Even allowing for the typical embellishments in advertising, Netgear's claims about the performance of its pre-802.11n equipment were always suspect. Some customers found that the suspicion was justified and decided to do something about it: They launched a class-action lawsuit charging the company with exaggerating the data speeds of its WiFi networking devices. The company has agreed to settle these claims.

In its last week's filing with the Securities and Exchange Commission (SEC), the company said it had agreed to pay $700,000 to settle a class-action lawsuit initiated in June 2004. A second lawsuit, filed in February 2005, was voluntarily dismissed in favor of the 2004 lawsuit. The settlement calls for customers who purchased Netgear wireless devices between January 1999 and November 2005 to receive a 15 percent discount on the purchase of a new wireless device. The agreement must be approved by the Santa Clara County Superior Court in California.

Netgear will change its advertising for WiFi gear to say that the data speeds advertised are the maximum rate but that "actual throughput will vary" depending on different factors. Netgear will also donate $25,000 worth of its equipment to charity as part of the settlement. In its SEC filing the company said that it does not admit any liability and that it "has agreed to enter into this agreement to avoid the further expense, inconvenience and distraction of burdensome and protracted litigation."

For more on the Netgear settlement
- see Grant Gross' IDG report

BACKGROUND: Netgear's claims, and the claims of many other WiFi gear makers, should be taken with a grain of salt. There are two reasons for this--the first has to do with MIMO technology and the second applies to WiFi more generally. The first reason addresses the question under what conditions MIMO technology makes a difference (the Netgear equipment in question was pre-802.11n, using MIMO). Typically, WiFi transmission is hobbled by a phenomenon called multipath: In an obstacle-rich environment the signal bounces off the obstacles (for example, office furniture), in the process generating another reflected signal which causes receivers to become confused. The receivers become confused because the obstacles cause the signal path lengths to be unequal, which results in several time-delayed copies of the information signal arriving at the receiver. The received signal is then the sum of all signals according to their phase relationships, and multipath results in received signals that are spread out in the time domain.

The beauty of MIMO technology is that it exploits the multipath signals to increase throughput. MIMO uses multipath to carry more information, which is recombined on the receiving side by the MIMO algorithms. There is a catch, however: When MIMO wireless is used in an environment without obstacles, the user gains nothing over more traditional WiFi technologies. In an environment without reflections there is no multipath, and MIMO is useless.

The second reason why WiFi gear makers' claims should be treated with caution is that a considerable portion of the WiFi transmission is taken by the set of rules governing the transmission. For example, the sender has to wait a period of DIFS (Distributed Coordination Function InterFrame Space) time of 50 Usec before the channel is presumed clear of traffic. Only then can a Data Frame or a Request to Send frame be sent. The receivers answer to this is with an ACK or Clear To Send, accordingly. The receiver has to wait a SIFS (Short InterFrame Space) time of 10 Usec before this reply is sent, and so it goes. The portion of the transmission which is actually used for the unadorned contents being transmitted is thus smaller than gear makers would typically claim for their equipment.

For a technical discussion of 802.11 throughput:
- see this Web site at Uninett, the Norwegian research network
For more on the multipath problem:
- see this Cisco document
- and Guy Kewney's Newswireless discussion