小波变换下ARMA改进模型预测话务总量的研究

发布时间：2018-05-25 13:24

本文选题：话务量 + 小波变换　；参考：《重庆大学》2015年硕士论文

【摘要】：进入21世纪以后,随着科技和网络的发展,我国通信行业迅速壮大,特别是移动、联通和电信三大运营商,其网络规模不断扩大,业务种类多种多样,客户数量大量增加。在这个行业中,话务量是一个重要概念,它的大小不仅关系到客户的通信质量,而且为运营商提供了发展的依据。虽然地面上基站数量的多少和运营商网络设备的性能以及用户通话次数直接决定了话务量的大小,但是我们要做的是根据已知的话务量历史数据如何提前准确地预测话务量大小,以便及时地作出调整,避免潜在的风险。话务量历史数据作为一种时间序列,我们可以用很多成熟的时间序列模型来进行预测,但是如何更加准确地来进行预测是我们研究的重要课题。在本文中把小波变换思想引入到ARMA模型中来对其进行改进而后预测。基于此,本文的主要研究内容如下:①首先详细介绍了小波变换思想和单支重构算法。②着重分析了话务量的几种预测模型,主要包括自回归模型(AR)、滑动平均模型(MA)、自回归滑动平均模型(ARMA)等。③对已经采集的话务量原始序列进行小波分解,得到近似部分和各细节部分,并分别单支重构到原级别上,对各个重构后的序列分别建立ARMA模型,进而对原序列进行预测。④基于预测模型的理论分析,本论文也给出常用的ARMA模型对已有数据的预测结果,并将其结果与改进模型下的预测结果进行了对比研究。⑤系统地提出了通信网络话务量预测系统的设计思路。上述几方面的研究,为通信运营商预测话务量及相关需求分析给出了理论指导,也为国内通信行业话务量需求预测的工程应用作了有益的探索。
[Abstract]:After entering the 21st century, with the development of science and technology and network, the communication industry of our country grows rapidly, especially the three major operators of mobile, Unicom and telecom, whose network scale is expanding constantly, the service types are various, and the number of customers is increasing greatly. In this industry, traffic is an important concept, its size not only relates to the customer's communication quality, but also provides the basis for the development of the operator. Although the number of base stations on the ground and the performance of the operator's network equipment and the number of user calls directly determine the size of the traffic, But what we need to do is how to predict the traffic accurately and ahead of time according to the historical data of known traffic so as to adjust in time and avoid the potential risks. As a kind of time series, traffic history data can be predicted by many mature time series models. However, how to predict more accurately is an important topic of our research. In this paper, wavelet transform is introduced into ARMA model to improve it and then predict it. Based on this, the main research contents of this paper are as follows: firstly, wavelet transform and single-branch reconstruction algorithm .2 are introduced in detail, and several prediction models of traffic are analyzed. It mainly includes autoregressive model, moving average model, autoregressive moving average model and so on. 3. The original sequence of traffic that has been collected is decomposed by wavelet, and the approximate parts and the detail parts are obtained, and each branch is reconstructed to the original level. The ARMA model is established for each reconstructed sequence, and then the theoretical analysis based on the prediction model is carried out for the original sequence. The paper also gives the prediction results of the existing data by the commonly used ARMA model. The results are compared with the prediction results under the improved model. 5. The design idea of the traffic prediction system in the communication network is put forward systematically. The above research provides the theoretical guidance for the communication operators to predict the traffic volume and the related demand analysis, and also makes a beneficial exploration for the engineering application of the traffic demand prediction in the domestic communication industry.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F626.12;F224

【相似文献】