基于移动蜂窝网的机器学习室外指纹定位方案

基于移动蜂窝网络技术的定位方案是提供网络优化、紧急救援、公安巡警和位置服务等应用的重要技术途径之一。传统的基于小区基站位置信息的定位方案定位精度低、定位误差大,无法满足某些定位应用需求。基于指纹定位的方案能够在基于小区粗定位方案基础上大幅度提升定位精度、节约计算成本、增强适用性,成为定位研究的热点。针对室外指纹定位的业务需求,深入研究分析了两种基于机器学习的栅格化和非栅格化室外指纹定位方案。通过参数加权、数据拟合等方法对于大规模指纹数据进行了清洗,提高数据源的有效性。通过划定研究区域、栅格化、构建指纹数据库、训练模型、修正模型、非栅格化、粗定位耦合、匹配参数、训练参数等子模块的实现,分析和优化了算法的运行效率和定位精度,确定了影响算法性能的关键指标。进而结合仿真结果,分析了两种基于指纹的定位方案的性能。最后介绍了基于机器学习的指纹定位方案在实际应用中的典型场景。     

一种基于量子细胞自动机的三维量子细胞神经网络

以量子细胞自动机为神经元,提出了一种三维的量子细胞神经网络结构;该量子细胞网络包含上下两层的量子细胞自动机阵列,并引入了A模板、B模板以及阈值的概念.以量子细胞自动机的极化率为像素值,通过选择不同的模板、阈值等参数.使得量子细胞神经网络实现了"与"、"或"、"非"操作以及边缘提取等图像处理功能,并利用MATLAB进行了仿真验证,数值仿真结果验证了其在图像处理上的有效性.与传统的细胞神经网络相比,量子细胞神经网络易于实现超大规模且具有超低功耗、超高集成度等优点.     

一种基于随机映射的网络状态评估方法

近年来,机器学习技术在网络管理领域得到了广泛使用。然而由于通信网络日益复杂,网络中的非线性和不确定因素使得机器学习变得十分困难。为了提升机器学习的效果,提出了一种采用随机映射的人工神经网络方案,其特点是引入机器学习和网络拓扑的随机性,使得神经网络对学习目标具有更大的适应性,并实现更快、更精确的收敛。相关成果已经在中国移动通信集团山西有限公司（以下简称山西移动）的实际网络中得到了应用并取得较好的效果。     

Classification with ensembles and case study on functional magnetic resonance imaging

The ensemble is a technique that strategically combines basic models to achieve better accuracy rates. Diversity, combination methods, and selection topology are the main factors determining ensemble performance. Consequently, it is a challenging task to design an efficient ensemble scheme. Even though numerous paradigms have been proposed to classify ensemble schemes, there is still much room for improvement. This paper proposes a general framework for creating ensembles in the context of classification. Specifically, the ensemble framework consists of four stages: objectives, data preparing, model training, and model testing. It is comprehensive to design diverse ensembles. The proposed ensemble approach can be used for a wide variety of machine learning tasks. We validate our approach on real-world datasets. The experimental results show the efficiency of the proposed approach.     

Estimating rainfall using machine learning strategies based on weather radar data

The precision of forecasting rainfall is vital owing to current world climate change. As deterministic weather forecasting models are usually time consuming, it becomes challenging to efficiently use this large volume of data in hand. Machine learning methods are already proven to be good replacement for traditional deterministic approaches in weather prediction. This paper presents an approach using recurrent neural networks (RNN) and long short term memory (LSTM) techniques to improve the rainfall forecast performance. This will be compared with the random forest classifier and XGBoost as well. The goal is to predict a set of hourly rainfall levels from sequences of weather radar measurements. Python libraries are utilized to forecast the time series data. The training set comprises of data from first 20 days of every month and the inference set data from the continuing days. This makes sure that both train and inference sets are more or less independent. The idea resides in implementing an end‐to‐end learning framework.     

