Switchboard: A Matchmaking System for Multiplayer Mobile Games - - PowerPoint PPT Presentation

Switchboard: A Matchmaking System for Multiplayer Mobile Games

Justin Manweiler, Sharad Agarwal, Ming Zhang, Romit Roy Choudhury, Paramvir Bahl

ACM MobiSys 2011

Battling demons and vampires

• n your lunch break…

Breakthrough of Mobile Gaming

2


Windows
Phone
7
 Top
10+
apps
are
games
 John Carmack (Wolfenstein 3D, Doom, Quake)…

“multiplayer in some form is where the breakthrough, platform-defining things are going to happen in the mobile space”

iPhone
App
Store
 350K
applica,ons
 20%
apps,
80%
downloads


47%
Time
on

 Mobile
Apps
 Spent
Gaming


Mobile Games: Now and Tomorrow

3


Increasing
Interac6vity


Single‐player
 Mobile
 
 (mobile
today)

 Mul6player

 Turn‐based
 
 (mobile
today)

 
 Mul6player

 Fast‐ac6on
 
 (mobile
soon)


Key Challenge

4


Game
Type Latency
Threshold First‐person,
Racing ≈
100
ms Sports,
Role‐playing ≈
500
ms Real‐,me
Strategy ≈
1000
ms

Challenge:
find
groups
of
peers
 than
can
play
well
together
 Bandwidth
is
fine:
250
kbps

 to
host
16‐player
Halo
3
game


Delay
bounds
are

 much
6ghter


The Matchmaking Problem

5


Match
to
sa+sfy
total
 delay
bounds


End-to-end Latency Threshold


Clients
 Connection Latency


Instability in a Static Environment

6


150
 170
 190
 210
 230
 250
 270
 290
 310
 9:36
 9:50
 10:04
 10:19
 10:33
 10:48
 11:02
 11:16
 11:31
 11:45
 12:00


Median
Latency
(ms)
 Time
of
Day
(AM)


Due
to
instability,
 must
consider
latency
distribu6on


End-to-end Latency over 3G

7


0
 0.2
 0.4
 0.6
 0.8
 1
 0
 100
 200
 300
 400
 500
 600


Empirical
CDF
 RTT
(in
ms)


AT&T
to
AT&T
Direct
 AT&T
to
AT&T
via
Bing
 AT&T
to
AT&T
via
Duke
 AT&T
to
AT&T
via
UW


First‐person
Shoot.
 Racing


Real‐6me
Strategy


Sports


Peer‐to‐peer
reduces
latency

 and
is

cost‐effec6ve


The Matchmaking Problem

8


Link
Performance
 P2P
Scalability
 Grouping
 Targe,ng
3G:
 play
anywhere


Latency
not
Bandwidth


interac+vity
is
key


Measurement
/
PredicTon


at
game
+mescales


Requirements for 3G Matchmaking

• Latency estimation has to be accurate
• Or
games
will
be
unplayable
/
fail

• Grouping has to be fast
• Or
impa,ent
users
will
give
up
before
a
game
is
ini,ated

• Matchmaking has to be scalable
• For
game
servers

• For
the
cellular
network

• For
user
mobile
devices


9


State of the Art

• Latency estimation
• Pyxida,
stable
network
coordinates;
Ledlie
et
al.
[NSDI
07]

• Vivaldi,
distributed
latency
est.;
Dabek
et
al.
[SIGCOMM
04]

• Game matchmaking for wired networks
• Htrae,
game
matchmaking
in
wired
networks;


Agarwal
et
al.
[SIGCOMM
09]


• General 3G network performance
• 3GTest
w/
30K
users;
Huang
et
al.
[MobiSys
2010]

• Interac,ons
with
applica,ons;
Liu
et
al.
[MobiCom
08]

• Empirical
3G
performance;
Tan
et
al.
[InfoCom
07]

• TCP/IP
over
3G;
Chan
&
Ramjee

[MobiCom
02]


10


Latency
es,ma,on
and
matchmaking
are

 established
for
wired
networks


A “Black Box” for Game Developers

11


Internet


IP
network
 RNC
 SGSN
 RNC
 SGSN
 GGSN
 GGSN


Link
Performance


(over
6me)
 End‐to‐end
 Performance
 “Black

 Box”


Crowdsourced
 Measurement


Latency Similarity by Time


Crowdsourcing 3G over Time

12


Time


Crowdsourcing 3G over Space

13


Latency Similarity by Distance


Can we crowdsource HSDPA 3G?

• How does 3G performance vary over time?
• How
quickly
do
old
measurements
“expire”?

• How
many
measurements
needed
to
characterize
the


latency
distribu,on?


• …

• How does 3G performance vary over space?
• Signal
strength?
Mobility
speed?


• Phones
under
same
cell
tower?

• Same
part
of
the
cellular
network?

• …


14


Details
of
parameter
space
lem
for
the
paper

 (our
goal
is
not
to
iden6fy
the
exact
causes)


Methodology

• Platform
• Windows
Mobile
and
Android
phones

• HSDPA
3G
on
AT&T
and
T‐Mobile

• Carefully deployed phones
• Con,nuous
measurements

• Simultaneous,
synchronized
traces
at
mul,ple
sites

• Several locations
• Princeville,
Hawaii

• Redmond
and
Seanle,
Washington

• Durham
and
Raleigh,
North
Carolina

• Los
Angeles,
California


15


Stability over Time (in a Static Environment)

16


0
 0.2
 0.4
 0.6
 0.8
 1
 120
 140
 160
 180
 200
 220
 240


Empirical
CDF
 RTT
(Msec)


Redmond,
AT&T,
15m
Intervals


Black
line
represents
phone
1

 (all
other
lines
phone
2)


Similar
latencies
under
 the
same
tower
 Performance
drims
over
 longer
,me
periods
 Live
characteriza,on
is
necessary
and
is
feasible


Stability over Space (at the same time)

17


0
 0.2
 0.4
 0.6
 0.8
 1
 0
 50
 100
 150
 200


Empirical
CDF
 RTT
difference
at
90th
percenTle
(ms)


S‐home
 Latona
 U
Village
 Herkimer
 Northgate
 1st
Ave


Similarity
at
the
 same
cell
tower
 Divergence
between
 nearby
towers
 Substan,al
 varia6on


Switchboard
Cloud
Service


• n
MSFT
Azure


Switchboard: crowdsourced matchmaking

18
 Game


Network
TesTng
 Service


Latency
 Data
 Latency
 EsTmator
 Measurement
 Controller


Grouping
Agent


Scalability through Reuse…

• Across Time
• Stable
distribu,on
over
15‐minute
,me
intervals

• Across Space
• Phones
can
share
probing
tasks
equitably
for
each
tower

• Across Games
• Shared
cloud
service
for
any
interac,ve
game


19


Client Matchmaking Delay

20


0
 0.2
 0.4
 0.6
 0.8
 1
 0
 100
 200
 300
 400
 500
 600
 700


Empirical
CDF
 Time
unTl
placed
in
group
(s)


Total
1
client/sec
 Total
10
clients/s
 10
client
arrival/sec
 1
client
arrival/sec
 Switchboard
clients
benefit
 from
deployment
at
scale


Conclusion

• Latency: key challenge for fast-action multiplayer
• 3G latency variability makes prediction hard
• Crowdsourcing enables scalable 3G latency estimation
• Switchboard: crowdsourced matchmaking for 3G

21


k
you!


cs.duke.edu/~jgm
 jgm@cs.duke.edu