The WiFox software developed at North Carolina State University promises performance gains between 400 and 700 per cent on congested Wi-Fi links.
Researchers at the North Carolina State University have released details of a software technique that could boost the throughput of congested Wi-Fi connections by up to 700 per cent.
Dubbed 'WiFox,' the system is designed to be easily incorporated into existing access points that deal with large numbers of users such as public hotspots in coffee shops, convention centres, offices, schools and universities. When installed, the team claims it can boost throughput on congested links by between 400 and 700 per cent.
WiFox works by monitoring the amount on traffic on a particular Wi-Fi channel, looking for backlogs of data caused by congestion. Should the access point become congested, the system gives it priority over client devices to clear the backlog. As the backlog clears, the priority is reduced, while the access point's own traffic receives no special priority when no backlog is present.
The results are startling: by allowing the access point to reduce its backlog without riding roughshod over users' data requests, the team claimed to boost their lab-based Wi-Fi system of 45 users by 700 per cent. When the system was less congested, the gains were less - but even with just 25 users active, the software enabled a claimed four-fold gain in data throughput.
The WiFox system should be relatively cheap for manufacturers to implement, too. '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,
' explained Arpit Gupta, a PhD student in computer science at the university and lead author of a paper describing the work. 'WiFox can be incorporated without overhauling a system.
Gupta's paper, 'WiFox: Scaling Wi-Fi Performance for Large Audience Environments,
' is due to be presented at the ACM CoNext conference in France next month. Thus far, no commercial partnerships have been announced to exploit the software.