WO2017095344A1 - A network traffic estimation system - Google Patents

A network traffic estimation system Download PDF

Info

Publication number
WO2017095344A1
WO2017095344A1 PCT/TR2016/000174 TR2016000174W WO2017095344A1 WO 2017095344 A1 WO2017095344 A1 WO 2017095344A1 TR 2016000174 W TR2016000174 W TR 2016000174W WO 2017095344 A1 WO2017095344 A1 WO 2017095344A1
Authority
WO
WIPO (PCT)
Prior art keywords
database
network traffic
information
estimation system
kept
Prior art date
Application number
PCT/TR2016/000174
Other languages
French (fr)
Inventor
Onur ONES
Omar OZENIR
Mehmet Onur SARKAN
Haci Ali GEMALMAZ
Ismail YILDIZ
Original Assignee
Turkcell Teknoloji Arastirma Ve Gelistirme Anonim Sirketi
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Turkcell Teknoloji Arastirma Ve Gelistirme Anonim Sirketi filed Critical Turkcell Teknoloji Arastirma Ve Gelistirme Anonim Sirketi
Priority to EP16831649.5A priority Critical patent/EP3384634B1/en
Publication of WO2017095344A1 publication Critical patent/WO2017095344A1/en

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
    • H04W16/18—Network planning tools

Definitions

  • the present invention relates to a network traffic estimation system which provides a solution architecture for estimating future network traffic via regression methods from machine learning approaches by looking at historical traffic data.
  • the teams controlling the network states gathers historical traffic data in excel files in order to make traffic estimation and they perform trend analysis manually.
  • increase of the network data which is required to be examined makes it impossible for the teams to examine each data. Therefore, the teams take small samples among the data reaching millions of records and then carry out the examination transaction. Gathering and examining so much data manually lead to a significant labour requirement. Large sizes of data may increase margin of error in data estimations of samples trend analysis of which are carried out.
  • the Chinese patent document no. CN102111284 discloses a method and device for predicting telecom traffic.
  • the method used in the said invention comprises steps of: determining the prediction granulation of the telecom traffic; selecting a historical and prediction sample; respectively calculating the initial values and the growth rates of the historical and prediction sample by using a unary linear regression model; calculating the traffic prediction of the historical and prediction sample; reading the actual traffic value of the historical sample; and calculating the second traffic prediction according to the deviation between the actual traffic value and the first traffic prediction.
  • Only historical traffic data are used in the said patent document. Therefore, rate of successful prediction is quite low.
  • the United States patent document no. US2003236084 discloses a system for estimating traffic rates of calls in environments wherein wireless personal communication services are provided.
  • the said system comprises a traffic parameter observation module for collecting real time observations on respective nodes and for making sets of nodes; a regression analysis module for performing regression analysis of observations to assume a prediction model for traffic rates of calls and to estimate traffic rates of internal -to-internal and external -to-internal calls, and a resource allocation module for directing resources. Whereas it is aimed to estimate densities in wireless personal communication networks instead of mobile phone networks in the said patent document.
  • the United States patent document no. US2010273493 discloses a radio access network management device, facility plan support system, and facility plan support method used therefor.
  • the invention is particularly related to allocation of base stations based on future-dated traffic estimation.
  • the traffic demand estimation unit included in the invention receives the past traffic demand data from near radio cells and uses them in regression analysis, and traffic demand occurs in the near future.
  • the said patent document relates to a solution for making regional estimation of traffic needs before investments such as base stations to be installed.
  • An objective of the present invention is to realize a network traffic estimation system whereby future network traffic are estimated automatically via regression methods from machine learning approaches by looking at historical traffic data.
