Embed Size (px)
Citation preview
UNDERSTANDING INTERNET USAGE AND NETWORK LOCALITY IN A RURAL COMMUNITY
WIRELESS MESH NETWORK
Adisorn Lertsinsrubtavee, Liang Wang, Nunthaphat Weshsuwannarugs, Arjuna Sathiaseelan, Apinun Tunpan,
Kanchana Kanchanasut, Jon Crowcroft
1
Computer Laboratory, University of Cambridge intERLab, Asian Institute of Technology
OUTLINE
➤ Internet in Rural Area of Thailand ➤ Community Network ➤ TakNet CWMN ➤ Social Interview ➤ Traffic Measurement & Data Analysis ➤ Discussion and Takeaways
2
INTERNET IN RURAL AREA OF THAILAND
3
28.92%ranked 6th in AEC, as of 2014
ITU -D
Thailand’s Internet Penetration
High cost for infrastructure investment
Low demand of using Internet
It is not cost-effective for ISP to invest network infrastructure
Digital Divide
Thai National Statistical Office, http://web.nso.go.th/
COMMUNITY NETWORK
4
Internet
Wireless ad-hoc network sharing the Internet gateway
Successful Community Networks !
TAKNET CWMN
5
Thai Samakhee a small rural village in northern Thailand
2 ADSL links provided by ISP
50 households with 300 population
Before 2013
28$/month for a subscription
TakNet CWMNInternet cost is shard among villagers
5$/month for a subscription
Attract villagers to use the Internet
TAKNET CWMN
6
Core Router
Access Router
ADSL gateway
R1
ADSL Gateway R2
R3
R4
R5
R7
R8
R9
R10
R12
R13
R14
Core Router
R6
R11
Ad hoc linkOLSR
WiFi
16 GB external storage
14 access routers (TPlink MR 3040) 1 core router (Unifi UAP) OpenWrt, Attitude Adjustment 12.04
UNDERSTANDING THE INTERNET USAGE
7
Traffic Measurement Social Interview
➤ Lightweight measurement ➤ Traffic volume
➤ ifconfig - 60 sec interval ➤ Internet usage
➤ tcpdump - HTTP request ➤ Filter out the URL
➤ Well-defined questionnaire ➤ Personal information ➤ Typical usage ➤ User feedback
➤ 30 mins interview ➤ Free-style conversation
SOCIAL INTERVIEW
8
18 3452
8
30
14
Kids Teens Adults8-16 16-21 over 22
140$ - 560$
40% 60%VS
91%Popular content
Device used
interviewees
User information
Teens
Adults
Kids
Monthly wages
SOCIAL INTERVIEW
9
User feedback85% of users install CM battery application (expect to improve their WiFi speed)
4 users opted out due to the extra cost incurred by electricity bill (just 1-2$/month)
Social Communications87%71%
33%81% of Line users have local contacts within the same village
10-20% of messages exchanged among local users
7.6%
21%
30%
Usage pattern
80% 17:00 - 22.0022:00 - 06.00
12:00 - 17.00
06:00 - 12.00
12.5%
4 hours per day on Internet usage
TRAFFIC USAGE
10
0 10 20 30 40 50 60 70 80 900
2000
4000
6000
8000
10000
12000
14000
Day (1 Febuary 2015 − 30 April 2015)
Tra
ffic
lo
ad
(M
B/d
ay)
Feb
Mar
Apr
Download
0 10 20 30 40 50 60 70 80 900
500
1000
1500
Day (1 Febuary 2015 − 30 April 2015)
Tra
ffic
lo
ad
(M
B/d
ay)
Feb
Mar
Apr
Upload
Traffic Growth Feb - Mar 15 Mar - Apr 15
Upload 25% 20%
Download 28.9% 15%
TRAFFIC USAGE PATTERN
11
0
0.5
1
1.5
2
2.5
3
3.5
Tra
ffic
lo
ad
(M
bp
s)
Time of day (Hours)
01.0
0
07.0
0
13.0
0
19.0
0
01.0
0
07.0
0
13.0
0
19.0
0
01.0
0
07.0
0
13.0
0
19.0
0
01.0
0
07.0
0
13.0
0
19.0
0
01.0
0
07.0
0
13.0
0
19.0
0
01.0
0
07.0
0
13.0
0
19.0
0
01.0
0
07.0
0
13.0
0
19.0
0
Mon Tue Wed Thu Fri Sat Sun
Dual off-peak hours 13:00 - 16:00 22:00 - 07:00
Traffic almost reached the committed speed (4Mbps)
How can we efficiently utilise these resources?
