Evaluation of Network Stack Optimization Techniques for Wireless Sensor Networks

We present a network stack implementation for a wireless sensor platform based on a byte-level radio. The network stack provides error-correction code, multi-channel capability and reliable communication for a high packet reception rate as well as a basic packet-level communication interface. In outdoor tests, the packet reception rate is close to 100% within 800 ft and is reasonably good up to 1100 ft. This is made possible by using error correction code and a reliable transport layer. Our implementation also allows us to choose a fre-quency among multiple channels. By using multiple frequencies as well as a reliable transport layer, we can achieve a high packet reception rate by paying additional retransmission time when collisions increase with additional sensor nodes.

Dynamic prediction of landslide displacement using singular spectrum analysis and stack long short-term memory network

An accurate landslide displacement prediction is an important part of landslide warning system. Aiming at the dynamic characteristics of landslide evolution and the shortcomings of traditional static prediction models, this paper proposes a dynamic prediction model of landslide displacement based on singular spectrum analysis(SSA) and stack long short-term memory(SLSTM) network. The SSA is used to decompose the landslide accumulated displacement time series data into trend term and periodic term displacement subsequences. A cubic polynomial function is used to predict the trend term displacement subsequence, and the SLSTM neural network is used to predict the periodic term displacement subsequence. At the same time, the Bayesian optimization algorithm is used to determine that the SLSTM network input sequence length is 12 and the number of hidden layer nodes is 18. The SLSTM network is updated by adding predicted values to the training set to achieve dynamic displacement prediction. Finally, the accumulated landslide displacement is obtained by superimposing the predicted value of each displacement subsequence. The proposed model was verified on the Xintan landslide in Hubei Province, China. The results show that when predicting the displacement of the periodic term, the SLSTM network has higher prediction accuracy than the support vector machine(SVM) and auto regressive integrated moving average(ARIMA). The mean relative error(MRE) is reduced by 4.099% and 3.548% respectively, while the root mean square error(RMSE) is reduced by 5.830 mm and 3.854 mm respectively. It is concluded that the SLSTM network model can better simulate the dynamic characteristics of landslides.

A Dynamic Protocol Stack Structure for Diversified QoS Requirements in Ad Hoc Network

A dynamic protocol stack(DPS) for ad hoc networks, together with a protocol stack construction scheme that is modeled as a multiconstrained knapsack problem is proposed. Compared to the traditional static protocol stack, DPS operates in a dynamic and adaptive manner and is scalable to network condition changes. In addition, a protocol construction algorithm is proposed to dynamically construct of the protocol stack each network node. Simulation results show that, the processing and forwarding performance of our scheme is close to 1 Gb/s, and the performance of our algorithm is close to that of the classical algorithms with much lower complexity.

Self-organization of Reconfigurable Protocol Stack for Networked Control Systems

In networked control systems （NCS）,the control performance depends on not only the control algorithm but also the communication protocol stack.The performance degradation introduced by the heterogeneous and dynamic communication environment has intensified the need for the reconfigurable protocol stack.In this paper,a novel architecture for the reconfigurable protocol stack is proposed,which is a unified specification of the protocol components and service interfaces supporting both static and dynamic reconfiguration for existing industrial communication standards.Within the architecture,a triple-level self-organization structure is designed to manage the dynamic reconfiguration procedure based on information exchanges inside and outside the protocol stack.Especially,the protocol stack can be self-adaptive to various environment and system requirements through the reconfiguration of working mode,routing and scheduling table.Finally,the study on the protocol of dynamic address management is conducted for the system of controller area network （CAN）.The results show the efficiency of our self-organizing architecture for the implementation of a reconfigurable protocol stack.

Stacked Attention Networks for Referring Expressions Comprehension

Referring expressions comprehension is the task of locating the image region described by a natural language expression,which refer to the properties of the region or the relationships with other regions.Most previous work handles this problem by selecting the most relevant regions from a set of candidate regions,when there are many candidate regions in the set these methods are inefficient.Inspired by recent success of image captioning by using deep learning methods,in this paper we proposed a framework to understand the referring expressions by multiple steps of reasoning.We present a model for referring expressions comprehension by selecting the most relevant region directly from the image.The core of our model is a recurrent attention network which can be seen as an extension of Memory Network.The proposed model capable of improving the results by multiple computational hops.We evaluate the proposed model on two referring expression datasets:Visual Genome and Flickr30k Entities.The experimental results demonstrate that the proposed model outperform previous state-of-the-art methods both in accuracy and efficiency.We also conduct an ablation experiment to show that the performance of the model is not getting better with the increase of the attention layers.

