a machine learning based approach to mobile network...
TRANSCRIPT
![Page 1: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/1.jpg)
A Machine Learning Based Approach to Mobile Network Analysis
Zengwen Yuan1, Yuanjie Li1, Chunyi Peng2, Songwu Lu1, Haotian Deng2, Zhaowei Tan1, Taqi Raza1
1 2
![Page 2: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/2.jpg)
Overview�2
![Page 3: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/3.jpg)
Background
Overview�2
![Page 4: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/4.jpg)
Background
Why machine learning for mobile network analysis
Overview�2
![Page 5: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/5.jpg)
Background
Why machine learning for mobile network analysis
Mobile network analysis: state-of-the-a; and our approach
Overview�2
![Page 6: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/6.jpg)
Background
Why machine learning for mobile network analysis
Mobile network analysis: state-of-the-a; and our approach
Case study: analyzing latency for mobile networks
• How mobile apps work over LTE
• How to breakdown app-perceived latency
• Challenges and ML scheme
• Primary results from crowdsourcing
Overview�2
![Page 7: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/7.jpg)
Background
Why machine learning for mobile network analysis
Mobile network analysis: state-of-the-a; and our approach
Case study: analyzing latency for mobile networks
• How mobile apps work over LTE
• How to breakdown app-perceived latency
• Challenges and ML scheme
• Primary results from crowdsourcing
Conclusion
Overview�2
![Page 8: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/8.jpg)
Ubiquitous cellular networks connect everyone, everything
�3
![Page 9: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/9.jpg)
The race to 5G opens many new oppoCunities�4
![Page 10: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/10.jpg)
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application StackWeb
server
Internet??
Cellular network (4G LTE)User space
Yet, access to mobile network analytics is barred�5
What’s going on in the 3G/4G/5G network??
Oh we cannot tell you unless you sign an NDA…
Researcher (you)
No privilege no talk, sorry!
Chipset/Mobile OS
Operators
![Page 11: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/11.jpg)
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application StackWeb
server
Internet??
Cellular network (4G LTE)User space
Yet, access to mobile network analytics is barred�5
What’s going on in the 3G/4G/5G network??
Oh we cannot tell you unless you sign an NDA…
Researcher (you)
No privilege no talk, sorry!
Chipset/Mobile OS
Operators
3G/4G operations remain closed both in device chipsets and network infrastructures :(
![Page 12: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/12.jpg)
Plus, mobile networks are complex & distributed
More complex functions on both control and data planes
Operations are distributed across layers
�6
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
TCP/IPProtocol Stack
Application Protocols
Radio Resource Control (RRC)
Mobility Management (EMM)
Session Management (ESM)
Packet Convergence (PDCP)
Radio Link Control (RLC)
Medium Access Control (MAC)
Physical Layer Protocols (PHY)
Control Plane
Data Plane
Signaling
![Page 13: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/13.jpg)
Moreover, analytics tasks are app speciNc�7
![Page 14: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/14.jpg)
Analytics for mobile networks is problem-speciCc, for example:
Moreover, analytics tasks are app speciNc�7
![Page 15: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/15.jpg)
Analytics for mobile networks is problem-speciCc, for example:
• Web browsers:✦ Why the time-to-first-byte (TTFB) is so long?
✦ What’s the major component of latency?
✦ …
Moreover, analytics tasks are app speciNc�7
![Page 16: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/16.jpg)
Analytics for mobile networks is problem-speciCc, for example:
• Web browsers:✦ Why the time-to-first-byte (TTFB) is so long?
✦ What’s the major component of latency?
✦ …
• Instant message apps:✦ Does the recipient read my message?
✦ Is my message delivered in time?
