Measuring The Latency of Cloud Gaming Systems

Post on 18-Dec-2014

201 views 1 download

description

Cloud gaming, i.e., real-time game playing via thin clients, relieves players from the need to constantly upgrade their computers and deal with compatibility issues when playing games. As a result, cloud gaming is generating a great deal of interest among entrepreneurs and the public. However, given the large design space, it is not yet known which platforms deliver the best quality of service and which design elements constitute a good cloud gaming system. This study is motivated by the question: How good is the real-timeliness of current cloud gaming systems? To address the question, we analyze the response latency of two cloud gaming platforms, namely, OnLive and StreamMyGame. Our results show that the streaming latency of OnLive is reasonable for real-time cloud gaming, while that of StreamMyGame is almost twice the former when the StreamMyGame server is provisioned using an Intel Core i7-920 PC. We believe that our measurement approach can be generally applied to PC-based cloud gaming platforms, and that it will further the understanding of such systems and lead to improvements.

transcript

Kuan-Ta Chen1, Yu-Chun Chang12, Po-Han Tseng1, Chun-Ying Huang3, and Chin-Laung Lei2

1Academia Sinica 2National Taiwan University 3National Taiwan Ocean University

Measuring The Latency of Cloud Gaming Systems

Conclusion We proposed a general methodology to measure the latency

components of cloud gaming systems

OnLive provides shorter latency for real-time gaming than SteamMyGame

Future work

Improve the proposed methodology

Apply the methodology to more platforms and derive general guidelines for designing quality cloud gaming systems

Cloud Gaming

Motivation Game software becomes more complex

The overhead of setting up a game is significant

Games have software and hardware compatibility with computers

Real-time game playing via thin clients

Advantages of cloud gaming

No set up overheads or compatibility issues

People can play games anywhere, anytime

Cloud gaming services

OnLive, StreamMyGame, GAIKAI, G-cluster Global, OTOY, T5-Labs

Which platform provides the best quality of service?

Measurement Methodology Response delay : In each experiment, send ESC key in the game and see

when the corresponding MENU frame will be displayed

Response delay (RD) = Network delay (ND) + Processing delay (PD) + Playout delay (OD)

Scatter Plot of tblock_succeed and tblock_failed

Measurement Results

How to determine the occurrence of t3

Blocking is successful : add tblock to the set tblock_succeeded

Blocking is failed : add tmenu to the set tblock_failed

We then estimate t3 as the point that yield the minimum sum of the

two density functions formed by tblock_succeeded and tblock_failed

Blocking : Block the incoming data on the client

Blocking is successful

Blocking is failed

Experiment Setup

• Game: Lego Batman: The VideoGame (Batman) Warhammer 40,000: Dawn of War (DOW) F.E.A.R. 2: Project Origin (FEAR)

• Platform: OnLive (OnLive) StreamMyGame (SMG)

Network delay (ND) = network RTT

Processing delay (PD) = t3 – t0 – ND

Playout delay (OD) = t4 – t3

tblock tblock+1 Time tmenu

Menu screen appears

tblock tblock+1 Time tmenu

Internet

LAN

LAN

Switch

Client (Intel Core i7-920)

Server (Intel Core i7-920)

OnLive Server

t0 t1 t2 t4

(controllable)

t3

(probable) (observable)

Processing delay Playout delay

Probing range

Time

Academia Sinica

Server Client

Game Servers

PC

Laptop

Mobile

Streaming

Streaming

Streaming

Internet

t0 (Key event sent) t1 (Key event received)

t2 (Frame sent) t3 (Frame received)

Menu frame

Menu screen shown

t4 (Frame displayed)