  • Another objective of the present invention is to realize a network traffic estimation system whereby hardware requirement and positioning of a network are provided more efficiently by the estimations made and thus service quality and customer satisfaction are increased while enabling decrease in investment costs.
  • Figure 1 is a schematic view of the inventive network traffic estimation system.
  • the inventive system (1) for making traffic estimation in a certain period of time comprises:
  • At least one application unit (4) which determines the locations and the complaint categories that are possible to receive customer complaint by using the regression method on the information kept on the database (2); and at least one estimation unit (5) which makes traffic estimation by the data obtained in the application unit (4).
  • the inventive network traffic estimation system (1) handles the traffic estimation as a machine learning regression problem.
  • Current (up-to-date) network performance indicators, network configuration planning information, planned activities, currently planned works, current network alarms, weather forecasts, current customer and future campaign information are kept on the database (2) on the basis of location/cell
  • the information belonging to the network traffic history are kept on the database (2).
  • the said information comprise the information of traffic amount experienced in the network in the past.
  • the application unit (4) uses the information of date when the measurement of the traffic kept on the database (2) is made and the information of location (cell information) wherein the traffic is measured.
  • Information of the customer complaints history are also kept on the database (2).
  • the said information are the customer complaints experienced in the past.
  • the application unit (4) uses the information of date of complaint, category of complaint and location (cell) where the complaint is received which is stored on the database (2).
  • KPI Key performance indicator
  • Information belonging to the network alarm history kept on the database (2) are the history of the warning messages received from the network equipment with respect to the problems and failures.
  • the application unit (4) uses these information kept on the database (2) on the basis of location (cell) and date.
  • Information belonging to the history of network configuration parameters are kept on the database (2).
  • Information belonging to the history of work orders are kept on the database (2).
  • the said information are breakdown and maintenance work information on the locations (cells).
  • Information belonging to the history of weather forecast are kept on the database (2).
  • the said information are the most evaluable temperature, humidity, pressure, wind direction intensity and similar information where any location (cell) is located in any date.
  • Information belonging to the history of planned works are kept on the database (2).
  • the said information are the information of renewal, readjustment and maintenance works on any location (cell) on the basis of history.
  • Information belonging to the campaign history are kept on the database (2).
  • the said information comprise the information of data/call campaign realized by marketing teams in the past. Except the above-mentioned information, all kinds of data which can affect the traffic that may be used as input for the application unit (4) and the history and future values of which are known can be stored on the database (2).
  • the application unit (4) creates a model for traffic estimation in a certain period of time by regression method and for deciding whether there is an investment need or not.
  • the estimation unit (5) can make estimations by aggregation for the higher locations in the location hierarchy with the traffic estimations obtained by thereof.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The present invention relates to a network traffic estimation system (1) which provides a solution architecture for estimating future network traffic via regression methods from machine learning approaches by looking at historical traffic data. The inventive network traffic estimation system (1) handles the traffic estimation as a machine learning regression problem. Current network performance indicators, network configuration planning information, planned activities, currently planned works, current network alarms, weather forecasts, current customer and future campaign information are kept on the database (2) on the basis of location/cell.

