiving scheme where interference signals are first detected and then canceled from the other received signals is proposed. The performance of the two interfering systems has been evaluated in terms of bit error rates (BERs) both by means of an analytical approach and computer simulations. The obtained results clearly demonstrate the effectiveness of the proposed method, in particular, by allowing efficient bandwidth sharing between the FMC system and the existing UMTS network.     

一种基于深度置信网络的通信辐射源个体识别方法

针对复杂电磁环境下通信辐射源个体识别问题，提出了一种小样本条件下基于深度置信网络的通信辐射源个体识别方法.首先分析通信辐射源信号频带内互调干扰信号的幅度和相位特性，建立基于互调干扰信号的通信辐射源个体特征；然后对辐射源信号进行预处理得到通信辐射源信号的矩形积分双谱，再采取对比散度的方法，利用高阶谱自底向上训练每个受限玻尔兹曼机，通过多次迭代得到合适的权重、隐藏层的偏差和可见层的偏差，从而提取出辐射信号的互调干扰信号特征；最后使用softmax分类器对训练模型进行微调，获得面向通信辐射源细微特征识别的深度学习网络.通过计算机的仿真得到了超过80%的识别率，进一步验证了该方法的有效性.     

安全高效的两方协同ECDSA签名方案

为了解决签名私钥易泄露和签名权利过度集中的问题,针对基于区块链技术的网络交易系统,提出了一种安全高效的两方协同ECDSA签名方案。通过预计算一次一密的Beaver三元组,进而利用基于Beaver三元组的安全两方乘法技术,有效避免使用计算繁重的同态加密和通信开销较大的不经意传输等操作,实现高效的两方协同ECDSA签名,保证2个签名参与方在不重构完整签名私钥的情况下输出合法的ECDSA签名。方案的安全性在通用可组合框架中的混合模型下被证明。理论分析与实验结果表明,与现有的2种两方协同ECDSA签名方案相比,所提方案在协同签名运行效率和带宽要求方面均具有明显优势。