peer-to-peer aoi voice chatting for massively multiplayer online games (p2p-nve 2007 workshop)...

23
Peer-to-Peer AOI Voice Chatting for Massively Multiplayer Online Games (P2P-NVE 2007 workshop) Jehn-Ruey Jiang and Hung-Shiang Chen Adaptive Computing and Networking Lab Dept. of CSIE, National Central University, Taiwan

Upload: angelica-wilson

Post on 30-Dec-2015

213 views

Category:

Documents


0 download

TRANSCRIPT

Peer-to-Peer AOI Voice Chatting for Massively Multiplayer Online Games

(P2P-NVE 2007 workshop)

Jehn-Ruey Jiang and Hung-Shiang Chen

Adaptive Computing and Networking LabDept. of CSIE, National Central University, Taiwan

Adaptive Computing and Networking Lab, CSIE, NCU

Communications in MMOGs

Massively Multiplayer Online Games (MMOGs) are growing rapidly World of Warcraft: 9 million subscribers 500,000 concurrent users

Players often type and read text instead of speaking and listening

Adaptive Computing and Networking Lab, CSIE, NCU

Player-initiated solutions

Teamspeak, Ventrilo, SkypeSet your own voice serverGroup chat for 5 (skype) to ~10 (teamspeak)

But…Need to set up own server (s)Support only limited, fixed user groupExplicit / preconfigured joining required

Adaptive Computing and Networking Lab, CSIE, NCU

Voice chat with fixed membership

A

B

D

C

E

Adaptive Computing and Networking Lab, CSIE, NCU

A better scenario…

AOI voice chatting Each player listens and talks to all users within its AOI

(area of interest) More natural More realistic

Our proposal If neighbors are known Talk to them directly

Adaptive Computing and Networking Lab, CSIE, NCU

Voice chat with dynamic membership

A

Adaptive Computing and Networking Lab, CSIE, NCU

A simple approach

NimbusCast: MMOG system first provides a list of AOI neighbors individual voice packets for all AOI neighbors

Adaptive Computing and Networking Lab, CSIE, NCU

But… bandwidth overload occurs

Number of supportable users very limited 256 kbps / 16 kbps = 16 users (theoretical maximum)

Adaptive Computing and Networking Lab, CSIE, NCU

Human conversation model

Conversations are made of short bursts of voices Talkspurts: single talk & double talk Pauses: gaps in speech or mutual silence

A person only talks 40% of time Talking rate of a listener is 6%

[ITU-T, 2003]

Adaptive Computing and Networking Lab, CSIE, NCU

Our proposal

Observation: idle bandwidth can be used to forward voice packets, supporting more users

Under two goals: Reasonable delay [ITU-T P.59]

150~250ms (good) 400ms (tolerable)

No bandwidth overload:Voice bursts should not exceed bandwidth limit

Adaptive Computing and Networking Lab, CSIE, NCU

QuadCast

Quadrant-based Forwarding A list of AOI neighbors is first obtained from the MMOG system Voice packets sent to forwarding assistants (FAs) FAs then forwards to remaining AOI neighbors in each quadrant A recipient list is attached to the packet, and continuously divided

Adaptive Computing and Networking Lab, CSIE, NCU

Player clustering in MMOGs

In MMOG, players usually group at some hot-spots

End-to-end latency in some quadrants may be long

Adaptive Computing and Networking Lab, CSIE, NCU

SectorCast

Similar to QuadCast except in player grouping Players are grouped into four sectors with equal sizes

according to polar angles

Adaptive Computing and Networking Lab, CSIE, NCU

Packet aggregation

Different packets might be sent to the same target: Header sharing (HS) Voice mixing (VM)

Adaptive Computing and Networking Lab, CSIE, NCU

Evaluation

Simulation parameters 200 ~ 1000 nodes (in 200 increments, i.e. 8 ~ 40 AOI neighbors) 1000 x 1000 area AOI radius: 100

40% talking, 60% silent Random movement model (10 units / step) 1000 time-steps of 100ms intervals

40 bytes header 100 bytes voice 256 kbps upload

Adaptive Computing and Networking Lab, CSIE, NCU

Alternatives to recipient list

The recipient list incurs bandwidth overhead Trade communication with computation

Append just the ID of the current FA to next FA Next FA determines forwarding area based on

position & AOI of the speaking node

SectorCast requires begin & end angles as well

Adaptive Computing and Networking Lab, CSIE, NCU

Total upload bandwidth (unlimited b/w)

Adaptive Computing and Networking Lab, CSIE, NCU

Total upload bandwidth (limited b/w)

Adaptive Computing and Networking Lab, CSIE, NCU

Dropping rate (limited bandwidth)

Adaptive Computing and Networking Lab, CSIE, NCU

Average latency (in hop counts)

Adaptive Computing and Networking Lab, CSIE, NCU

Comparisons

Latency Drop rate Computation Scenarios

NimbusCast Low HighLow

(none)Low density

QuadCast High MediumMedium

(grouping)Uniform

SectorCast Medium LowHigh

(sorting)Clustering

Adaptive Computing and Networking Lab, CSIE, NCU

Conclusion

AOI voice chatting proposed for MMOGs

Idle bandwidth & human behavior exploited Lower dropping rate & bandwidth overloading QuadCast: suitable in uniform distributions SectorCast: suitable in clustered distributions

Usable with both client-server or P2P MMOGs

Adaptive Computing and Networking Lab, CSIE, NCU

Q & A

Thank you!