Deep reinforcement learning based resource allocation algorithm in cellular networks

In order to solve multi-objective optimization problem,a resource allocation algorithm based on deep reinforcement learning in cellular networks was proposed.Firstly,deep neural network (DNN) was built to optimize the transmission rate of cellular system and to complete the forward transmission process of the algorithm.Then,the Q-learning mechanism was utilized to construct the error function,which used energy efficiency as the rewards.The gradient descent method was used to train the weights of DNN,and the reverse training process of the algorithm was completed.The simulation results show that the proposed algorithm can determine optimization extent of optimal resource allocation scheme with rapid convergence ability,it is obviously superior to the other algorithms in terms of transmission rate and system energy consumption optimization.     

采用梯度提升树的战机型号快速识别方法

针对战机型号快速准确识别问题,提出一种利用梯度提升树的战机型号快速识别方法.以多传感器融合的战机航迹解译信息为数据基础,通过分析航迹数据特征,构建战机航迹数据特征工程,利用boosting集成学习思想,训练基于梯度提升决策树的战机型号分类器,可准确识别每个航迹点对应的战机型号.实测数据实验结果表明,所提模型识别准确率达...     

An efficient bandwidth allocation scheme for hierarchical cellular networks with energy harvesting: an actor-critic approach

In this paper, we propose an efficient bandwidth allocation strategy for multiclass services in hierarchical cellular networks that consist of an operation controller, several small-cell base stations (SBSs), and a number of mobile users. Each SBS is equipped with a finite-capacity battery that is regularly recharged by a solar harvester. We aims to find the optimal bandwidth allocation policy in order to enhance the network performance in terms of user satisfaction and energy efficiency under energy harvesting and bandwidth sharing constraints. Since the arrivals of harvested energy and traffic requests are unknown due to users’ mobility and stochastic request generation, it is necessary to design a learning framework for the controller in order to predict these dynamics through interaction with the environment. For this purpose, we first formulate the resource allocation problem as the framework of a Markov decision process, and then, we employ an actor-critic algorithm to find the optimal policy under which the controller can effectively allocate the limited bandwidth to the SBSs for their data transmissions. We evaluate the performance of the proposed scheme through comprehensive simulations with different settings, and show that the proposed bandwidth allocation scheme can enhance the network’s performance in the long run.     

An efficient resource allocation scheme in a dense RFID network based on cellular learning automata

Radio‐frequency identification (RFID) is a wireless communication technology. Radio frequencies can cause interference in a dense RFID system, thus decreasing efficiency. In recent years, many protocols have been proposed to reduce reader collisions based on multiple‐access techniques. The main weakness of Time Division Multiple Access (TDMA)‐based schemes is the random selection of resources. Additionally, they do not consider the distance between the interfering readers. Therefore, the likelihood of interference in an RFID system will be increased. To address this problem, we propose a new scheme for allocating resources to readers using a learning technique. The proposed scheme takes into account the distance between interfering readers, and these readers acquire the necessary knowledge to select new resources based on the results of the previous selection of neighboring readers using cellular learning automata. This approach leads to reduced interference in an RFID system. The proposed scheme is fully distributed and operates without hardware redundancy. In this scheme, the readers select new resources without exchanging information with each other. The simulation results show that the percentage of kicked readers decreased by more than 20%, and the proposed scheme also provides higher throughput than do state‐of‐the‐art schemes for dense reader environments and leads to further recognition of tags.     

