Author: HQH
Released: March 9, 2004
Updated: N/A
|
Below are the setup of the equipments and what I am going to test. Also listed are the conditions and settings applied to the
testing.
D-Link equipments
D-Link DI-624 wireless router:
- Hardware revision: C1
- Firmware: 2.28, dated: November 3, 2003
D-Link DWL-G520 wireless desktop adapter:
- Hardware revision: A3
- Firmware: 2.36
- Driver set: 2.48
- Core driver: 2.1.3.1
D-Link DWL-G650 wireless laptop adapter:
- Hardware revision: B3
- Firmware: 2.36
- Driver set: 2.48
- Core driver: 2.1.3.1
Notes
I did not use firmware 2.37 (available on 2/26/04) that is publicly available for D-Link DI-624 wireless router due to on going
stability problems since D-Link released firmware 2.37 that superseded 2.36 that supposedly to superseded 2.28. Since my 2.28
firmware was stable and I didn't have any problems with my wireless router, I opted out upgrading the firmware based on stability
viewpoint and that the fact there wasn't a 2.28 Nov '03 version to revert back to (there was only 2.28 Oct '03 version).
Driver set version is for the entire package (driver + utility) that you can download from
D-Link's website. Core driver is the driver component. This can be located
by selecting the file ar5211.sys, right click, select properties, select version tab, and it should say File Version: x.x.x.x at
the top, or if you select File Version from Item name box under Other version information section. ar5211.sys can be found within
the driver set package.
The D-Link utility was not installed. I used Windows wireless management system (Wireless Zero Configuration) instead to have
my wireless connection configured.
Wireless Router Settings
Permanent settings are settings that do not change when performing benchmarks. Variable settings are settings that changes
throughout the benchmarks, wherever noted in the results. Settings listed below are in the order that the menus appear in the
DI-624 router configuration. Listing of the settings below are what really matters for benchmark purposes. All other settings
are omitted if they do not contribute benchmark changes:
- Home: Wireless,WAN
- Advanced: Filters, Performance
- Tools: Misc.
- Status: Log
| Permanent Settings |
| Feature | Value(s) |
| Channel | 6 |
| MTU | 1500 |
| MAC Filters | 5 MAC addresses allowed |
| TX Rate | 108 Mbps |
| Beacon interval | 100 |
| RTS Threshold | 2346 |
| Fragmentation | 2346 |
| DTIM interval | 1 |
| SSID Broadcast | Disabled |
| 802.11g Only Mode | Enabled |
| CTS Mode | None |
| Super G Mode | Super G with Static Turbo |
| UPNP Settings | Disabled |
| Gaming Mode | Enabled |
| Log | Log Type: all boxes unchecked |
| Variable Settings |
| Feature | Value(s) |
| Authentication | Open System (WEP: Disabled, 64-bit, 128-bit), WPA-PSK |
| Preamble Type | Short Preamble, Long Preamble |
Notes: In order to achieve Super G mode, the following settings took place: 802.11g Only Mode - Enabled; CTS Mode - None;
Super G Mode - Super G with Static Turbo. These settings are reflected in the above table. Settings other than those would yield
me 54 Mbps (plain 802.11g).
Computers
Below lists what computer uses which equipment with their specs and their purpose.
FTP Server: Celeron 300A Desktop
- Intel Celeron 300A overclocked to 450 MHz
- 320 MB of PC-133 SDRAM, no virtual memory (swap file)
- Abit BH6 v1.2, BIOS version SP
- Maxtor, Model: 54610H6, Capacity: 46.1 GB, Interface: UDMA/ATA-100, RPM: 7200, Cache: 2 MB, note: due to Abit BH6 motherboard,
it will only support max ATA-33 transfer mode
- Netgear FA311 10/100 Mbps PCI network card
- Windows XP Professional with SP1
- FTP Server Serv-U Personal Edition v4.1
DWL-G520: P4 2.4C Desktop
- Intel Pentium 4 2.4C overclocked to 3.0 GHz
- 1 GB of PC3500 (DDR434) DDRRAM, no virtual memory (swap file)
- Abit IC7-G, BIOS version 22
- Western Digital, Model: WD2500JBRTL, Capacity: 250 GB, Interface: UDMA/ATA-100, RPM: 7200, Cache: 8 MB
- Windows XP Professional with SP1
- FTP Server Serv-U Personal Edition v4.1 (for DWL-G520 to DWL-G650 transfer testing)
DWL-G650: Compaq Presario Notebook 17XL260
- Intel Pentium 3 500 MHz
- 256 MB PC-133 SDRAM
- Ultra DMA 6 GB HDD
- Windows XP Professional with SP1
Benchmarking Notes
The test file that I used to measure transfer rate will be a mp3 file that is about 10.0 MB in size. The exact file size is
10,577,107 bytes, or 10,579,968 bytes stored on an actual NTFS disk (retrieved from right clicking on the file, selecting
properties, information found under general tab).
House scenario is a three story house (2 floors plus basement). DI-624 wireless router is in the basement, desktop computer
containing DWL-G520 is on the 2nd floor. They are virtually vertically aligned together. Distance between them is roughly 2 floors
between each other. So I'd guess about close to 20 feet between the router and the client. For the laptop containing DWL-G650, it
will be stationed in two places: 2 feet from the wireless router in the basement for client to server transfer and then back
upstairs on the 2nd floor near desktop computer with DWL-G520 to do a client to client transfer through wireless router. Note that
client to client transfer through the wireless router is not done as ad-hoc mode. It is done to test how the entire wireless 108
Mbps bandwidth is used up. The ftp server computer is downstairs in the basement.
The desktop and laptop containing the DWL-G520 & DWL-G650 respectfully, will be uploading to the FTP server computer using
Microsoft's FTP program (C:\WINDOWS\system32\ftp.exe), in the form of MS-DOS command line prompt. The FTP program will tell how
fast the transfer rate is, which is sufficient enough for my benchmarking. There will be three transfers total and an average will
be calculated to reach a solid average value for a single test condition. Due to wireless performance fluctuation caused by radio
noise (read: interference), results may look weird, but that's why three transfers are performed and multiple tests are done. For
client to client transfer through wireless router, the laptop DWL-G650 will upload to desktop DWL-G520.
My benchmarks are analyzed with all the conditions listed on this page and explained in a detailed manner so that it can be
referenced in your decision making or for your entertainment. My performance analysis does not mean you will achieve the exact same
results. Your results highly depend on your environment and how much radio noise is in your area. It can be better than my results,
equal to my results, or less than my results. So this means that this editorial is to be taken as a reference, but not an actual
raw and true performance of the 802.11g equipments. To sum it up: performance analysis cannot be done in ideal conditions to
measure the raw and true performance of D-Link's 802.11g equipments. This editorial's analysis is as good as it gets in my location
and conditions.
Unfortunately, measuring the signal and link quality cannot be done since the D-Link utility was not installed. Microsoft's
5 bars in the wireless properties is also useless, and it'll say excellent no matter what. The D-Link utility wasn't installed due
to a couple of reasons: crashes my computer and no wireless connectivity. If I just use Microsoft's built in wireless management,
everything works fine. The opposite happens when using D-Link's utility.
There are no 2.4 GHz phones in my house nor was the microwave used at the time of testing. However, there were a few other 2.4
GHz access points detected during test time, however, those wireless access points were houses away (read: neighbors' wireless
stuff).
Test Results
|
|
|
