Indicator: Traffic per User

10th World Telecommunication/ICT

Indicators Meeting (WTIM-12) Bangkok, Thailand, 25-27 September 2012

Contribution to WTIM-12 session Document C/12-E 26 September 2012

English

SOURCE: Nokia Siemens Network

TITLE: Measuring mobile broadband data traffic – End user and access network centric

approach

1 © Nokia Siemens Networks 2012

ITU ICT Indicators Meeting Measuring Mobile Broadband Data Traffic

– End user and access network centric approach

Henri Helanterä, MSc

Solution Architect

Nokia Siemens Networks


Table of Contents

1. Introduction: Future of Mobile Communications

2. Mobile Traffic Indicators

– Indicator: Traffic per User

– Indicator: Traffic per Device Model / Device Type

– Indicator: Hourly Share of Daily Traffic

– Indicator: Traffic per Cell/Cluster

– Indicator: Traffic per Technology

– Indicator: Traffic per Access Point

– Indicator: Traffic per Application Protocol / Protocol Category

– Indicator: Traffic per Domain / Service

– Indicator: Mobile Data Service Usage by Age

3. Data Collection and Dissemination

4. Conclusions and Recommendations

Henri Helanterä / 25.09.2012


The Future of Mobile Communications


Communities, sharing

Navigation

Business on the move, email

Location

File Transfer, Personal Cloud

New applications and devices drive usage

Voice Data

Traffic

Revenue

Time

Network cost (4G)

Network cost

(2G/3G)

Profit

Mobile network traffic

/ Cost

~6X

Mobile data traffic booms

~3X

1.2 EB

7 EB

2012 2017e

Global Monthly Mobile Data Traffic Forecast, in total*

0.5 GB

1.5 GB

2012 2017e

Global Monthly Mobile Data Traffic Forecast, per user*

Source： NSN consulting； Cisco VNI mobile，2011；*1EB=1024PB，1PB=1024TB，1TB=1024GB

Video Streaming, Mobile TV

Online Gaming

Voice over IP

Music Downloading, Audio Streaming


Indicator: Traffic per User


Indicator Definition

• Traffic (UL+DL) in MB per User

• Users’ Share of Total Traffic (UL+DL)

– Users can be further grouped by data usage (i.e. # of users in

certain usage range)

Data sources

• Real-Time Traffic monitoring tools

• Gn/Gi interface probes

• xDRs

Source： NSN customer project

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

60.00%

70.00%

80.00%

90.00%

100.00%

0-100 MB 100-500 MB 500-1000 MB 1 GB +

Monthly Data Usage Histogram

Share of Users (%)

Share of Traffic (%)


Indicator: Traffic per Device Model / Device Type

0

50000

100000

150000

200000

7 8 9 10 11 12 13 14

Hourly Traffic for iPhone 4

Total (MB)

Uplink (MB)

Downlink (MB)

Share of Subscribers

Subscribers’ share of Total Traffic

Sh

are

of

Tra

ffic


• Traffic (UL+DL) in MB per Device Model / Device Type

Data sources



100%

50%

0 % 0% 50% 100%



Modem

iPhone

Other

smartphones


Indicator: Hourly Share of Daily Traffic


Busy hour is 7% of daily traffic


• Share of Traffic (UL/DL) per Hour of Daily Traffic

Data sources

• OSS PM counters




0

100

200

300

400

0

5000

10000

15000

20000

25000

30000

35000

00 03 06 09 12 15 18 21

Voice traffic (Erlang)

Data traffic (GB)


Indicator: Traffic per Cell/Cluster


• Traffic (UL+DL) in MB/GB per Cell

– Cells can be grouped into clusters to provide a geographical level for analysis

