As part of a broader organisational restructure, data networking research at Swinburne University of Technology has moved from the Centre for Advanced Internet Architecture (CAIA) to the Internet For Things (I4T) Research Lab.

Although CAIA no longer exists, this website reflects CAIA's activities and outputs between March 2002 and February 2017, and is being maintained as a service to the broader data networking research community.

This page is part of the GENIUS project.

Quake 3 Packet and Traffic Characteristics

CAIA Technical report 021220A, December 2002

Mark Pozzobon


This preliminary investigation of Quake 3 in-game network traffic aims to report upon the inter-packet arrival times, packet size distributions, packet per second rates and server/client data rates. Traffic characteristics of network games such as Quake 3 are particularly useful to Internet Service Providers (ISPs) who are interested in identifying such applications utilising their resources. The results of this investigation will form the basis for further research into player sensitivity to jitter and latency.

A game sequence is generally conducted as follows:

  1. Player joins
  2. Game is quiet for a few seconds
  3. Map loads
  4. Interactive-phase (in-game)
  5. Game ends
  6. Short pause
  7. Repeat from step 3. until player leaves

Although the joining, leaving and idle phases of networking gaming are important to any game play, the most interesting and crucial part of the game cycle is the traffic during the interactive phase. In the results of this investigation we have attempted to eliminate histograms containing traffic of idle clients.

Generating and Capturing Quake 3 Traffic

Game traffic was captured in a LAN environment with typically 2 - 5 players being 1 hop away from the game server (approx. 2ms). This paper looks at game play traffic over a 2 day period, although games were played over two months with largely similar results.

Quake 3 Client Computer Configurations

Computer Details
Client A B C D E
Brand Compaq Compaq Compaq Compaq Dell
Model EVO D500 EVO D500 EVO D500 EVO D500 Dimension
CPU Pentium 4 1.6MHz Pentium 4 1.6MHz Pentium 4 1.6MHz Pentium 4 1.6MHz Pentium 4 1.5MHz
RAM 256Mb 256Mb 256Mb 256Mb 256Mb
Graphics Card 16Mb Nvidia Vanta/Vanta LT 16Mb Nvidia Vanta/Vanta LT 16Mb Nvidia Vanta/Vanta LT 32Mb Nvidia GeForce2 MX/MX 16Mb ATI RAGE 128 Ultra
OS Microsoft Windows 2000 Microsoft Windows 2000 Microsoft Windows 2000 Microsoft Windows 2000 Microsoft Windows 2000

Please note that the Default Quake 3 System Setup (Graphics, Display, etc.) settings.

Equipment used for generating/capturing Quake 3 traffic

Quake 3 Equipment Setup

Each host computer is under 2-5ms from the server.
Game Server/Packet Sniffing Computer (
 Brand: Compaq
 Model: EVO D500
 CPU: Pentium 4 1.6MHz
 RAM: 256Mb
 OS: FreeBSD 4.5

 Ethernet Port Details (fxp0)
 Brand: Intel
 Model: Pro/100 Ethernet
 Ethernet address: 00:08:02:3B:5B:71

Quake 3 (v1.31) and Pkthisto 0.1.2 software running on the Game Server/Packet Sniffing Computer. The Game Server for Quake 3 was configured to operate on port 27963 and used the following map cycle: runtfest, coralctf, platform6.

The configuration settings used for Pkthisto can be found in the following file. Each histogram represents 2000 packets of traffic flow. A flow is only recorded if it saw more than 200 packets, and packets arrived more frequently than once every 800 milliseconds.


The data set used for analysis was collected over a two day period (31/07/02 - 01/08/02). Three game playing sessions were observed, where 3, 4 and 5 players participated.

Summary and Further Analysis

A Quake 3 Arena game server sends updates to each of the clients every 50msec on average, where updates are N packets back-to-back, one packet to each active client. Generally speaking for aggregate server to client traffic, (100*(N-1)/N)% of intervals will be less than or equal to 100usec and the remaining will be scattered closely around 50msec.

The table below is a summary of the inter-packet arrival times, packet size distributions, packet per second rates and server/client data rates observed from Quake 3 in-game network traffic.

  Aggregate Traffic Individual Client Traffic
Server to Client Client to Server Server to Client Client to Server
Inter-packet Arrival Times Many back-to-back packets (<100usec) and 50msec.

Dependant upon number of players

100usec - tens of milliseconds.

Upper limit dependant upon number of players.

Mean: 50msec 10 - 70msec
Packet Lengths Mean: 100 bytes

90% of packets within 50 - 200 bytes

Mean: 60 bytes

90% of packets within 57 - 70 bytes

Mean: 100 bytes

90% of packets within 65 - 170 bytes

Mean: 65 bytes

90% of packets within 57 - 70 bytes

Packets per second 20 - 100 pps 36 - 265 pps 20 pps 30 - 50 pps (Clients A - C, E)

82 - 90 pps (Client D)

Data Rate 10 - 90 kbit/sec 15 - 135 kbit/sec 12 - 19 kbit/sec 14 - 26 kbit/sec (Clients A - C, E)

40 - 46 kbit/sec (Client D)

In the table above, all of the figures obtained for Packet per second and Data rates were dependant upon the number of players simultaneously participating. The lower values generally represent one client participating and the upper values are for 5 clients.

From the aggregate results, the table below lists some formulae for relationships to express the the average aggregate packet per second and data rates given the number of players (N) participating in a game.

Aggregate Traffic Relationships
Packets Per Second
Data Rate
Server to Client
Client to Server
Server to Client
Client to Server
PPSCS = 57N - 21
RATESC = 20N - 10
RATECS = 30N -15

Our next steps involve analysing the effect of jitter, packet loss and link latency on game play and the sensitivity of each. Also, to extend the above results to include players and traffic statistics from the Internet. Finally, we wish to develop model(s) for the traffic seen. As a result, it will provide the ability to identify and monitor game traffic on a network and the ability to deliver the correct Quality of Service for game play.


Swinburne Homepage Site Map Search Index

Swinburne Copyright and disclaimer Privacy Feedback

Last Updated: Friday 20-Dec-2002 12:48:44 EST
Maintained by: Grenville Armitage
Authorised by: Grenville Armitage

IndexSearchSite MapSwinburne Home Page