✦ …
Moreover, analytics tasks are app speciNc�7
![Page 17: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/17.jpg)
State-of-the-aC mobile network analytics�8
![Page 18: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/18.jpg)
Current 4G network analytics is primarily “infrastructure-based”:
State-of-the-aC mobile network analytics�8
![Page 19: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/19.jpg)
Current 4G network analytics is primarily “infrastructure-based”:
State-of-the-aC mobile network analytics�8
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application StackWeb
server
Internet
Cellular network (4G LTE)
Base stations
Gateway
Gateway
User profile server
Mobility controller
User space
![Page 20: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/20.jpg)
Current 4G network analytics is primarily “infrastructure-based”:
State-of-the-aC mobile network analytics�8
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application StackWeb
server
Internet
Cellular network (4G LTE)
Base stations
Gateway
Gateway
User profile server
Mobility controller
User space
![Page 21: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/21.jpg)
Current 4G network analytics is primarily “infrastructure-based”:
State-of-the-aC mobile network analytics�8
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application StackWeb
server
Internet
Cellular network (4G LTE)
Base stations
Gateway
Gateway
User profile server
Mobility controller
User space
Not Scalable
![Page 22: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/22.jpg)
Current 4G network analytics is primarily “infrastructure-based”:
State-of-the-aC mobile network analytics�8
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application StackWeb
server
Internet
Cellular network (4G LTE)
Base stations
Gateway
Gateway
User profile server
Mobility controller
User space
Not Scalable Incomplete View
![Page 23: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/23.jpg)
Current 4G network analytics is primarily “infrastructure-based”:
State-of-the-aC mobile network analytics�8
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application StackWeb
server
Internet
Cellular network (4G LTE)
Base stations
Gateway
Gateway
User profile server
Mobility controller
User space
Not Scalable Incomplete View Opaqueness
![Page 24: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/24.jpg)
ML-based approach is a must-have feature for mobile network analytics
�9
Not Scalable Incomplete View Opaqueness
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 25: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/25.jpg)
Device-centric ML approach brings new hope
ML-based approach is a must-have feature for mobile network analytics
�9
Not Scalable Incomplete View Opaqueness
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 26: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/26.jpg)
Device-centric ML approach brings new hope
ML-based approach is a must-have feature for mobile network analytics
�9
Not Scalable Incomplete View Opaqueness
Scalability
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 27: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/27.jpg)
Device-centric ML approach brings new hope
ML-based approach is a must-have feature for mobile network analytics
�9
Not Scalable Incomplete View Opaqueness
Scalability Device QoE View
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 28: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/28.jpg)
Device-centric ML approach brings new hope
ML-based approach is a must-have feature for mobile network analytics
�9
Not Scalable Incomplete View Opaqueness
Scalability Device QoE View Availability
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 29: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/29.jpg)
It is probably true that machine learning is a must-have approach, rather than a nice-to-have one, to our Celd for mobile network analysis
![Page 30: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/30.jpg)
Our proposal: two-level device-centric ML approach �11
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 31: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/31.jpg)
Local level: sensing mobile network data inside each sma;phone
• Via hardware-software coordination (e.g. MobileInsight [ACM MobiCom’16])
• Via higher-layer (application/transport/IP) and lower-layer (cellular-specific) integration
• Via ML-assisted data plane prediction from control plane protocol reconstruction
Our proposal: two-level device-centric ML approach �11
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 32: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/32.jpg)
Local level: sensing mobile network data inside each sma;phone
• Via hardware-software coordination (e.g. MobileInsight [ACM MobiCom’16])
• Via higher-layer (application/transport/IP) and lower-layer (cellular-specific) integration
• Via ML-assisted data plane prediction from control plane protocol reconstruction
Global level:
• Crowdsourcing-based dataset
• Cloud-synthesized insights
Our proposal: two-level device-centric ML approach �11
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Mob
ile A
pps
Ba
seba
nd
Mob
ile O
S
TCP/
IP s
tack
LTE
inte
rfac
e
Appl
icat
ion
Stac
k
User
spa
ce
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 33: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/33.jpg)
Local analysis�12