ConvNeXt网络及Stacked BiLSTM-Self-Attention在轴承剩余寿命预测中的应用

在滚动轴承剩余使用寿命预测方面,采用传统方法时存在鲁棒性差、精度低等各种问题。近些年来深度学习的发展为解决这些问题提供了新的思路。为了进一步提高对轴承寿命的预测精度,提出了一种基于ConvNeXt网络、堆叠双向长短时记忆网络(SBiLSTM)和自注意力机制(Self-Attention)的滚动轴承寿命预测方法。首先,采用连续小波变换(CWT)构造了振动信号的时频图,以更好地捕捉信号的时域和频域特征;然后,将得到的时频图输入到构建的ConvNeXt网络中,通过卷积、池化和层归一化等操作,对时频图的关键特征进行了提取;最后,将提取后的特征输入到SBiLSTM-Self-Attention模块中,进一步提取了时序信息和特征权重分配数据,利用PHM2012挑战数据集进行了验证,通过实验分析了该方法的均方根误差(RMSE)和平均绝对误差(MAE)。研究结果表明:相较于现有技术方法,该方法的平均RMSE为0.031;与其他三种方法,即卷积神经网络(CNN)、深度残差双向门控循环单元(DRN-BiGRU)和深度卷积自注意力双向门控循环单元(DCNN-Self-Attention-BiGRU)相比,其平均RMSE值分别下降了79%、74%和55%,MAE值分别下降了78%、73%和53%,说明该方法在滚动轴承剩余寿命预测中有较好的性能。

基于Stacking融合的LSTM-SA-RBF短期负荷预测

为了解决单个神经网络预测的局限性和时间序列的波动性,提出了一种奇异谱分析(singular spectrum analysis,SSA)和Stacking框架相结合的短期负荷预测方法。利用随机森林筛选出与历史负荷相关性强烈的特征因素,采用SSA为负荷数据降噪,简化模型计算过程;基于Stacking框架,结合长短期记忆(long and short-term memory,LSTM)-自注意力机制(self-attention mechanism,SA)、径向基(radial base functions,RBF)神经网络和线性回归方法集成新的组合模型,同时利用交叉验证方法避免模型过拟合;选取PJM和澳大利亚电力负荷数据集进行验证。仿真结果表明,与其他模型比较,所提模型预测精度高。

Nonlinear modeling based on RBF neural networks identification and adaptive fuzzy control of DMFC stack

The temperature models of anode and cathode of direct methanol fuel cell （DMFC） stack were established by using radial basis function （RBF） neural networks identification technique to deal with the modeling and control problem of DMFC stack. An adaptive fuzzy neural networks temperature controller was designed based on the identification models established, and parameters of the controller were regulated by novel back propagation （BP） algorithm. Simulation results show that the RBF neural networks identification modeling method is correct, effective and the models established have good accuracy. Moreover, performance of the adaptive fuzzy neural networks temperature controller designed is superior.

基于Stacking集成学习的枣树智能灌溉系统设计与试验

南疆降雨量少,气候干燥,农业用水紧张,水资源节约尤为重要,针对此问题设计一套智能灌溉系统。系统使用阿里云服务器作为上位机,树莓派作为下位机,并搭建相应的操作页面。根据Penman-Monteith公式中需要的气象数据、过去7天需水量以及前1天气象数据为输入向量,作物需水量为输出向量,构建基于随机森林、BP神经网络与岭回归的Stacking集成学习预测模型。结果表明Stacking集成学习预测模型拟合系数R 2为0.973,且MAE、RMSE、MAPE三类误差更小,Stacking集成学习预测模型预测效果更强。灌溉试验中自动灌溉决策正确,系统运行稳定,为新疆地区农业提高水资源利用问题提供思路。