Description

DESCRIPTION
A NETWORK TRAFFIC ESTIMATION SYSTEM Technical Field
The present invention relates to a network traffic estimation system which provides a solution architecture for estimating future network traffic via regression methods from machine learning approaches by looking at historical traffic data.
Background of the Invention
A team consisting of network planning experts follow traffic, alarm, breakdown and maintenance works in networks in order to continue quality services provided in networks. The teams controlling the network states gathers historical traffic data in excel files in order to make traffic estimation and they perform trend analysis manually. However, increase of the network data which is required to be examined makes it impossible for the teams to examine each data. Therefore, the teams take small samples among the data reaching millions of records and then carry out the examination transaction. Gathering and examining so much data manually lead to a significant labour requirement. Large sizes of data may increase margin of error in data estimations of samples trend analysis of which are carried out.
The Chinese patent document no. CN102111284 discloses a method and device for predicting telecom traffic. The method used in the said invention comprises steps of: determining the prediction granulation of the telecom traffic; selecting a historical and prediction sample; respectively calculating the initial values and the growth rates of the historical and prediction sample by using a unary linear regression model; calculating the traffic prediction of the historical and prediction sample; reading the actual traffic value of the historical sample; and calculating the second traffic prediction according to the deviation between the actual traffic value and the first traffic prediction. Only historical traffic data are used in the said patent document. Therefore, rate of successful prediction is quite low.
The United States patent document no. US2003236084 discloses a system for estimating traffic rates of calls in environments wherein wireless personal communication services are provided. The said system comprises a traffic parameter observation module for collecting real time observations on respective nodes and for making sets of nodes; a regression analysis module for performing regression analysis of observations to assume a prediction model for traffic rates of calls and to estimate traffic rates of internal -to-internal and external -to-internal calls, and a resource allocation module for directing resources. Whereas it is aimed to estimate densities in wireless personal communication networks instead of mobile phone networks in the said patent document. The United States patent document no. US2010273493 discloses a radio access network management device, facility plan support system, and facility plan support method used therefor. The invention is particularly related to allocation of base stations based on future-dated traffic estimation. The traffic demand estimation unit included in the invention receives the past traffic demand data from near radio cells and uses them in regression analysis, and traffic demand occurs in the near future. Whereas the said patent document relates to a solution for making regional estimation of traffic needs before investments such as base stations to be installed.
Summary of the Invention
An objective of the present invention is to realize a network traffic estimation system whereby future network traffic are estimated automatically via regression methods from machine learning approaches by looking at historical traffic data. Another objective of the present invention is to realize a network traffic estimation system whereby hardware requirement and positioning of a network are provided more efficiently by the estimations made and thus service quality and customer satisfaction are increased while enabling decrease in investment costs.
Detailed Description of the Invention
"A network traffic estimation system" realized to fulfill the objectives of the present invention is shown in the figure attached, in which:
Figure 1 is a schematic view of the inventive network traffic estimation system.
The components illustrated in the figure are individually numbered, where the numbers refer to the following: 1. System
2. Database
3. Combination unit
4. Application unit
5. Estimation unit
The inventive system (1) for making traffic estimation in a certain period of time comprises:
- at least one database (2) wherein history inputs gathered in the light of essentially network, seasonality, new field information and campaign subjects are stored;
at least one combination unit (3) wherein the method of machine learning is applied by combining the information located on the database (2) on the basis of location (cell) and date;
at least one application unit (4) which determines the locations and the complaint categories that are possible to receive customer complaint by using the regression method on the information kept on the database (2); and at least one estimation unit (5) which makes traffic estimation by the data obtained in the application unit (4).
The inventive network traffic estimation system (1) handles the traffic estimation as a machine learning regression problem. Current (up-to-date) network performance indicators, network configuration planning information, planned activities, currently planned works, current network alarms, weather forecasts, current customer and future campaign information are kept on the database (2) on the basis of location/cell
The information belonging to the network traffic history are kept on the database (2). The said information comprise the information of traffic amount experienced in the network in the past. In one preferred embodiment, the application unit (4) uses the information of date when the measurement of the traffic kept on the database (2) is made and the information of location (cell information) wherein the traffic is measured.
Information of the customer complaints history are also kept on the database (2). The said information are the customer complaints experienced in the past. In one preferred embodiment, the application unit (4) uses the information of date of complaint, category of complaint and location (cell) where the complaint is received which is stored on the database (2).
Information belonging to the history of network performance indicators (KPI(key performance indicator)) are kept on the database (2). The said information are a KPT set which indicate the network performance and these information are kept together with the location (cell) and date information.
Information belonging to the network alarm history kept on the database (2). The said information are the history of the warning messages received from the network equipment with respect to the problems and failures. In one preferred embodiment, the application unit (4) uses these information kept on the database (2) on the basis of location (cell) and date.
Information belonging to the history of network configuration parameters are kept on the database (2). The said information configuration parameters of any location (cell) on any date.
Information belonging to the history of work orders are kept on the database (2). The said information are breakdown and maintenance work information on the locations (cells).
Information belonging to the history of weather forecast are kept on the database (2). The said information are the most evaluable temperature, humidity, pressure, wind direction intensity and similar information where any location (cell) is located in any date.
Information belonging to the history of planned works are kept on the database (2). The said information are the information of renewal, readjustment and maintenance works on any location (cell) on the basis of history.
Information belonging to the campaign history are kept on the database (2). The said information comprise the information of data/call campaign realized by marketing teams in the past. Except the above-mentioned information, all kinds of data which can affect the traffic that may be used as input for the application unit (4) and the history and future values of which are known can be stored on the database (2).
In the inventive traffic estimation system (1), the application unit (4) creates a model for traffic estimation in a certain period of time by regression method and for deciding whether there is an investment need or not. In one preferred embodiment of the invention, the estimation unit (5) can make estimations by aggregation for the higher locations in the location hierarchy with the traffic estimations obtained by thereof. Within these basic concepts; it is possible to develop a wide range of embodiments of a network traffic estimation system (1), the invention cannot be limited to examples disclosed herein and it is essentially according to claims.