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 34: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/34.jpg)
Step 1: open up the “black-box” operations
• At/above IP network data: TCPDUMP
• Below IP network data: MobileInsight
Local analysis�12
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 35: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/35.jpg)
Step 1: open up the “black-box” operations
• At/above IP network data: TCPDUMP
• Below IP network data: MobileInsight
Local analysis�12
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 36: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/36.jpg)
Step 1: open up the “black-box” operations
• At/above IP network data: TCPDUMP
• Below IP network data: MobileInsight
Step 2: automated data preprocessing
• Data cleansing and integration of two sources
Local analysis�12
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
![Page 37: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/37.jpg)
Step 1: open up the “black-box” operations
• At/above IP network data: TCPDUMP
• Below IP network data: MobileInsight
Step 2: automated data preprocessing
• Data cleansing and integration of two sources
Local analysis�12
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Preprocessing
![Page 38: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/38.jpg)
Step 1: open up the “black-box” operations
• At/above IP network data: TCPDUMP
• Below IP network data: MobileInsight
Step 2: automated data preprocessing
• Data cleansing and integration of two sources
Step 3: local ML-based analysis
• Control plane for protocol operations
• Data plane for performance
Local analysis�12
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Preprocessing
![Page 39: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/39.jpg)
Step 1: open up the “black-box” operations
• At/above IP network data: TCPDUMP
• Below IP network data: MobileInsight
Step 2: automated data preprocessing
• Data cleansing and integration of two sources
Step 3: local ML-based analysis
• Control plane for protocol operations
• Data plane for performance
Local analysis�12
Mobile Apps
Baseband
Mobile OS
TCP/IP stack
LTE interface
Application Stack
User space
Preprocessing ML analysis
![Page 40: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/40.jpg)
Global analysis
Enabled by cloud-based crowdsourcing (e.g. cniCloud [HotWireless’17])
Analytical Insights for:
• Geographical location
• Operators
• Phone models
• …
�13
…
SQL Query
SQL Response
Fine-grainedlogging&sharing
EfficientDataManagement
StructuredQuery
![Page 41: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/41.jpg)
Case study: latency analysis in mobile networks
![Page 42: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/42.jpg)
Every millisecond of latency maSers!
Mobile network users want fast access
• 1 second latency in page response → 7% reduction in PageView [KissMetrics 2011]
Developers lose revenue due to long latency
• Every 100 ms costs Amazon 1% ($1.6 bn) in sales
• An extra 400 ms latency drops daily Google searches per user by 0.6%
Latency does maZer a lot!
�15
![Page 43: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/43.jpg)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
![Page 44: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/44.jpg)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
![Page 45: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/45.jpg)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet?
![Page 46: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/46.jpg)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet?
![Page 47: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/47.jpg)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
![Page 48: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/48.jpg)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
![Page 49: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/49.jpg)
Cellular network (4G LTE)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
![Page 50: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/50.jpg)
Cellular network (4G LTE)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
Base stations
Gateway
Gateway
User profile server
Mobility controller
![Page 51: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/51.jpg)
Cellular network (4G LTE)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
(a)
(b)
Base stations
Gateway
Gateway
User profile server
Mobility controller
![Page 52: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/52.jpg)
Cellular network (4G LTE)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
(a)
(b) (c)
(c)
(c)
Base stations
Gateway
Gateway
User profile server
Mobility controller
![Page 53: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/53.jpg)
Cellular network (4G LTE)
(d)
(d) (d)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
(a)
(b) (c)
(c)
(c)
Base stations
Gateway
Gateway
User profile server
Mobility controller
![Page 54: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/54.jpg)
Cellular network (4G LTE)
(d)
(d) (d)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
(a)
(b) (c)
(c)
(c)
Base stations
Gateway
Gateway
User profile server
Mobility controller
(f)
(e)
![Page 55: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/55.jpg)
Cellular network (4G LTE)
(d)
(d) (d)
Background: how do mobile apps work over 4G LTE?
What happens under the hood?
How LTE impacts perceived latency on mobile web/IM app?