CONTENT POPULARITY
12
ksmobile naver baidu google 3g.cn umeng instagram fb youtube animaljam0
2
4
6
8
10
12
14
Domain (url)
Fra
cti
on
of
req
ue
sts
(%
)
Zipf distribution with alpha = 0.57
➤ Commonly known alpha (0.9 - 1.1) ➤ Mixture of misbehaviour domains ➤ Some valuable domains such as education and local newspaper
are pushed to the tail
CONTENT POPULARITY
13
naver google instagram fb youtube apple msn ytimg thscore dek−d0
2
4
6
8
10
12
14
Domain (url)
Fra
cti
on
of
req
ue
sts
(%
) Zipf distribution with alpha 1.05
Education website appears
Remove all suspicious domains
ANALYSIS OF SUSPICIOUS DOMAINS
14
Jain’s fairness indexThe higher value indicates the requests are more uniformly distributed
ksmobile naver baidu google 3g instagram umeng fb youtube animaljam0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Jain
Fair
ness In
dex
Large amount of requests were generated by some specific users
APPLICATION BEHAVIOUR
15
Inter arrival time of suspicious domain
baidu.com
Almost 80% of requests are made with an inter arrival
less than 2s
These requests were generated by the baidu
browsers0 5 10 15 20 25
0
200
400
600
800
1000
1200
1400
Nu
mb
er
of
req
ue
sts
Interval (s)
MISINFORMED KNOWLEDGE
➤ Several users misuse an application ➤ Considering ksmobile domain
➤ Android application —> CM Battery ➤ To save the battery power ➤ But! the villagers believe that it can use to accelerate the WiFi
speed ➤ Fact! it generates a lot of request to ksmobile domain ➤ Observation! advertisements are automatically downloaded to
users’ mobile phone ➤ 85% of villagers use this application
16
LOCALISED COMMUNICATION
17
0 2 4 6 8
x 104
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Ro
tue
r ID
Time of day (second)
Line application (naver)
HTTP request to Line server from each router
Line server
Sender Receiver
Send msg
Notification
Retrieve msg
LOCALISED COMMUNICATION
18
3.2 3.205 3.21 3.215 3.22 3.225 3.23
x 104
1
2
3
4
5
6
7
8
9
10
11
12
13
14
A pair of communication represents the localised communication
Achieve10% - 15% of identified pairs Varying window size 1s - 5s
From interview, 10% - 20% were sent to the local contacts
TAKEAWAYS
19
What are the impacts of TakNet ?➤ TakNet is able to create a demand within the community for
Internet access ➤ Number of Internet users in TakNet is increased significantly ➤ Villagers gain significant benefits form the Internet ➤ TakNet is a catalyst for changes: ISPs expand more backhaul to
cover the villages
TAKEAWAYS
20
Is there a universal model for all rural settings?
➤ Traffic pattern ➤ Asia1 - TakNet: Dual-peak pattern ➤ Africa2 and Europe3: Single peak:
➤ Localised communication ➤ TakNet: 10 - 15%, Africa2: ~50%
➤ Social Communications ➤ OSN (e.g., FB, Twitter) and email are popular services in rural
Africa2. ➤ Instant messaging is the most dominant service in TakNet
1 B. Du, et al. Analysis of www traffic in cambodia and ghana. In WWW ’06. ACM, 2006. 2 D. L. Johnson, et al. Network traffic locality in a rural african village. In ICTD. ACM, 2012. 3 A. Sathiaseelan, et al. A feasibility study of an in-the-wild experimental public access wifi network. In ACMDEV, 2014
TAKEAWAYS
➤ The available 4 Mbps bandwidth may be saturated soon in the near future.
➤ Can we simply expand the the link capacity or add more gateway? ➤ Villagers are very sensitive to the cost
➤ Can we utilise the off-peak hours with content/service caching ? ➤ Identify the true valuable contents
➤ Efficiently remove the suspicious domains ➤ New technologies
➤ Information Centric Network ➤ Service migration - virtualisation, container
21
What are the potential solutions to improve TakNet?
