On the 802.11 Turbulence of Nintendo DS and Sony PSP Handheld Network Games

1
On the 802.11 Turbulence of Nintendo DS and Sony
PSP Hand-held Network Games

• Research Seminar Course
• Polyxeni Kleniati
• Department of Computing
• Imperial College London
• Monday, 27 February 2006

2
About the Author

• Mark Claypool
• Associate Professor, CS Department, Worcester
  Polytechnic Institute
• Director, WPI s Interactive Media and Game
  Development undergraduate degree

3
About the Author

• Research Methodology
• Observe, Hypothesize, Design, Experiment,
  Analyze, Report
• "The fundamental principle of science, the
  definition almost, is this the sole test of the
  validity of any idea is experiment." -- Richard
  P. Feynman
• Research Interests
• Distributed Systems, Multimedia, Performance
  Analysis, Collaborative Filtering, Perceptual
  Quality, Wireless Network and Network Games
• Publications
• Journals (9), Conferences (51), Books (2),
  Demos, Posters, Technical Reports and so on
• January 2006, his paper The Effects of Frame
  Rate and Resolution on Users Playing First Person
  Shooter Games, presented in Proceedings of
  ACM/SPIE Multimedia Computing and Networking
  (MMCN), won the best paper award

4
About the Presenter

• Polyxeni Margarita Kleniati
• 1st year PhD student
• Supervisor Professor Berç Rustem
• Group High Performance Computing (HPC)
• Research Area Computational Management Science
  (CMS)
• The research of our group is focused on decision
  making under uncertainty, with applications to
  finance, scheduling, worst-case analysis and
  chemical engineering
• My Topic Semidefinite Programming and Global
  Optimization of Polynomials Using Higher Order
  Moments

5
About the Paper

• Presented in Proceedings of the 4th ACM SIGCOM
  workshop on Network and System Support For Games
• Session Game Traffic Characterization
• Year of Publication 2005
• Publisher ACM Press, New York, USA
• Best Paper award!

6
Outline

• Introduction
• Motivation Objective of the paper
• B.I. on IEEE 802.11, Nintendo DS Sony PSP
• Methodology
• Games Selected
• Wireless Framework, Gameplay Sniffing
• Experimental Setup
• Results
• Game Phases

7
Outline

• Results
• Turbulence
• Nintendo DS vs. Sony PSP (bitrates)
• Frame Sizes (CDFs)
• Inter Frame Times (CDFs)
• Wireless Connection Quality
• Impact on TCP throughput
• Conclusions
• Discussion

8
Motivation

• Recently, the computer game industry has seen a
  substantial growth
• Computer game sales in US comparable with
  Hollywood box office
• Increase in residential broadband Internet
  connections with high capacities
• Hand-held gaming consoles were primarily
  single-player until 2004
• Spring 2005, Sony PSP and Nintendo DS released
  and support multi-player gameplay, using IEEE
  802.11 (WLAN)

9
Motivation

• This is significant for many reasons
• More more home users access the internet via a
  WLAN
• Growth in WLAN increases the likelihood that
  concurrent hosts compete on the same channels
• The traffic generated by one host on a WLAN can
  degrade the performance of other hosts on the
  WLAN
• Essential to know the network load caused by
  hand-held game traffic
• Little research has been done on hand-helds using
  802.11 protocol

10
Objective

• This paper examines the network characteristics
  for a variety of network games on the Nintendo DS
  and the Sony PSP
• The author uses the terminology turbulence to
  denote the network characteristics
• size and frequency of data sent
• overall bitrate

11
Objective Questions

• What is the network turbulence for hand-held
  network games?
• Does the network turbulence differ between PSP
  and DS?
• Does the network turbulence differ among
  different games on the same hand-held?
• Does the network turbulence for hand-held games
  differ from PC games?
• Does hand-held game traffic affect the Internet
  traffic on the same wireless channel?

12
IEEE 802.11

• IEEE 802.11 denotes a set of Wireless LAN/WLAN
  standards developed by working group 11 of the
  IEEE LAN/MAN Standards Committee (IEEE 802)
• The term IEEE 802.11 is also used to refer to the
  original 802.11, which is also called 802.11
  legacy
• The 802.11 family currently includes six
  modulation techniques that all use the same
  protocol
• The most popular techniques are those defined by
  the b, a and g amendments to the original
  standard
• The information is taken from Wikipedia, the
  free on-line encyclopedia.