�16
Safari WhatsApp
modem chipset
mobile OS
TCP/IP stack
LTE interface
Application(HTTP/DNS)
Web server
Internet
(a)
(b) (c)
(c)
(c)
Base stations
Gateway
Gateway
User profile server
Mobility controller
(f)
(e)
LTE control-plane operations pose sizable latency on mobile apps
![Page 56: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/56.jpg)
Timing breakdown of control plane operations�17
![Page 57: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/57.jpg)
Timing breakdown of control plane operations�17
P1a: RRC connection setup request
![Page 58: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/58.jpg)
Timing breakdown of control plane operations�17
P1a: RRC connection setup request
(Random Access)
![Page 59: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/59.jpg)
Timing breakdown of control plane operations�17
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
![Page 60: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/60.jpg)
Timing breakdown of control plane operations�17
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
![Page 61: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/61.jpg)
Timing breakdown of control plane operations�17
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
![Page 62: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/62.jpg)
Timing breakdown of control plane operations�17
P3. Authentication (non-mandatory)
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
![Page 63: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/63.jpg)
Timing breakdown of control plane operations�17
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
![Page 64: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/64.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
![Page 65: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/65.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
P5b. Security mode complete
![Page 66: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/66.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
![Page 67: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/67.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
![Page 68: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/68.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
Data bearer Data bearer
![Page 69: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/69.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
Data bearer Data bearerUL data
![Page 70: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/70.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
Data bearer Data bearerUL data
![Page 71: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/71.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup complete
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
Data bearer Data bearerUL data
![Page 72: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/72.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup completeT
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
Data bearer Data bearerUL data
![Page 73: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/73.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup completeT
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
Data bearer Data bearerUL data
![Page 74: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/74.jpg)
Timing breakdown of control plane operations�17
P5a. Security mode command
P3. Authentication (non-mandatory)
P4. Initial Context Setup
P1a: RRC connection setup request
(Random Access)
P1b: RRC connection setup
P1c: RRC connection setup completeT
P2. Service request
P5b. Security mode complete
P6a. RRC connection reconfig
P6b RRC connection reconfig complete
T
Data bearer Data bearerUL data
![Page 75: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/75.jpg)
Learning latency: latency data sensing�18
Three-tiered timing data collection:
• App-specific semantic timing (e.g. Navigation Timing API, IM timing model)
• TCP/IP stack timing (from TCPDUMP)
• LTE stack timing (from MobileInsight)
DNS query TCP connection
TCP SYN
Data access request
LTE data
SYN ACK
HTTP request
LTE control plane
LTE data plane
OS overhead
HTTP transmission
Pagerendering
TCP dataTCP layer
unloadEventStartfetchStart
domainLookupStartdomainLookupEndconnectStartconnectEnd
requestStart
responseEndresponseStart
domInteractive
loadEventEnd
![Page 76: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/76.jpg)
Challenge: timestamp alignment�19
How to align timestamps at these layers?
• Domain-specific event tracing and mapping
• Machine-learning assisted
App TCP/IP LTE chipset
Base station
Web server
LTE control plane msgs
network request
LTE data plane pkts Server’s ACK
TappTtcpdump
TLTE
App log tcpdump
log MobileInsightlog
network response
![Page 77: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/77.jpg)
Pinpoint latency boSleneck in LTE: An example�20
![Page 78: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/78.jpg)
Pinpoint latency boSleneck in LTE: An example
Run a small webpage (4 KB) in Chrome on Android
• User is static, under good 4G LTE signal (-95 dBm), T-Mobile
�20
![Page 79: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/79.jpg)
Pinpoint latency boSleneck in LTE: An example
Run a small webpage (4 KB) in Chrome on Android
• User is static, under good 4G LTE signal (-95 dBm), T-Mobile
Total Latency: 473 msec
• Clicking URL → page loading complete, Steps (a)—(f)
�20
![Page 80: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/80.jpg)
Pinpoint latency boSleneck in LTE: An example
Run a small webpage (4 KB) in Chrome on Android
• User is static, under good 4G LTE signal (-95 dBm), T-Mobile
Total Latency: 473 msec
• Clicking URL → page loading complete, Steps (a)—(f)
Pinpointing the latency boZleneck
• How to breakdown?
�20
![Page 81: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/81.jpg)
Control-plane latency breakdown: local analysis I�21
![Page 82: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/82.jpg)
Control-plane latency breakdown: local analysis I
Major component from Navigation Timing API: DNS lookup, 250 ms out of 473 ms�21
![Page 83: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/83.jpg)
Control-plane latency breakdown: local analysis I
Major component from Navigation Timing API: DNS lookup, 250 ms out of 473 ms
Is the DNS server slow to handle connection?
�21
![Page 84: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/84.jpg)
Control-plane latency breakdown: local analysis I
Major component from Navigation Timing API: DNS lookup, 250 ms out of 473 ms
Is the DNS server slow to handle connection?