Claims

1. A network traffic estimation system (1) for making traffic estimation in a certain period of time; characterized by:
- at least one database (2) wherein history inputs gathered in the light of essentially network, seasonality, new field information and campaign subjects are stored;
at least one combination unit (3) wherein the method of machine learning is applied by combining the information located on the database (2) on the basis of location (cell) and date;
at least one application unit (4) which determines the locations and the complaint categories that are possible to receive customer complaint by using the regression method on the information kept on the database (2); and
- at least one estimation unit (5) which makes traffic estimation by the data obtained in the application unit (4).
2. A network traffic estimation system (1) according to Claim 1, characterized by the database (2) wherein the information belonging to the network traffic history are kept.
3. A network traffic estimation system (1) according to Claim 2, characterized by the application unit (4) which uses the information of date when the measurement of the traffic kept on the database (2) is made and the information of location (cell information) wherein the traffic is measured.
4. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information of the customer complaints history are kept.
5. A network traffic estimation system (1) according to Claim 4, characterized by the application unit (4) which uses the information of date of complaint, category of complaint and location (cell) where the complaint is received that is stored on the database (2).
6. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information belonging to the history of network performance indicators (KPI(key performance indicator)) are kept.
7. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information belonging to the network alarm history are kept.
8. A network traffic estimation system (1) according to Claim 7, characterized by the application unit (4) which uses these information kept on the database (2) on the basis of location (cell) and date.
9. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information belonging to the history of network configuration parameters are kept.
10. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information belonging to the history of work orders are kept.
11. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information belonging to the history of weather forecast are kept.
12. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information belonging to the history of planned works are kept.
13. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein information belonging to the history of campaign history are kept.
14. A network traffic estimation system (1) according to any of the preceding claims, characterized by the database (2) wherein all kinds of data which can affect the traffic that may be used as input for the application unit (4) and the history and future values of which are known can be stored on the database (2).
15. A network traffic estimation system (1) according to any of the preceding claims, characterized by the application unit (4) which creates a model for traffic estimation in a certain period of time by regression method and for deciding whether there is an investment need or not.
16. A network traffic estimation system (1) according to any of the preceding claims, characterized by the estimation unit (5) which can make estimations by aggregation for the higher locations in the location hierarchy with the traffic estimations obtained by thereof.
PCT/TR2016/000174 2015-12-04 2016-12-02 A network traffic estimation system WO2017095344A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP16831649.5A EP3384634B1 (en) 2015-12-04 2016-12-02 A network traffic estimation system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
TR2015/15448A TR201515448A2 (en) 2015-12-04 2015-12-04 A NETWORK TRAFFIC FORECASTING SYSTEM
TR2015/15448 2015-12-04

Publications (1)

Publication Number Publication Date
WO2017095344A1 true WO2017095344A1 (en) 2017-06-08

Family

ID=57910094

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/TR2016/000174 WO2017095344A1 (en) 2015-12-04 2016-12-02 A network traffic estimation system

Country Status (3)

