Faster and Higher
One method that almost all the developers have utilized is to reduce the overhead described by the standard protocols. For example, the standard provides an obligatory pause (Distributed Inter-Frame Space – DIFS) after transmission of each Wi-Fi frame: the transmitting station is listening to the air and begins to transmit the next frame only if other stations haven’t contended for the channel within this period of time. So, the inter-frame space is necessary to separate the data-transfer medium.
It’s clear that you can achieve a considerable performance growth if you’re transmitting several frames in a row without DIFS pauses, even though the medium has to be monopolized for the whole period of transmission. Referred to as frame bursting, this technique has got widely popular in proprietary technologies intended to increase the performance of wireless networks. In fact, frame bursting is part of the recently ratified standard extension IEEE 801.11e (Quality of Service, QoS), but it’s hard to tell what degree of compatibility there is between the existing solutions and this extension.
Besides frame bursting, a considerable performance boost (about 30%) can be achieved by merging two (Atheros) or more (Texas Instruments) frames due to a reduction of the frame transmission overhead (this method is referred to as “fast frames”, “frame concatenation”, etc.)
Some exclusive solutions also use compression of Wi-Fi frames with the Ziv-Lempel algorithm, but this compression isn’t generally very effective when applied to small-size frames.
Yet another method to get more performance is implemented by Atheros. Its point is in combining two physical radio channels into one logical high-speed channel between the receiver and transmitter (his can be viewed as analogous to Cisco Fast EtherChannel or Link Aggregation Control Protocol (IEEE802.3ad) for Ethernet networks and to the multilink PPP protocol). This Super G Turbo technology helps double the bandwidth of the channel to achieve a theoretical data throughput rate of 108Mbps. There is one hitch, however, Atheros is much criticized for. I won’t delve into details, but equipment on the Atheros chipset goes beyond the standard use of the 802.11g frequency spectrum when in Super G Turbo mode and may interfere with operation of other networks. Realizing this, Atheros provided two Super G Turbo sub-modes: dynamic and static. In dynamic mode, the utilization of the entire 802.11g range is constantly being analyzed and the bandwidth is doubled only when there are free channels available. The dynamic mode is backwards compatible with any 802.11b/g equipment which does not support Super G Turbo. The static mode is asocial in the meaning that it can interfere with neighboring networks and is only compatible with itself. Instead, it provides the maximum data-transfer speed.
There are other approaches to doubling the radio interface bandwidth. For example, Airgo solved this problem in an elegant way by improving the spectrum efficiency instead of doubling the frequency band as Atheros did.