Scalability of Future Mobile Networks Through Network Independent - - PowerPoint PPT Presentation

Scalability of Future Mobile Networks

Through Network Independent Application Layer Mobility and Context Monitoring

Florian Metzger 2015/12/10

Modeling of Adaptive Systems htups://www.mas.wiwi.uni-due.de/en

Motivation

QoS and QoE of TCP Streaming in Mobile Networks?

5000 10000 15000 2014 2015 2016 2017 2018 2019

year traffic (exabytes per month)

traffic type video web, data, and voip filesharing audio streaming

source: Cisco VNI 2014

Interactions and Optimizations?

Modeling of Adaptive Systems 2015/12/10 2/13

Motivation

QoS and QoE of TCP Streaming in Mobile Networks?

5000 10000 15000 2014 2015 2016 2017 2018 2019

year traffic (exabytes per month)

traffic type video web, data, and voip filesharing audio streaming

source: Cisco VNI 2014

UTRAN RNC Node B Core Gi SGSN GGSN Uu Iub Iu Gn BSS BSC BTS Abis E-UTRAN eNB Mobile Devices Um MSC HSS X2 Uu Iur A IuCS CS-MGW PSTN PSTN Mc S-GW P-GW MME S1-U S1-MME S3 S11 S4 S5 SGi C D Nc Gr Gc S6a PCRF Gx S9 Gx Gxc GMSC

Interactions and Optimizations?

Modeling of Adaptive Systems 2015/12/10 2/13

Motivation

QoS and QoE of TCP Streaming in Mobile Networks?

5000 10000 15000 2014 2015 2016 2017 2018 2019

year traffic (exabytes per month)

traffic type video web, data, and voip filesharing audio streaming

source: Cisco VNI 2014

UTRAN RNC Node B Core Gi SGSN GGSN Uu Iub Iu Gn BSS BSC BTS Abis E-UTRAN eNB Mobile Devices Um MSC HSS X2 Uu Iur A IuCS CS-MGW PSTN PSTN Mc S-GW P-GW MME S1-U S1-MME S3 S11 S4 S5 SGi C D Nc Gr Gc S6a PCRF Gx S9 Gx Gxc GMSC

Interactions and Optimizations?

Modeling of Adaptive Systems 2015/12/10 2/13

Interaction of Complex Systems

Mobile Nets vs TCP Streaming

Control plane is communicated explicitly by the network entities and baseband All user traffic is encapsulated into tunnels Mobility support, signaling and anchors in the net Other, off-path entities hold and communicate more state (e.g. MME, HSS, PCRF) Fixed protocol stack (TCP+HTTP) distinguished only through buffering and quality level adaptation strategies TCP and TCP-streaming not designed with mobility in mind Impact of mobile nets and signaling

• n QoS and QoE?

Modeling of Adaptive Systems 2015/12/10 3/13

Interaction Scenario #1

Segment Retrieval and High Latency

30000 60000 90000 120000 100 1000

additional latency (ms) video stall duration (s)

low quality standard quality high quality 50 100 150 100 1000

additional latency (ms) number of video stalls

low quality standard quality high quality

HQ video did not complete in time at higher latencies Demonstrates the effects of a bad streaming strategy:

New segments were only requested when the previous one had arrived → Stop-and-wait behavior, overly sensitive to latency In this case a simple change helps: Request new segments ahead of time to ensure back-to-back transmission

High latency (and variation) is a common feature

• f mobile nets!

→ Need for testing to find other unexpected interactions!

Modeling of Adaptive Systems 2015/12/10 4/13

Interaction Scenario #2

High Bandwidths and Buffering

2000 4000 6000 10 1000

bandwidth (Mb/s) video stall duration (s)

low quality standard quality high quality 30 60 90 10 1000

bandwidth (Mb/s) number of video stalls

low quality standard quality high quality

Stalls at higher bandwidths Suspiciously appears above the radio capacity (≈ 80 Mbit

s )

Possible explanations:

Excessive buffering in the net Negative interaction of TCP and HARQ Packet buffers fill, nothing will be dropped, TCP won’t back down → Bufferbloat Further investigation needed

Modeling of Adaptive Systems 2015/12/10 5/13

Interaction Scenario #3

Device Mobility

eNodeBs UE SGW/ PGW streaming server X2 d d/2 d/2 y

→ Low buffer and stall events at further distances Streaming players need to be aware

• f drops in connectivity

Modeling of Adaptive Systems 2015/12/10 6/13

Mobile Streaming and Context Factors/Monitoring

Extend the handover blackout to other events of variable lengths, e.g.:

Horizontal and vertical handovers Areas with low radio coverage and insufficient throughput (Car) traffic tunnels Subway, metro traffic and tunnels, etc.