基于F-Stack的高性能ICN网关设计与实现

部署信息中心网络(ICN)协议转换网关是实现ICN与现有IP网络兼容演进的一种方式。为了提升网关在数据流量的处理与转发方面的效率,避免网关的数据处理瓶颈影响ICN提供的性能增益,本文提出一种基于F-Stack开发框架的高性能ICN网关设计方法。该网关系统利用DPDK用户态协议栈快速处理大量TCP连接与流量,并结合DPDK的共享内存与无锁环形队列机制实现进程间通信,不仅具备良好的性能,而且可以降低模块间的耦合性。实验结果表明,在8核CPU资源配置下,本文方法具有良好的性能优势:数据传输速率可达75%网卡线速,同时支持87万个以上的并发连接,且平均处理时延在45μs以下。

基于残差注意力自适应去噪网络和Stacking集成学习的局部放电故障诊断

针对传统局部放电(Partial Discharge,PD)故障诊断方法在处理复杂含噪PD信号存在局限性并依赖于人工去噪和专家经验,难以学习到PD特征多样化表达等问题,分别提出残差注意力自适应去噪网络(Residual Attention Adaptive Denoising Network,RAADNet)和基于Stacking集成学习的PD故障诊断模型。RAADNet基于残差网络结构设计,通过集成CAM注意力机制和软阈值函数实现自适应去噪;Stacking集成模型的基学习器分别由基于卷积神经网络的RAADNet、基于多头自注意力机制的Transformer以及基于Boosting集成策略的XGBoost多个差异化模型共同构建构成。实验结果表明,提出的RAADNet优于其他先进方法,识别准确率达到93.99%,Stacking集成模型则通过学习多样化特征表达,进一步提高模型性能,达到96.79%识别准确率。

基于Z-Stack协议栈的ZigBee网络节能算法的研究

ZigBee技术是实现无线传感器网络的重要技术手段之一,其中,Z-Stack协议栈是TI公司推出的一种实现ZigBee技术的软件架构,已被广泛使用。以CC2530为硬件平台,Z-Stack协议栈为软件平台,通过对Z-Stack协议栈的分析研究,指出了现有的Z-Stack协议栈在实际应用条件下存在的功耗问题,提出了相应的节能算法:针对终端节点的功率优化调节算法和针对路由器节点的低功耗算法,并通过实验进行验证。实验结果表明,与原始的Z-Stack协议相比,本文提出的基于Z-Stack协议栈的节能算法可以有效地减少节点在工作过程中的能耗,从而提高网络的整体寿命。

基于Z-Stack协议栈的WSN能量管理策略

在分析基于ZigBee协议的Z-Stack1.4.3协议栈基础上,实现无线传感器网络(WSN)数据的自动周期获取,通过设置协议栈中休眠条件,不仅能最大限度地降低功耗,而且可以保证网络的同步性。构建一个簇树结构的低功耗ZigBee网络,设计具有网络数据、拓扑显示等功能的监控管理系统。应用结果表明,该系统能在低功耗下稳定地工作,完成环境数据监测。

基于Z-Stack协议栈的船舶机舱环境监测系统设计

为提高船舶机舱的安全性,针对传统有线监测方式存在布线困难、成本高和扩展性差的缺点,设计了一套基于Z-Stack协议栈的监测系统。该系统综合了Zigbee无线网络自行愈合、组网灵活和超低功耗的优点,以CC2530射频模块为基础组成网络节点,由传感器节点把测得的温湿度值、CO_(2)浓度和空气流速等参数经路由器或直接传递到协调器,最终传至上位机监控界面来实时显示、记录数值。试验结果表明,该系统具有功耗低、稳定性高、可开发性强等特点,满足机舱里监测实时性和数据精度的要求,同时对提高船舶自动化水平具有一定的参考和借鉴意义。

