Home & Office

Wi-Fi 'traffic cop' software boosts hotspot performance by 'up to 700pc'

A team of researchers from North Carolina State University say they may have cracked the problem of Wi-Fi hotspots buckling under the pressure of too many concurrent users, through a tool they have developed called WiFox.
Written by David Meyer, Contributor

A team of US researchers has come up with a piece of software for drastically improving the ability of Wi-Fi hotspots to cope with dozens of users.

Anyone who has been to a conference or tried to log onto an access point (AP) at an airport will be familiar with the problem: a slow connection that frequently conks out. The North Carolina State University researchers say this is because the 802.11 protocol's distributed coordination function was not intended to support so many people — the data sent between an AP and all the clients connected to it travels over a single channel that quickly gets clogged up.

According to the team, their WiFox software can improve throughput by up to 700 percent, with performance actually scaling up as more people connect to the same AP.

"One of the nice things about this mechanism is that it can be packaged as a software update that can be incorporated into existing Wi-Fi networks," NC State computer science student Arpit Gupta said in a statement.

The WiFox software solves a key problem with the way APs work with a multitude of clients, namely the fact that 802.11 gives AP and clients equal status when it comes to sharing channel resources.

It has not previously been practical to simply give the AP higher channel access priority, as this would throttle the ability of the clients to send user requests to the AP. WiFox comes at the problem from a different angle, giving the AP priority only when it is developing a data backlog.

"The amount of priority the access point is given depends on the size of the backlog — the longer the backlog, the higher the priority. In effect, the program acts like a traffic cop, keeping the data traffic moving smoothly in both directions," the team explained.

The researchers added that, in their lab test, "improvements ranged from 400 percent with approximately 25 users to 700 percent when there were around 45 users". They said this effectively meant an average fourfold improvement in the amount of time the AP takes to respond to user requests.

Editorial standards