Take a MIMO: WLAN ubiquity at last

Anybody who's used a wireless LAN (WLAN) for even a little while knows there's a big difference between the theoretical maximum speed of the WLAN, and the actual speed you get. Factors such as interference, congestion, transmission power, building materials, and even the weather can all play havoc with WLAN throughput in often unpredictable ways.

Aiming to improve the situation, WLAN vendors now offer both conventional equipment and MIMO (Multiple Input Multiple Output) gear, which combines several antennae to boost signal strength and equipment sensitivity. Vendors say MIMO delivers longer range, faster speeds and better penetration of objects -- but how much of a difference does it really make in practice? We decided to find out.

The test set up
Using our test NEC Versa 4400 notebook PC with various combinations of MIMO (Netgear RangeMax WPN511 PC Card and RangeMAX WPN824 access point) and non-MIMO equipment (Netgear WG511 PC Card and Netgear WGR614 base station), we benchmarked a standard, unencrypted WLAN at different points around a standard-sized suburban Melbourne house. Base stations were connected directly to an Optus cable modem Internet service. Test sites included the back bedroom, where the WLAN access point was installed; the kitchen; the living room; the front bedroom; the front yard; the back yard; and the rear garage. Each location was a different distance from the WLAN base station, and required that the WLAN carrier pass through a different combination of building materials.

At each location, signal strength data was collected from the Netgear software's strength monitor; Windows XP's connection monitor; and Passmark Wirelessmon, which measures signal strength in dB and percentage of optimum. Real-world performance was tested by running three online bandwidth testers (CNET Bandwidth Meter Speed Test, Numion, and PCPitStop), three times each, then averaging the results.

Each set of tests was run four times at each location, allowing for four combinations of MIMO and non-MIMO base station and WLAN card. To simulate real-world activity, we timed the copying of a 6.36MB file from a hardwired network host to the WLAN-connected notebook.

The results
While it's easy to dismiss hype, our testing confirmed that MIMO offers a significantly better WLAN signal. In every location, particularly those where walls significantly hampered non-MIMO communication, the introduction of MIMO gear boosted both reported connection speeds and real-world performance - in some cases, dramatically. MIMO gear delivered 54Mbps and 108Mbps connections from every location tested, turning in performance ratings that were consistently double or more the speed of non-MIMO equipment. It wasn't necessary to replace both base station and PC Card to get results: speed also increased even when just the PC Card or base station was upgraded.

Using a MIMO base station and PC Card brought a stable WLAN signal to the brick garage, front bedroom and front yard -- three places where the non-MIMO equipment had struggled to reach. In fact, only one equipment combination -- MIMO base station and PC Card -- was able to penetrate into the spare room inside the brick garage, lighting up that area for the first time ever.

Locating local internet providers

Even your neighbours may benefit from MIMO, whether you want them to or not. Using the MIMO base station and PC Card, we got a usable WLAN signal from six houses up the street -- far short of the claimed 1km range, but much farther than was possible without MIMO.

Click here for a full table of the results for each room.

The results table
The following tables show the full test results, but the conclusions are clear: MIMO costs a bit more, but it delivers. If your house suffers frustrating WLAN black spots or you're struggling to get enough speed from your office in the garage, spending a bit more on MIMO equipment may be your answer.

Location	Distance from base station (m)	# brick walls	# plaster walls	WirelessMon (dB / %)	Config (MIMO PC Card / MIMO base station)	NetGear meter (out of 8 / Mbps)
Ground Zero	1	0	0	-43/80	1 (N/N) 2 (Y/N) 3 (N/Y) 4 (Y/Y)	8/54 8/54 8/54 8/54
Kitchen Table	8	0	3	-70/67	1 (N/N) 2 (Y/N) 3 (N/Y) 4 (Y/Y)	4/18 7/54 4/12 6/108
Front Bedroom	12	0	3	-79/63	1 (N/N) 2 (Y/N) 3 (N/Y) 4 (Y/Y)	6/18 5/36 4/6 4/24
Living Room	11	0	3	-74/66	1 (N/N) 2 (Y/N) 3 (N/Y) 4 (Y/Y)	4/12 7/48 4/12 6/36
Front Yard	20	1	3	-84/61	1 (N/N) 2 (Y/N) 3 (N/Y) 4 (Y/Y)	4/1 3/5.5 5/1 3/18
Back Yard	6	1	0	-76/64	1 (N/N) 2 (Y/N) 3 (N/Y) 4 (Y/Y)	6/18 5/48 4/12 5/36
Garage	12	2	1	-89/58	1 (N/N) 2 (Y/N) 3 (N/Y) 4 (Y/Y)	3/6 3/1 3/1 2/2

Location	Win XP (speed/ quality)	CNET speed (kbps)	Numion speed (kbps)	PCPitStop speed (kbps)	Seconds to copy 6.36MB file
Ground Zero	54/Excellent 54/Excellent 54/Excellent 108/Excellent	383 890 96 895	37 198 106 169	80 354 168 467	5 3 5 3
Kitchen Table	24/Good 54/Good 24/Low 108/Good	298 839 146 709	135 201 77 158	48 572 107 600	4 6 4 4
Front Bedroom	18/Good 54/Low 2/Low 108/Very Low	283 233 195 702	35 210 42 184	110 422 101 270	12 45 12 7
Living Room	18/Low 54/Good 12/Low 108/Low	76 839 132 620	33 209 63 176	136 430 143 446	41 DNC 12 7
Front Yard	1/Very Low 54/Very Low 12/Low 108/Low	DNC 436 107 519	DNC 135 DNC 149	DNC 446 55 501	DNC DNC 125 12
Back Yard	12/Low 54/Low 12/Low 108/Low	412 814 222 812	65 193 45 230	181 393 55 564	12 10 11 4
Garage	2/Low 54/Very Low 1/Very Low 108/Very Low	DNC DNC 48 286	DNC DNC DNC 109	DNC DNC DNC 160	DNC DNC DNC 65

Non-MIMO PC Card: Netgear WG511
Non-MIMO base station: Netgear WGR614
MIMO PC Card: Netgear RangeMax WPN511
MIMO base station: Netgear RangeMax WPN824

DNC = Test did not complete

Internet Benchmarks used:
CNET Bandwidth Meter Speed Test Numion YourSpeed

PCPitStop Download Test Other resources

• Catch the new wave in wireless networking: 802.11n
• FAQ: Wi-Fi alphabet soup
• Wireless: the next lap