基于利用BP神经网络进行Stacking模型融合算法的电力非节假日负荷预测研究

短期负荷预测尤其是非节假日负荷预测对提升电力系统整体调度、支撑电网运营工作起着十分关键的作用。目前针对非节假日负荷预测的理论、方法和应用层出不穷,但是预测精度和使用范围都受到一定限制,并且经济发展对短期负荷预测的精度提出越来越高的要求,传统的机器学习算法已经难以满足人们的需求。为了提高负荷预测的精度,本文提出了利用BP神经网络进行Stacking模型融合算法,它是基于集成学习的思想,首先挑选五种预测精度较高的单模型,然后利用Stacking模型融合方法将其集成为预测精度更高的综合模型。本文采用此算法预测某省2018年非节假日负荷,结果表明该算法可以有效提高预测精度。

基于Z-Stack的ZigBee协议栈组网过程研究

在ZigBee协议栈Z-Stack基础上,重点分析了Z-Stack协议栈的结构和组网类型,通过详细研究TI公司的Z-Stack协议栈核心的实现过程,从代码级别对ZigBee协议栈组网过程进行剖析和节点组网过程中的重要函数调用进行详细分析,对实际工程项目开发有一定的指导作用。

基于Stacking多GRU模型的风电场短期功率预测

为提高风电场短期功率预测的准确度,在深度学习的基础上提出利用Stacking算法集成融合多个GRU(Gated Recurrent Unit)模型的风电场短期功率预测的方法。该方法首先搭建3个多层GRU神经网络模型建立第1级模型,深度提取高维的时序特征关系,通过第1级模型的预测结果构建训练集,然后利用新生成的训练集训练第2级GRU模型,第2级的GRU模型采用单层结构,能发现并且纠正第1级模型中的预测误差,提升整体的预测结果。最终得到两级模型嵌入的Stacking融合模型。以宁夏太阳山风电场历史数据为例对该模型的准确性进行验证。实验结果表明,通过Stacking算法融合的GRU模型相比其他算法预测平均绝对百分比误差提高了0.63,总体预测效果较为理想,预测准确度提升明显。

0.13 μm CMOS Stacked-FET两级功率放大器设计

基于TSMC 0.13μm CMOS工艺设计了一款适用于无线传感网络、工作频率为300~400 MHz的两级功率放大器。功率放大器驱动级采用共源共栅结构,输出级采用了3-stack FET结构,采用线性化技术改进传统偏置电路,提高了功率放大器线性度。电源电压为3.6 V,芯片面积为0.31 mm×0.35 mm。利用Cadence Spectre RF软件工具对所设计的功率放大器电路进行仿真,结果表明,工作频率为350 MHz时,功率放大器的饱和输出功率为24.2 d Bm,最大功率附加效率为52.5%,小信号增益达到38.15 d B。在300~400 MHz频带内功率放大器的饱和输出功率大于23.9 d Bm,1 d B压缩点输出功率大于22.9 d Bm,最大功率附加效率大于47%,小信号增益大于37 d B,增益平坦度小于±0.7 d B。

基于Z-Stack协议的网络地址分配机制研究

当代无线传感器网络得到了快速的发展,无线传感器网络协议标准——Zigbee 2007也日渐规范,与Zigbee 2007标准完全兼容的Z-Stack协议,因为在构建网络方面的强大优势,其在各行业的应用前景十分广泛。基于Z-Stack分布式地址分派机制,分析了其网络地址分配原理,搭建了一种通用的网络信息传递架构,通过实验验证了分布式地址分配机制的正确性,为其它Z-Stack组网应用提供参考。

适合酒店应用的Z-Stack协议的研究与实现

应用先进的信息化、智能化技术来改变现有的酒店管理和经营方式,可以有效地解决现有酒店管理效率低,人工成本高等问题,而这些技术的应用都需要高效、安全的网络来支撑。本文根据Z-Stack协议的特点,以及在酒店对所使用的无线网络的特殊性及要求,在原有Z-Stack协议的基础上设计了专门用于酒店自动化的ZStack协议。论文从网络结构、固定地址处理、安全与加密以及低功耗等几个方面对所设计的协议进行了阐述,最后给出了所设计的Z-Stack协议在酒店门锁管理系统中的一个具体应用。本文提出的专用于酒店管理的ZStack协议不仅数据传输可靠性高、安全性高,并且成本低、功耗低,能有效提高酒店的经济效益,提升酒店的服务水平。