Fu;her breakdown: LTE service request takes 172 ms before the DNS setup
�21
![Page 85: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/85.jpg)
Control-plane latency breakdown: local analysis I
Major component from Navigation Timing API: DNS lookup, 250 ms out of 473 ms
Is the DNS server slow to handle connection?
Fu;her breakdown: LTE service request takes 172 ms before the DNS setup
�21
QueueingStalledDNS LookupInitial ConnectionRequest SentWaiting (TTFB)Content Download
8.98 ms2.97 ms250.04 ms30.11 ms0.36 ms137.41 ms43.48 ms
LTE Data Access LatencyLTE service request
![Page 86: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/86.jpg)
Data-plane latency breakdown: local analysis II
Fu;her zoom in and breakdown the remaining LTE data access latency (291 ms):�22
DNS-Wait GrantDNS (IPv6)DNS-Wait GrantDNS (IPv4)APP-OS overheadTCP SYN-Wait GrantTCP SYN-Send DataTCP ACK (local processing)HTTP GET (send request)HTTP GET-Wait GrantHTTP GET-req sentHTTP-server RTT+ DL latencyLTE-to-TCP overheadHTTP page DL transmissionHTTP DL retransmission
2.02 ms11 ms
18 ms0.02 ms
First bit of HTTP response
0.36 ms12 ms
8 ms110 ms6.1 ms
40 ms3 ms
16 ms
17 ms26 ms
12 ms
![Page 87: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/87.jpg)
Latency mapping for failures: local analysis III
Example: data plane suspension due to radio reconnection and head-of-line blocking during handover
�23
![Page 88: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/88.jpg)
Latency mapping for failures: local analysis III
Example: data plane suspension due to radio reconnection and head-of-line blocking during handover
�23
BlockingRequest SentWaiting GrantUplink TransmissionHandover Disruption — No dataHandover Disruption — Duplicate recv’d dataWaiting (TTFB, due to parallel TCP connection)Content Download
5.05 ms0.58 ms4 ms130 ms263 ms36 ms275 ms33.16 ms
![Page 89: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/89.jpg)
Machine learning scheme
We leverage domain-speciCc knowledge for ML-based predictions
Control plane: predict handover using a decision tree classiCer
• Features from 3GPP standards
• Predicts handover 100ms before it occurs with >99% accuracy
Data plane: predict NACK/ACK hip at MAC layer
�24
![Page 90: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/90.jpg)
Synthesizer: global crowdsourcing analysis�25
![Page 91: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/91.jpg)
Synthesizer: global crowdsourcing analysis
Four US carriers + Google Project Fi�25
![Page 92: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/92.jpg)
Synthesizer: global crowdsourcing analysis
Four US carriers + Google Project Fi
23 phone models, 95,057 data sessions
�25
![Page 93: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/93.jpg)
Synthesizer: global crowdsourcing analysis
Four US carriers + Google Project Fi
23 phone models, 95,057 data sessions
Overall latency: 77 — 2956 ms in 500K samples
�25
![Page 94: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/94.jpg)
Synthesizer: global crowdsourcing analysis
Four US carriers + Google Project Fi
23 phone models, 95,057 data sessions
Overall latency: 77 — 2956 ms in 500K samples
• Varies among different mobile carriers
�25
Average Latency by LTE Data Access Setup (no mobility)
0
50
100
150
200
AT&T T-mobile Sprint VerizonProject Fi
162153147165
196
![Page 95: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/95.jpg)
Synthesizer: global crowdsourcing analysis
Four US carriers + Google Project Fi
23 phone models, 95,057 data sessions
Overall latency: 77 — 2956 ms in 500K samples
• Varies among different mobile carriers
• Insensitive to varying radio link quality
�25
-130-120-110-100 -90 -80 -70 -60 -50 -40
50
100
200
500
1,000
3,000
Signal Strength (dBm)
TotalLatency(ms)
![Page 96: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/96.jpg)
LTE data access latency: how frequent?
Frequent data access setup operations
• every 58.8 sec (median); 133.6 sec (average)
• cause: frequently entering power-saving mode
Sho1-lived Radio connectivity lifetime
• every 10.8 sec (median); 17.3 sec (average)
• cause: inactivity timer (regulated by standards)
�26
![Page 97: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/97.jpg)
LTE data access latency: how frequent?