Country Link
EP (1) EP3384634B1 (en)
TR (1) TR201515448A2 (en)
WO (1) WO2017095344A1 (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2002044901A2 (en) * 2000-11-29 2002-06-06 Netuitive Inc. Computer performance forecasting system
US20030236084A1 (en) 2002-06-20 2003-12-25 Ki-Dong Lee System for estimating traffic rate of calls in wireless personal communication environment and method for the same
US20100273493A1 (en) 2007-12-12 2010-10-28 Nec Corporation Radio access network management device, facility plan support system, and facility plan support method used therefor
CN102111284A (en) 2009-12-28 2011-06-29 北京亿阳信通软件研究院有限公司 Method and device for predicting telecom traffic
US20130198767A1 (en) * 2012-01-30 2013-08-01 Board Of Regents, The University Of Texas System Method and apparatus for managing quality of service
EP2750432A1 (en) * 2012-12-28 2014-07-02 Telefónica, S.A. Method and system for predicting the channel usage
WO2014127051A1 (en) * 2013-02-14 2014-08-21 Adaptive Spectrum And Signal Alignment, Inc. Churn prediction in a broadband network
US20150195149A1 (en) * 2014-01-06 2015-07-09 Cisco Technology, Inc. Predictive learning machine-based approach to detect traffic outside of service level agreements

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2002044901A2 (en) * 2000-11-29 2002-06-06 Netuitive Inc. Computer performance forecasting system
US20030236084A1 (en) 2002-06-20 2003-12-25 Ki-Dong Lee System for estimating traffic rate of calls in wireless personal communication environment and method for the same
US20100273493A1 (en) 2007-12-12 2010-10-28 Nec Corporation Radio access network management device, facility plan support system, and facility plan support method used therefor
CN102111284A (en) 2009-12-28 2011-06-29 北京亿阳信通软件研究院有限公司 Method and device for predicting telecom traffic
US20130198767A1 (en) * 2012-01-30 2013-08-01 Board Of Regents, The University Of Texas System Method and apparatus for managing quality of service
EP2750432A1 (en) * 2012-12-28 2014-07-02 Telefónica, S.A. Method and system for predicting the channel usage
WO2014127051A1 (en) * 2013-02-14 2014-08-21 Adaptive Spectrum And Signal Alignment, Inc. Churn prediction in a broadband network
US20150195149A1 (en) * 2014-01-06 2015-07-09 Cisco Technology, Inc. Predictive learning machine-based approach to detect traffic outside of service level agreements

Also Published As

Publication number Publication date
EP3384634B1 (en) 2020-09-16
TR201515448A2 (en) 2017-06-21
EP3384634A1 (en) 2018-10-10

Similar Documents

Publication Publication Date Title
US9503919B2 (en) Wireless communication network using multiple key performance indicators and deviations therefrom
KR101676743B1 (en) Prediction and root cause recommendations of service access quality of experience issues in communication networks
US8255524B2 (en) Arrangement and a method relating to performance monitoring
Torres et al. Data analytics for forecasting cell congestion on LTE networks
EP2750432A1 (en) Method and system for predicting the channel usage
CN104113869B (en) A kind of potential report user's Forecasting Methodology and system based on signaling data
US11509518B2 (en) Network fault diagnosis
JPWO2009075246A1 (en) Wireless access network management apparatus, equipment plan support system, and equipment plan support method used therefor
CN113947260A (en) User satisfaction prediction method and device and electronic equipment
KR20190103661A (en) Method of predicting fine dustreflecting both indoor and outdoordata
US10085157B2 (en) Reconfiguring a mobile network based on cellular network state information
Slimen et al. Anomaly prevision in radio access networks using functional data analysis
Moysen et al. Big data-driven automated anomaly detection and performance forecasting in mobile networks
CN107465559A (en) A kind of network security supervising platform
JP5023998B2 (en) Wireless communication network quality evaluation system, communication quality improvement area extraction method and program
EP3384634B1 (en) A network traffic estimation system
Pimpinella et al. Crowdsourcing or network kpis? A twofold perspective for QoE prediction in cellular networks
US20220251817A1 (en) Systems and methods for managing stormwater regulations
Khatib et al. Degradation detection algorithm for lte root cause analysis
Hoang et al. Integrating energy efficiency-based prognostic approaches into energy management systems of base stations
CN113556773A (en) Method for detecting sleeping cell based on AI model
Shen et al. Research on Network Quality Intelligent Monitoring Technology Based on Prophet Model
Fong Data mining for resource planning and QoS supports in GSM networks
US20230134035A1 (en) Systems and methods for prioritizing repair and maintenance tasks in telecommunications networks
CN113543187B (en) Cell control method and device

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16831649

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE