doc.: ieee 802.11-13/1305r0 submission w.carney, et. al. (sony) slide 1 simplification of hew...
TRANSCRIPT
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)Slide 1
Simplification of HEW Traffic Model Simulations
Date: 2013-11-12Name Affiliations Address Phone email
William Carney Sony Electronics 16530 Via Esprillo San Diego, CA, 92127
+1 858 774 9865 william.carney @am.sony.com
Kåre Agardh Sony Mobile Nya Vattentornet 221 88 Lund, Sweden
+46 (10) 801 3618 kare.agardh @sonymobile.com
Hideyuki Suzuki Sony Corporation 2-10-1 Osaki Shinagawa-ku, Tokyo, 141-8610
+81 (50) 3750 2715
hideyukia.suzuki @jp.sony.com
Lachlan Michael Sony Corporation 4-16-1 Okata Atsugi Kanagawa, 243-0021
+81 (50) 3141 3138
lachlan.michael @jp.sony.com
Guoqing Li Intel Corporation 2111NE 25th Ave, Hillsboro, OR 97124
+1 503 712 2089 [email protected]
Laurent Cariou
Orange 4 rue du clos Courtel 35512 Cesson Sévigné France
+33 299124350
Thomas Derham
Orange 9F Keio Shinjuku Oiwake Bldg. Shinjuku 3-1-13, Tokyo, Japan
+81 3 5312 8563
Jean-Pierre Le Rouzic
Orange 4 rue du Clos Courtel 35512 Cesson Sevigne France
+33 299124893
Leif Wilhelmsson Ericsson AB Scheelevägen 23 Lund, Sweden
+46 706 216 958 leif.wilhelmsson @ericsson.com
Filip Mestanov Ericsson AB Färogatan 6, Stockholm, Sweden
+46 725 298 161 [email protected]
Authors:
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
Abstract
We discussed in 13/0728r0 and 13/1144r1 that small packets consume a relatively large part of the capacity in today’s wireless networks. [1], [3], [4]
To achieve a better representation of real usage and to simplify simulations, the Use Cases defined by HEW should be combined in simulation environments, one Use Case per STA.
Recommending annex updates to 13/1001r_ to account for these environments.
Slide 2
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
• Video is clearly the ‘killer app’ and will dominate traffic in HEW networks.
• Sony is in the business of delivering entertainment content to consumers with high QoS/QoE expectations.
• Small packets are already a reality in today’s networks and their increasing presence will have an effect on video traffic quality.– HEW simulations must accurately account for and
understand the issues.
Slide 3
Motivation
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
Review of Prior Materials: MAC/PHY Inefficiency for Small PacketsPacket Size distribution MAC Efficiency (e.g., 11a)
30% of used capacity from 5% of traffic volume64 128 512 1500 Bytes per packet
20% 15% 10% 55% % of total number of packets2.2% 2.6% 5.8% 89.4% => % of traffic volume (Bytes)
15.4% 12.5% 8.3% 63.8% => % of airtime on network
Ref [2] – Efficiency for streaming data without aggregation. Doufexi, et. al.
Ref [1] – Packet sizes in real networks, Ericsson 2012
Slide 4
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
Mixed Usage in a Network is CommonHEW simulations should model this usage
Example: Expected traffic within each network in dense apartment building (3a)
Streaming to living room Television
Kid: PlaystationNetwork gaming
Dad: Paying bills and browsing Kid: Social
networking
Mom: Video Skype with kid away from home
Slide 5
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
• With the number of defined scenarios, there is the possibility of creating very complex traffic patterns.
• Option 1: Simplify by deciding on traffic mix (constant packet size distribution) for each device. All STAs have the same mix.- Cons: i) Ignores differences between users/use cases; ii) Over-optimistic in data
packet aggregation opportunities.- Pros: Simulation traffic model complexity is simplified.
• Option 2: Simplify by defining sets, each with one packet size indicative of a main usecase. Distribute these sets between STAs.- Cons: i) Ignores background traffic from other simultaneous usecases; ii) Ignores
real combined usecases.- Pros: Simulation traffic model complexity is simplified.
- Option 3: Combination of 1 and 2 by defining a set of traffic mixes that each refer to one main UC. Distribute these between STAs.- Cons: Very complex simulations.- Pros: Most representative of realistic HEW network usage.
Slide 6
How to Account for Small Packets in HEW Simulation Environments?
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
Recommended Simplified Simulations
• Traffic mix to be simulated with one traffic type per STA.– E.g. in Dense Apartment 3a.
• UL video STA.• DL video STA.• Browsing Traffic.• Video call.• Gaming.
• Distribution of use cases to be defined per Scenario.– Distribution shall approximate real traffic as seen from available
network statistics.
Slide 7
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
• The recommendation and conclusion in previous slides complement the content of the draft test case template pioneered by Qualcomm in 13/1001.
• One simplified traffic definition to be included in each traffic profile T1…Tn. Propose to add a clear statement of the one most relevant MAC packet size per UL/DL.
• Example of content for each traffic profile definition:
• Combination of usecases within a network achieved by defining multiple STAs with different traffic profiles.
Proposal for Additions to Test Case Template
ref [5] : IEEE 802.11-13/1001r2
Traffic Model #
Traffic model name Description Application traffic(Forward / Backward)
Application Load (Mbps)(Forward / Backward)
Max A-MPDU Size (B)(Forward / Backward)
T1 Local file transfer FTP/TCP transfer of large file within local network
FTP file transfer / FTP TCP ack
Full buffer/ 0.1
4k / 64
Slide 8
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
• A mixed use case simulation environment allows HEW network behavior to be clearly understood, especially the distribution of packets sizes and their effects.
• The purpose is to account for the effects of all traffic types and packet sizes on the most important traffic – video.
• By accounting for such traffic, HEW can better achieve its primary goal of improving network efficiency.
Slide 9
Summary
doc.: IEEE 802.11-13/1305r0
Submission W.Carney, et. al. (SONY)
References
[1] Ericsson measurement, 2012/Q4, smartphone-dominated mature LTE/HSPA/2G network
[2] A. Doufexi, S. Armour, P. Karlsson, M. Butler, A. Nix, D. Bull, J. McGeehan, “A Comparison of the HIPERLAN/2 and IEEE 802.11a Wireless LAN Standards,” IEEE Communications Magazine, May 2002, Vol. 40, No. 5.
[3] Network optimizations for expected traffic patterns. IEEE July 2013 11-13-0728-00-0hew-Network-optimization-for-expected-traffic-patterns_v01
[4] Simplified Traffic model, IEEE September 2013, 11-13-1144-01-0hew-simplified-traffic-model-based-on-aggregated-network-statistics
[5] 11-13-1001-02-0hew-simulation-scenarios-document-template
Slide 10