基于优化决策树的时延敏感流智能感知调度

目前流量调度策略无法做到智能按需化,尤其对于网络突发故障造成的拥塞以及高价值业务的护航场景,无法按需保障时延敏感的业务体验。通过分析研究不同网络业务流量时延敏感性属性需求,探索挖掘不同网络业务流量的行为特征与其时延敏感性需求之间的内在关联关系。然后利用AI技术对这种内在的关联关系进行学习,构建其映射关系,实现了时延敏感流智能感知调度。同时,考虑AI模型的可解释性及可部署性实际问题,采用强化学习剪枝优化可解释性决策树模型,提高模型的鲁棒性同时使模型更轻量化,易于设备部署实现。通过真实网络流量实验,强化学习优化后的决策树模型在单次推理情形下感知正确率提高1.75%,推理速度提升约30%;同时,实验也证明了使用局部微观统计特征多次推理有助于提高模型感知正确率。在所有实验中,强化学习优化的决策树模型规模缩小了60.0%～87.2%,并且Saras比Q-learning具有更好的优化表现。     

基于大系统方法的上行CDMA蜂窝系统性能分析

结合大尺度和小尺度衰落,采用大系统分析方法对CDMA蜂窝系统上行链路的性能进行分析。首先分析了小区内多址干扰和小区间干扰的情况;其次推导出了通过匹配滤波器后的信干噪比值,该值只与小区负载、用户功率、信道增益有关;最后分析了系统的误码率、吞吐量和频谱效率,并通过仿真验证了分析的准确性。仿真结果发现,随着小区负载的增大,系统的误码率增大,吞吐量下降,但频谱效率有所提高。     

基于决策树算法的无线参数优化方法

本文针对大规模无线参数历史数据,以自学习的数据挖掘方法建立决策树智能分析模型,并对现网数据实现联机分析处理,是一种智能的主动发现无线参数配置问题的算法模型。     

基于认知自选择决策树的垂直切换方法研究

在异构无线网络环境下高速移动的终端用户需要随时随地接入网络,然而,在垂直切换过程中,由于终端用户对网络选择的偏好不同以及切换之后网络连接时间和状态难以得到保障,无法满足用户的切换需求。因此,提出了一种支持终端用户高速移动的基于认知自选择决策树的垂直切换方法。该方法首先根据网络属性和终端运动趋势建立相应的切换概率分布,然后根据用户偏好选择相应的决策树进行决策,最后通过对切换后终端的业务类型和运动状态进行反馈分析,提出了反馈认知决策方法。仿真结果表明,所提的方法不仅能够保证高速移动终端在异构无线网络下的切换质量,而且减少了不必要的切换,保证了网络的及时更新。     

基于移动蜂窝网的定位技术探讨

移动蜂窝网本身就具有一定的定位服务,在5G新基建的支撑下,定位服务进一步增强,定位精度也大幅提升。在5G新基建的移动互联网时代,定位服务获得了前所未有的关注与青睐。结合多年在定位服务领域的研究,文章对基于移动蜂窝网的定位服务进行了综述,分析了各类定位技术的优势与不足,总结了定位服务的指标体系,通过行业的用例对定位服务进行了说明。文章对移动蜂窝网的定位服务进行梳理,旨在为研发与管理人员提供支撑。     

一种应用于GNSS/蜂窝网混合定位接收机的多标准频率合成器

A fully integrated hybrid integer/fractional frequency synthesizer is presented.With a single multiband voltage-controlled-oscillator(VCO),the frequency synthesizer can support GPS,Galileo,Compass and TDSCDMA standards.Design is carefully performed to trade off power,die area and phase noise performance.By reconfiguring between the integer mode and fractional mode,different frequency resolution requirements and a constant loop bandwidth for each standard can be achieved simultaneously.Moreover,a long sequence length,reduced hardware complexity multi-stage-noise-shaping(MASH).-.modulator is employed to reduce fractional spur in the fractional mode.Fabricated in a 0.18 m CMOS technology,the frequency synthesizer occupies an active area of 1.48 mm2 and draws a current of 13.4-16.2 mA from a 1.8 V power supply.The measured phase noise is lower than-80 dBc/Hz at 100 kHz offset and-113 to-124 dBc/Hz at 1 MHz offset respectively,while the measured reference spur is-71 dBc in integer mode and the fractional spur is-65 dBc in fractional mode.     

