Forecasting Telecommunication Network States on the Basis of Log Patterns Analysis and Knowledge Graphs Modeling
Author:
Affiliation:
1. Saint-Petersburg Electrotechnical University “LETI”, Russia
2. St. Petersburg Federal Research Centre of the Russian Academy of Sciences (SPCRAS), Russia
Abstract
The article proposes a state forecasting method for telecommunications networks (TN) that is based on the analysis of behavioral models observed on users' network devices. The method applies user behavior that makes it possible to forecast with more accuracy both the network parameters and the load at various back-ends. Suggested forecasts facilitate implementing reasonable reconfiguration of the TN. The new method proposed as a further development of TN states the forecasting method presented by the authors before. In this new version, forecasting algorithm users' behavioral models are involved. The models refer to a class of time diagrams of device transitions between different states. The novelty of the proposed method is that resulting TN models enable forecasting device state transitions represented in a device state diagram in the form of knowledge graph, in particular changes in loads of different back-ends. The provided case study for a subgroup of network devices demonstrated how their states can be forecasted using behavioral models obtained from log files.
Publisher
IGI Global
Subject
General Computer Science
Reference59 articles.
1. Allen, R. (1997). A formal approach to software architecture [Ph.D. Thesis]. Carnegie Mellon University, Pittsburgh, PA. CMU Technical Report CMU-CS-97-144.
2. A formal basis for architectural connection
3. Anand, Scoglio, & Natarajan. (2008). GARCH Non-Linear Time Series Model for Traffic Modeling and Prediction. IEEE.
4. Survey on traffic prediction in smart cities
5. Building an IoT Data Hub with Elasticsearch, Logstash and Kibana
Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Synthesis of multilevel knowledge graphs: Methods and technologies for dynamic networks;Engineering Applications of Artificial Intelligence;2023-08
1.学者识别学者识别
2.学术分析学术分析
3.人才评估人才评估
"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370
www.globalauthorid.com
TOP
Copyright © 2019-2024 北京同舟云网络信息技术有限公司 京公网安备11010802033243号 京ICP备18003416号-3