Does network-assisted mobility help or hinder in such scenarios for streaming? Could application-layer mobility in conjunction with context-monitoring even provide similar or betuer results?

Mobility as context factor Requires good predictors, e.g. by deriving information from past patuerns Provide an interface with every available context information to applications so that they can conduct appropriate QoE optimizations themselves Passing information up the protocol stack, no network assistance

Modeling of Adaptive Systems 2015/12/10 7/13

“Tunnel” Scenario

Context-Based Solution

Use context and context predictors in adaptive streaming strategies Compute optimal context-based buffering and quality level selection strategy to ensure best QoE

Prevent stalling, but still do not excessively buffer ahead Optimize segment quality level while still avoiding stalling

Knowledge of upcoming event through context, advance time is critical

Stalling Tunnel end Buffered video playtime Real time Time spent to download a segment (low layer) Time to play any segment (fixed) Tunnel warning notification (tadv)

Context- aware HAS Standard HAS

Proactive lower layer selection No stalling (high layer) Tunnel start

Low quality High quality segment

100 200 300 400 500 600 700 800 50 100 150 200 250

Time (sec) Buffer playtime (sec)

Context−aware HAS Standard HAS Optimal HAS

Tunnel Start Tunnel End

Modeling of Adaptive Systems 2015/12/10 8/13

Network Independent Application-Layer Mobility

Through Context Monitoring

“Tunnel” scenario easily transferable from TCP video streaming to other applications Is mobility support really necessary for many applications today?

Web/HTTP traffic consists of many small objects, could easily completely forgo network-assisted mobility and just reorder/schedule around VoIP and other real-time communication: More tricky, but, e.g. adapt existing over-the-top mobility solutions (SIP proxy)

→ Streamlining for future mobile architectures

Remove global mobility support Provide a trimmed down architecture with only the essentials Increase scalability/performance by removing most control plane procedures, just provide a bare-minimum bit-pipe access Solve remaining issues and provide missing features over-the-top

Modeling of Adaptive Systems 2015/12/10 9/13

Streamlining for Future Mobile Architectures

High-Level Abstract Architecture

Access Provider Nets

3GPP base station

Internet Access

• ptional

mobilty link WiFi AP cloud context service

Internet

3GPP base station 3GPP base station 3GPP base station

Modeling of Adaptive Systems 2015/12/10 10/13

Streamlining for Future Mobile Architectures

Application-layer Mobility Example for Handover

mobile device apps 5G transceiver #0 5G transceiver #1 5G transceiver #2 wifi AP #0 cloud context service

• ther users

connected nodes in range probing context request context upload node selection early handover notification mobility connect connected disconnect context upload and request probing node selection context request probing node selection context request

Modeling of Adaptive Systems 2015/12/10 11/13

Summary

Much potential for (negative) interactions and feedback loops between mobile signaling and (amongst others) streaming

E.g. stalling events during handover Requires good understanding and deep investigation

Applications could orchestrate their own mobility using context monitoring Mobility scenarios and mobility prediction merit further investigation Goal: Reduce network complexity, increase scalability by moving exiting network features and relinquish control to the application layer

Modeling of Adaptive Systems 2015/12/10 12/13

Thanks!

• E. Liotou, T. Hoßfeld, C. Moldovan, F. Metzger, D. Tsolkas, and N. Passas.

“Enriching HTTP Adaptive Streaming with Context Awareness: A Tunnel Case Study”. In: ICC (2016). Under review, submitued in October 2015.

• F. Metzger, E. Liotou, C. Moldovan, and T. Hoßfeld.

“TCP Video Streaming and Mobile Networks: Not A Love Story, But Betuer With Context”. In: Computer Networks Special Issue on Big Data (2016). Under review, submitued in October 2015.

• F. Metzger. “Evaluating Reliable Streaming in Mobile Networks”. PhD thesis. Apr. 2015.
• F. Metzger, C. Steindl, and T. Hoßfeld. “A Simulation Framework for Evaluating the QoS and

QoE of TCP-based Streaming in an LTE Network”. In: Teletraffic Congress (ITC 27), 2015 27th International. Sept. 2015, pp. 168–176.

Qvestions!

Modeling of Adaptive Systems 2015/12/10 13/13