LTE-A网络中断检测技术研究

 2021-12-16 08:12

论文总字数:33192字

摘 要

近年来,随着智能手机和各种移动网络终端的普及以及4G网络的部署完善,移动互联网业务得到了爆炸性的增长。这让各大运营商的通信服务网络面临着前所未有的压力挑战。在节假日或大型活动举办地由于突发的巨大业务量而造成的基站设备故障时有发生。新技术和新网络的引入增强了服务承载能力也带来了兼容性的问题,使网络维护问题越来越复杂和困难。且新的应用场景带来了新的业务增长点同时也带来了新的故障案例。这些新问题对运营商的运维能力提出了更高的要求。

当蜂窝小区由于各种不同原因而发生中断故障时,能够迅速地将其检测出来并及时处理变得越来越重要。在自组织网络的概念中,自我修复部分的功能就是自动的处理一些网络故障。这需要针对明确的故障进行一系列操作以恢复正常服务,而进行这些处理操作的前提就是要发现这些故障。在一个庞大的移动通信网络中,一个区域内会有多个蜂窝设备同时工作,而每个网络设备都有不同的监控参数。但这些参数的变化所表示的含义并不是显而易见的,在传统的警报式监控端工作人员难以及时从海量数据中发现故障,同时也难以由人工方式逐一排查可能的故障原因,因此我们需要对这些参数进行分析处理以检测出问题。

本文将讨论两种中断检测算法。首先,是基于统计分析的样本对比算法,通过对监控参数进行实时对比分析,检测可能出现故障的参数。另外,采用基于AP(Affinity Propagation)聚类的算法实现中断设备的发现,并进行定位。针对不同的监控参数可以有不同的处理策略,本文将算法实现到几种监控参数上作为参考实例来验证算法的可行性。通过此算法可实现简洁高效的中断检测流程,有较高的实用性。

关键词:中断检测,自组织网络,AP聚类算法

Research of Cell Outage Detection Technology

in LTE-A System

04011633 Wu Siqi

Supervisor Pan Zhiwen

Abstract

For the past few years, smart phones and other portable devices become very popular as 4G networks are deploying in many countries. Thus, the demand for broadband mobile network services experience a great boost across the world. However, the communication networks of the operators will be under extreme pressure in particular scenario. Cell outage in holiday and traditional festival become regular because of the unimaginably large data traffic within hours even minutes. The introduction of new technologies and new networks brings stronger load capacity but also leads to compatibility problems which make the maintenance work more complicated. The new usage scenarios bring up new points of increase in business and new failure scenarios for network equipment. All of these call for higher level of operation and maintenance works of the operators.

When the cells get into outage conditions, it is important for the operators to detect it and fix it on time. In the concept of Self-Organizing Networks, self-healing mechanisms aim at automatically reducing the impacts from faulty equipment. These require the ability to spot such failing equipment in order to take further measures. There are several equipment with different monitoring parameters within certain area in a complex mobile network. It is not easy to recognize the meaning of each changing of these parameters. In traditional networks, it is hard to identify a failure from such massive data and hard to find out the reason of the failure. So we need to build a mechanism to solve these problem.

In this paper, approaches of automatic cell outage detection algorithm are discussed. First, there are a statistic analysis towards the monitoring parameters. Comparison between actual CDFs and sample CDFs are processed in order to identify anomaly of the parameters. In addition, an algorithm based on affinity propagation clustering is introduced to spot and locate the outage cell. For different monitoring parameters, there will be different strategies to work with. This is a simple approach of cell outage detection with high efficiency and it will be illustrated in a simulating LTE-A system.

Keyword: Outage detection, Self-organizing network, AP Clustering Algorithm

目录

  1. 绪论……………………………………..………………………7
    1. 背景和意义……………………………………………….................7
    2. 研究现状………………………………………………….................9
    3. 本文主要内容………………………………………………...........11
    4. 文章结构………………………………………………..………….11

第二章 LTE-A中断检测……………………………………………...13

2.1 应用场景…………………………………………………...............13

2.2 问题描述…………………………………………………...............13

2.3 解决方案………………………………….……………..................14

第三章 中断检测算法…………………………………………………15

3.1 聚类算法………………………………………….……................. 15

3.1.1 AP聚类算法概述……………………………………………..17

3.1.2 AP聚类算法实现………………………………………..……18

3.2 统计样本算法…………………………………………...................19

3.2.1 统计样本算法概述………………………..…………..……..19

3.2.2 统计样本算法实现…………………………………………..20

第四章 LTE-A系统仿真平台…………………………………………23

4.1 LTE-A系统概述………………………………………….…………23

4.2 系统模型搭建……………………………………………...............24

4.3 仿真流程…………………………………………….…..................25

第五章 结果分析………………………………………………………27

5.1 仿真场景………………………………………………..................27

5.2 仿真结果………………………………………………..................34

5.2.1 AP聚类算法仿真结果………………………………………..37

5.2.2 统计样本算法仿真结果……………………………………..34

第六章 总结……………………………………………………………40

致谢…………………………………………………………………..…42

参考文献………………………………………………………………..43

  1. 绪论
    1. 背景和意义

近年来各大网络运营商的4G网络覆盖逐步完善,宽带移动网络及其服务应用得到广泛普及。第四代通信技术有着更好的网络承载能力和网络性能,使得更高质量更丰富的网络服务得以在移动终端上实现。电信运营商更好的业务水平也让用户得以在移动终端上得到更好的使用体验。当前百舸争流的移动互联网行业乘着4G网络普及的大潮得到了爆炸性的增长。各种业务创新利用新网络更强的承载能力创造了新的商业模式,开拓了前所未有的新市场,也改变了人们传统的生活方式,让人们得以更便捷更迅速的实现互相交换信息的需求。但这同时也考验着电信运营商的运营维护人员对设备服务水平的保障能力。

逢年过节,亲朋好友之间祝福或问候的短信在短短几天甚至几小时内如潮水般涌来,使网络的峰值承载能力受到考验。但近年来移动互联网的发展使得短信的潮水转变成了各种即时通信应用造成的信令风暴。在大型体育场馆或展览会场,极高密度的人群带来了极高密度的通话服务需求,考验的是运营商的应急通信保障能力。近年来第一时间的新闻报道和人们的社交分享需求则使通话需求转变成大流量的数据传输需求。在偏远山区,在海岛景点,在地铁深处,在各种难以部署完善网络的地点,网络的覆盖能力也受到考验。由于4G网络工作在较高频率上,无线信号受环境影响较大,运营商需要比以往部署更多的基站。当前,虽然不断升级的移动网络已经带来了数百倍于以往的网络性能,但移动互联网业务的发展却更快,也在不断发生这需求变化,使得许多新问题出现在了运营商的日常维护中。

剩余内容已隐藏,请支付后下载全文,论文总字数:33192字

您需要先支付 80元 才能查看全部内容!立即支付

该课题毕业论文、开题报告、外文翻译、程序设计、图纸设计等资料可联系客服协助查找;