Frequent data access setup operations
• every 58.8 sec (median); 133.6 sec (average)
• cause: frequently entering power-saving mode
Sho1-lived Radio connectivity lifetime
• every 10.8 sec (median); 17.3 sec (average)
• cause: inactivity timer (regulated by standards)
�26
![Page 98: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/98.jpg)
LTE data access latency: how frequent?
Frequent data access setup operations
• every 58.8 sec (median); 133.6 sec (average)
• cause: frequently entering power-saving mode
Sho1-lived Radio connectivity lifetime
• every 10.8 sec (median); 17.3 sec (average)
• cause: inactivity timer (regulated by standards)
�26
![Page 99: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/99.jpg)
Overall latency and breakdown for major carriers�27
AT&TT-Mobile SprintVerizon Project Fi
![Page 100: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/100.jpg)
Findings Summary
Tradio: Radio connectivity setup
• It contributes 67.5 −1665.0 ms of the overall LTE access latency.
• On average, it contributes 39.7%, 44.0%, 61.9%, 64.2% and 43.7% of total latency in T-Mobile, AT&T, Verizon, Sprint and Project-Fi, respectively.
Tctrl: Connectivity state transfer
• It contributes 28.75 ms to 2286.25ms of the overall LTE access latency.
• On average, it contributes 60.3%, 56.0%, 38.1%, 35.8% and 56.3% of total latency in T-Mobile, AT&T, Verizon, Sprint and Project-Fi, respectively.
�28
![Page 101: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/101.jpg)
Impact on mobile Web app: Chrome
Average page loading time for tested webpage: 411 ms
• LTE data access setup: 174 ms
• 42.3% total latency perceived
Similar results for Safari latency on iOS
�29
DNS query TCP connection
TCP SYN
Data access request
LTE data
SYN ACK
HTTP request
LTE control plane
LTE data plane
OS overhead
HTTP transmission
Pagerendering
TCP dataTCP layer
unloadEventStartfetchStart
domainLookupStartdomainLookupEndconnectStartconnectEnd
requestStart
responseEndresponseStart
domInteractive
loadEventEnd
0 25 50 75 1000
200
400
600
800
Normalized sorted sample (%)
Latency(ms)
Total latency
LTE latency
![Page 102: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/102.jpg)
Impact on instant-messaging: WhatsApp
Average time Crst data packet being ACKed: 341 ms
• LTE data access setup: 175 ms
• 51.4% total latency perceived
�30
DNS query TCP connection
SYN
Data access request
LTE data
SYN ACK
LTE control plane
LTE data plane
OS overhead SSL Data
TCP dataTCP layer
App init
TCP ACK
SSL Data
TCP ACK
App connect w/ server
New messageidle
Server ACK
Latency
Next message
0 25 50 75 1000
200
400
600
800
Normalized sorted sample (%)
Latency(ms)
Total latency
LTE latency
![Page 103: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/103.jpg)
Discussion: reducing LTE latency
Data plane walk-arounds
• Mask the data setup latency by waking device in connected mode in advance
Control plane acceleration
• Speed up connectivity state transfer between the base station and the mobility controller (e.g. DPCM [ACM MobiCom’17])
• Handover prediction
Other issues
• Extending to other network metrics (e.g. loss, throughput, …)
• Theoretical bounds
• Privacy issues
�31
![Page 104: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/104.jpg)
Conclusion: ML-based analysis for next-gen mobile networks
Mobile networks are successful and will continue to prosper (5G, self driving, …)
Mobile network analysis: paradigm shin to device-centric, ML-based scheme
• Device-centric: unveil the tightly-guided operation issues over 4G/5G mobile networks
• Two-tiered approach: a more open solution approach for the research community
�32
![Page 105: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/105.jpg)
Q & A
![Page 106: A Machine Learning Based Approach to Mobile Network Analysisweb.cs.ucla.edu/~yuanjie.li/slides/icccn18-mobile-analytics-slides.pdf · A Machine Learning Based Approach to Mobile Network](https://reader036.vdocuments.site/reader036/viewer/2022081607/5ed816c4cba89e334c6739ab/html5/thumbnails/106.jpg)
Backup