Traffic

City

60 GB

30 GB

0 GB


Data sources

• OSS Performance Management counters



• xDRs



Indicator: Traffic per Cell/Cluster

10% of cells contribute from 43% to 73% of

total traffic (data volume) depending on

hour

Approximately 54% of total traffic is

contributed by 15% of the cells during the

busy hour




Indicator: Traffic per Technology


80.24%

19.76%

Share of Traffic

3G

2G

0

500

1000

1500

2000

0% 0-20% 20-50% 50-80% 80-100%

Share of 3G Usage Time for 3G Devices

# of devices (thousands)


• Traffic (UL+DL) per Radio Access Technology (2G/3G)

• Time spent on 2G/3G per User

Data sources





Indicator: Traffic per Access Point


Blackberry.net – Hourly Traffic (GB)

Blackberry.net – HourlyTraffic per User (MB) Less users at night

but more traffic per

user


• Traffic (UL+DL) in GB per Access Point

Data sources





Indicator: Traffic per Application Protocol / Protocol Category

22%

Share of Traffic (DL)

Web

P2P

Web Media

News Groups

Others


• Traffic (UL+DL) in GB per Application Protocol / Protocol Category

Data sources






Indicator: Traffic per Domain / Service



• Total Unique Users per Internet Domain (Service)

• Traffic (UL+DL) in GB per Internet Domain (Service)

Data sources



0

50000

100000

150000

200000

250000

300000

Facebook Youtube Twitter

Total Unique Users

0

20

40

60

80

100


Success Rates (%)

HTTP GET HTTP POST

0

10

20

30

40

50

60

70

80

90

100


Total Traffic (GB)

Text Image Video



Indicator: Mobile Data Service Usage by Age


• Use of Mobile Data Services by Age / Age Group

• Traffic (UL+DL) per Mobile Data Service by Age / Age Group

Data sources

• Surveys



• Customer Relationship Management Systems Source： NSN study; base: 21,000 mobile phone owner respondents from maturing and emerging markets;

multiple answers possible


15-26 Years

26-35 Years

36-45 Years

46-55 Years

56-70 Years


Data Collection and Dissemination


Operator PLMN Operator services

ISP

Internet

E-mail

WAP

MMS Streaming

Downloading

Gn Gi Iu Operator IP network

Partner Corporate

• RRC/RAB Failures

• LAC, SAC, RNC • IMSI

• MSISDN • APN • Billing Type • Data Volumes • Throughput • IP-Address

• MSISDN • APN • IMEI • Qos • Data Volumes • Cause Codes

Mediation

RNC GGSN

RNC

SGSN

Core network (SGSN, GGSN, Gn, Gi) is the best source of subscriber and application level data

Charging


Web clients

Data processing

and enrichment

Data collection

Real-time analysis Long-term reporting Customer Analytics

Customer repository

CRM

BSS

Inventory

Mediation & control

Data warehouse

Real-Time Traffic Monitoring tools / Signalling logs

SGSN GGSN RNC

Gn, Gi probes IP Flows + DPI CDRs from SGSN, GGSN, ICD, PoC, IMS

Terminal Agents

Customer Care

Management Customer Care

Network Engineering

Marketing & Sales

Product Lifecycle Network Operations

… …

Data Collection and Dissemination (cont’d)



Conclusions and Recommendations

Mobile data traffic behaves differently from traditional Voice and SMS

• Device evolution and diverse range of applications drive growth in data traffic

• Analysing mobile data usage across different dimensions (user, application/service, device, cell/cluster, time) is important to better understand the relation between traffic and revenue – Understand how capacity is utilized in different parts of the network at different times (e.g. Central Business

District and sub-urban housing areas)

– Understand different user profiles and how they drive traffic and revenue

– Understand how Over-The-Top services and applications contribute to traffic and revenue

– Applying Fair Usage Policy to balance traffic, revenue and QoS

Sourcing and processing indicator data

• Sourcing and processing the transactional data needed for this analysis sets new requirements for mobile operators’ IT systems (big data) – Volume of transactional data is far greater than that of traditional Network Management statistics data

– Analyzed information and insights need to be made available in near real-time for timely and automated decision making and actions


Indicator: Traffic per User

Documents

Transcript of Indicator: Traffic per User