基于蜂窝网络的物联网解决方案研究

移动通信技术向5G演进,推动工业革命向物联网时代迈进. “万物互联”的时代已经来临,物联网的发展进入了风口期,运营商都在考虑提前布局,培育市场,寻找新的收入增长点.本文将在探讨蜂窝网络的物联网解决方案的基础上,针对运营商如何部署物联网进行深入思考.     

Random forest classifier‐based safe and reliable routing for opportunistic IoT networks

Designing a safe and reliable way for communicating the messages among the devices and humans forming the Opportunistic Internet of Things network (OppIoT) has been a challenge since the broadcast mode of message sharing is used. To contribute toward addressing such challenge, this paper proposes a Random Forest Classifier (RFC)‐based safe and reliable routing protocol for OppIoT (called RFCSec) which ensures space efficiency, hash‐based message integrity, and high packet delivery, simultaneously protecting the network against safety threats viz. packet collusion, hypernova, supernova, and wormhole attacks. The proposed RFCSec scheme is composed of two phases. In the first one, the RFC is trained on real data trace, and based on the output of this training, the second phase consists in classifying the encountered nodes of a given node as belonging to one of the output classes of nodes based on their past behavior in the network. This helps in proactively isolating the malicious nodes from participating in the routing process and encourages the participation of the ones with good message forwarding behavior, low packet dropping rate, high buffer availability, and a higher probability of delivering the messages in the past. Simulation results using the ONE simulator show that the proposed RFCSec secure routing scheme is superior to the MLProph, RLProph, and CAML routing protocols, chosen as benchmarks, in terms of legitimate packet delivery, probability of message delivery, count of dropped messages, and latency in packet delivery. The out‐of‐bag error obtained is also minimal     

基于随机森林的大学英语四级通过率预测模型

为了得到一种实用性较强且具有较高精度的大学英语四级通过率的预测模型,本文尝试将随机森林模型应用到大学英语四级通过率预测中,以学生基本情况(性别、民族、专业)、高考英语成绩、大学英语成绩(共计4学期)、大学生课外英语使用统计数据为输入变量,以通过英语四级和未通过英语四级作为分类变量,建立基于随机森林预测模型.实验结果表明...     

A new model for network traffic forecast based on user's mobility in cellular networks with highway stretches

Wireless communications have always faced the scarceness of bandwidth and until today the cellular division has solved this problem by making radio cells always smaller. However, as cell size is reduced, more users will probably require handoffs, and signaling can overload the system. Further, the decrease in cell dimension cannot ensure the best utilization of system resources, because user density in the cell is not constant over the time. Dynamic allocation might be the solution, but it is a complex task, basically due to complexity of the models for the mobility prediction, especially in those environments where users move very fast (e.g. highways). This last one is a typical environment where high mobility of vehicles does not allow small cell sizes and thus system capacity is intrinsically bounded. In this paper, we deploy an analytical model for a cellular network operating in a high mobility environment. Such a model is capable of representing and forecasting wireless system evolution in terms of channel occupancy, starting from the current state of the mobile network and road traffic. This model has been defined by integrating a vehicular traffic model with a wireless cellular communication one. With such approach the dynamic behaviour of the most critical mobile terminals, those associated with vehicles, are described in a very precise and specific way with respect to mobility. Many results obtained with specific simulation tools and herein reported show the good behaviour of our proposal. Copyright © 2004 John Wiley & Sons, Ltd.     

中继蜂窝网中基于功率控制的频谱分配方案

中继蜂窝网络是近年来移动通信领域的热点,为了解决传统蜂窝网络的缺点,本文提出了在中继蜂窝网中基于功率控制的频谱分配方案,通过中继方式,解决了远近效应的问题,通过频率复用可以显著的提高系统频率资源利用率和系统性能,通过采用基站功率控制技术解决了由于资源复用带来的链路干扰过大问题。