Researchers at the University of California, San Diego (UCSD), have developed a diagnostic system that monitors Wi-Fi activity in a building and determines why traffic slows, signals dip, and laptops get kicked off the network. The researchers say that elements of the system, which consists of traffic-monitoring hardware and specialized software that analyzes the activity, could easily be deployed in offices and buildings to help network administrators find and fix problems more easily.

“The magic is the software that takes all the data from all these different points [around the building] and stitches it back together to see what it means,” says Stefan Savage, a professor of computer science and engineering at UCSD and one of the lead researchers on the work. He says that the software is designed to infer why certain problems occur when they do, and to produce a report that breaks down the problems into their components. For instance, a video streaming to someone’s laptop might stutter because a microwave oven is using the same wireless frequency, because a large number of users are trying to access the same signal, or because the host site is experiencing heavy traffic.

Wi-Fi tends to be an unreliable form of wireless communication, says Savage, in part because the set of instructions that governs how it passes data through the air was never designed for the widespread use that Wi-Fi currently experiences. Today, network administrators try to fix Wi-Fi problems by looking at a number of things that could interfere with a signal: hardware malfunctions, software bugs, and outside devices, such as microwave ovens and cordless phones. But these things can all change quickly, making Wi-Fi failures difficult to anticipate. What’s more, says Savage, it’s nearly impossible to accurately diagnose problems after the fact, which is why an automated system is so desirable.

In the past, researchers designed automated diagnostic systems that monitor how individual components of a Wi-Fi network affect its performance, but no one had looked at the problem comprehensively, says Savage. “In the end, you can’t look at [the components] in isolation because they all interact,” he says.

To identify Wi-Fi network problems across the computer-science building at UCSD, Savage and his team set up 192 traffic-monitoring radios throughout the building, where there are 40 wireless access points, the boxes that send wireless data from a wired network. The radios collect information about the traffic and report all wireless events–packets of data sent and collected, and signal-strength dipping, for instance–to a storage server. Software called Jigsaw then merges the data from all the different radios and creates a single, unified report of building-wide wireless operation.

Monitoring wireless networks is much more challenging than monitoring those that are wired, says David Wetherall, director of Intel Research Seattle. In wired networks, he explains, one can attach a box to the network hardware that will reliably count the number of packets in and out. But in a wireless network, it’s impossible to collect all the packets, and these lost bits invariably obscure the picture of actual activity. To solve this problem, the UCSD researchers developed a novel set of algorithms that infer wireless activity that isn’t directly measured. For instance, if a monitoring radio sees that a laptop received a packet, but didn’t see that a packet was sent to the laptop, the algorithm can infer that a packet was sent. The researchers’ model, explains Savage, infers behavior about activity at many different levels of the network. The model takes into account the structure of the underlying wired network, the method used to encode data into wireless signals, and the manner in which access points and wireless cards in laptops send and receive information.

The researchers have “done a nice job extending and synthesizing known inference techniques into a useful system,” says Wetherall. He suspects that it wouldn’t take much to use this approach to make a commercial system.

While the approach couldn’t solve the problem of spotty wireless coverage in an apartment building where access points aren’t connected to a single wired network, it could be modified to monitor citywide Wi-Fi. Savage says that his team is looking at better ways to monitor large-area Wi-Fi, which is more difficult to do than monitoring Wi-Fi in a building. “If you’re trying to cover a large area, then there may be some places that you don’t see at all, or you just might have one data point for.” In this case, he says, the inference algorithms would have to be tweaked to make more guesses based on much less information.