13
IEEE 802.11 Protocols

• 802.11 legacy The original version of the
  standard IEEE 802.11 released in 1997 specifies
  two raw data rates of 1 Mbits/s and 2 Mbits/s to
  be transmitted in the Industrial Scientific
  Medical frequency band at 2.4 GHz
• The original standard also defines Carrier Sense
  Multiple Access with Collision Avoidance
  (CSMA/CA) as the media access method
• Legacy 802.11 was rapidly supplemented by
  802.11b. Widespread adoption of 802.11 networks
  only occurred after 802.11b was approved and as a
  result few networks run on the 802.11 standard

14
IEEE 802.11 Protocols

• 802.11b The 802.11b amendment to the original
  standard was approved in 1999. 802.11b has a
  maximum raw data rate of 11 Mbits/s and uses the
  same CSMA/CA media access method defined in the
  original standard. Due to the CSMA/CA protocol
  overhead, in practice the maximum 802.11b
  throughput that an application can achieve is
  about 5.9 Mbit/s over TCP and 7.1 Mbit/s over UDP
• 802.11b is usually used in a point-to-multipoint
  configuration, wherein an access point
  communicates via an omni-directional antenna with
  one or more clients that are located in a
  coverage area around the access point
• 802.11b cards can operate at 11 Mbit/s but will
  scale back to 5.5, then 2, then 1 Mbit/s
  (Adaptive Rate Selection), if signal quality
  becomes an issue. Since the lower data rates use
  less complex methods of enconding the data, they
  are less susceptible to corruption due to
  interference and signal degradation

15
Nintendo DS

• System
• Resolution 256192 pixels, 260k colours
• Dual ARM processors
• ARM 9 at 67 MHz
• ARM 7 at 33 MHz
• 4 Mbytes RAM
• Touch-pad
• Network
• WLAN is IEEE 802.11b
• Short preamble
• Wireless capacities of 1 Mbps or 2 Mbps
• Ad hoc mode
• No IP stack (UDP/TCP)

16
Sony PSP

• System
• Resolution 480272 pixels, 24-bit colour
• Dual MIPS R4000 processors at 333 MHz
• 32 Mbytes RAM
• Analog joystick
• Network
• WLAN is IEEE 802.11b
• 1, 2, 5.5, 11 Mbps
• Ad hoc mode and infrastructure mode
• IP stack

17
Games Selected
18
Wireless Framework

• First floor of wooden house
• Used 2 Sony PSP and 2 Nintendo DS close to each
  other (10 feet), except when deliberately taken
  further apart
• Sniffing via Ethereal (v0.10.3) on laptop
• 3 APs were visible by the sniffer, but no
  additional traffic apart from the beacons
• Games on WLAN channels 1, 6, 11

19
Gameplay Sniffing

• Sniffer able to capture the entire frame (MAC
  headers, extra physical/MAC layer information)
• Each game was played first in single player mode
• Then, two-player version of each game
• 3 runs, gather all data
• For two games, 3 wireless locations (Bad, Good,
  Excellent)
• For one game, bulk download on PC

20
Experimental Setup
Dell Inspiron 8600 Netgear WG511 802.11 b/g with
Intersil s PRISM GT chipset
21
Game Phases

• Setup
• Minimal network traffic
• Synchronization
• High bitrates
• Play
• Moderate bitrates
• Transition
• Low bitrates

22
Game Phases
23
Nintendo DS vs. Sony PSP
24
Nintendo DS vs. Sony PSP
25
Frame Sizes Cumulative Distribution Functions
Nintendo DS
Sony PSP
26
Inter Frame Times Cumulative Distribution
Functions
Nintendo DS
Sony PSP
27
Wireless Connection Quality
28
Impact on TCP throughput
29
Conclusions

• What is the network turbulence for hand-held
  network games?
• Frequent sends of small frames of data
• Typical hand-held frame sizes are less than 100
  bytes
• Typical inter frame times are less than 20 msecs
  (excluding ACKs)
• Does the network turbulence differ between PSP
  and DS?
• DS games send smaller frames that the PSP games
• Some PSP games send mostly broadcast traffic,
  while all DS games studied use direct addressing

30
Conclusions

• Does the network turbulence differ among
  different games on the same hand-held?
• Games on the Nintendo DS have network turbulence
  similar to each other
• Games on the Sony PSP vary considerably in
  bitrate, frame size, frame frequency, and
  fraction of broadcast traffic
• Does the network turbulence for hand-held games
  differ from PC games?
• Data frames sent by hand-helds are comparable in
  size to those of PC or console games
• Hand-helds send data more frequently, resulting
  in bitrates that are higher overall than typical
  PC or console games

31
Conclusions

• Does hand-held game traffic affect the Internet
  traffic on the same wireless channel?
• Some game phases may have adverse affects on
  throughput for Internet applications sharing the
  same 802.11 wireless channel

32
Additional Conclusions

• Some Sony PSP games primarily use 802.11
  broadcast as a means of communicating between the
  hand-helds
• Implications on other wireless devices that share
  the same channel
• Under poor connection quality, the Nintendo DS
  game studied sends data at higher bitrates
• Sony PSP game studied increases bitrates slightly
  and has difficulty staying connected

33
Discussion

• Is research on games legitimate? (In my opinion,
  it is!)
• The ultimate goal is to improve games, or to
  improve protocols, so improving games?
• What is the real meaning/importance of the
  influence of games on internet traffic?
• Is there any room for doubt about protocol
  802.11b being really supported in both games?

34
The End.Any Questions?

• Thank you.