Skip to Content
Uncategorized

A Faster Wireless Web

A new protocol called fasp-AIR promises speedier mobile downloads.
February 16, 2010

Transfers of large amounts of data across the Internet to wireless devices suffer from a key problem: The Transmission Control Protocol (TCP) used to send and receive that data can be unnecessarily slow.

Mobile speed: Fasp-AIR will make its first appearance in the wild as an iPhone application designed to speed media uploads over wireless networks.

A company called Aspera has now announced an alternative protocol designed to accelerate wireless transfer speeds. Called fasp-AIR, it includes new proprietary approaches to addressing problems of data transfer that are unique to wireless communications. The original fasp protocol is already used to boost regular Internet transfers. It was used, for instance, to speed up the transfer of files from New Zealand to the U.S. during production of the movie Avatar.

The main problem with the TCP protocol, which was designed before wireless connections to the Internet were commonplace, is that it doesn’t know the difference between packets of data that are lost because of network congestion and those that are lost because of a weak wireless signal. TCP automatically throttles the speed of data transfer when it sees dropped packets, so that congestion doesn’t overwhelm the network. That’s fine when packets are lost because of congestion, but when the problem is a weak signal, it causes an unnecessary drop in transfer speeds that can bring downloads and uploads to a crawl.

For some applications, like streaming video and Internet telephony, it’s possible to use an alternative like the User Datagram Protocol (UDP), which doesn’t bother to confirm that all data has arrived intact. The price of UDP’s speed is dropped packets of data–a result familiar to anyone who has endured the degraded quality of a video stream or telephone conversation when at the limits of a wireless network’s range.

Fasp-AIR achieves faster speeds than TCP but doesn’t result in any dropped packets, making it suitable for transferring data that must arrive complete and intact. “The drop-off in performance we see with fasp-AIR is almost linear,” says Aspera CEO Michelle Munson. “So a 10 percent loss in the available bandwidth means we’re still getting transfer rates that are 90 percent of what’s specified.”

At first, fasp-AIR will be available as an iPhone app that can be used to access enabled servers. Fasp-AIR requires that both the client and the server are running software developed by Aspera. In the future, Aspera hopes that developers will incorporate fasp-AIR into their applications directly. Aspera licensees currently include Amazon and several other large Internet companies.

Fasp-AIR certainly isn’t the only novel approach being used to speed up transfers of wireless data. Jon Crowcroft, Marconi Professor of Communications Systems at the University of Cambridge, says that some wireless carriers use a proxy server between the wireless and the wired networks to intelligently adapt to changing network conditions. This gets around the problem of whether or not a TCP alternative like FaspAIR is hogging bandwidth on a congested network.

Because Aspera’s algorithms are proprietary, Crowcroft says it’s impossible to tell whether fasp-AIR “is a huge win compared with other solutions.” However, he notes, if fasp-AIR includes, as its makers claim, an algorithm that can distinguish between whether packets are being dropped because a network is congested or simply because the wireless signal is weak, fasp-AIR represents a workable solution. “I don’t know how they figure out which scenario they’re in,” says Crowcroft, “and that’s the clever bit.”

Crowcroft’s own solution to this problem, first implemented with Vodafone almost a decade ago, involved configuring the proxies that connected the wireless network to the wired data and voice backbones so that they could monitor all TCP traffic. This allowed the proxy to tell individual devices when they were in a “lossy” part of the wireless coverage area. “It’s not a general solution,” says Crowcroft. “If [Aspera] has a general solution, good luck to them.”

Another wireless network company, the San Francisco-based Meraki, takes a similar approach, mapping a network’s usage and lossiness and then using that information to optimize it at the level of individual wireless routers.

“What we have found is that when it’s deployed properly, 802.11n can be pretty darn fast, even with standard TCP traffic,” says Kiren Sekar, product marketing manager at Meraki. Sekar sees Meraki’s approach as complementary to efforts like fasp-AIR.

One issue that fasp-AIR does not resolve is the limited capacity of existing networks, a problem that has affected parts of AT&T’s 3G network, largely due to overwhelming iPhone data traffic. “Networks are already over-provisioned, even without good utilization [such as what’s achievable with fasp-AIR],” says Munson. “Things like fasp-AIR will create pressure on AT&T, and I don’t know where it will lead.”

Munson imagines that in the future, wireless providers could use more sophisticated data-transfer protocols to charge more for faster deliveries of large files. But this prospect raises the specter of violating the principles of Net neutrality, which hold that all data should be treated equally by Internet service providers. BitTorrent already exploits weaknesses in TCP by opening up multiple connections at once, Munson notes.

“If one application is greedy, then there’s no way within TCP to enforce bandwidth limits [on that application],” says Munson. “When using fasp-AIR, however, you could control bandwidth on a flow-by-flow basis so one application can’t get around observing whatever caps there are.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.