Clearer Signals for Faster Phone Downloads

New technology for mobile devices is being designed to cut through traffic-jamming interference to improve downloads.

Cell-phone service providers are running into a jam. More and more people are signing up for smart phones that can download large amounts of data in the form of Web pages, music, and video. And with more data traveling the wireless pathways, bandwidth -- the data-carrying capacity of wireless channels -- is decreasing, slowing download speeds and causing more calls to get dropped from the network.

To help eke out more bandwidth, TensorComm, a Westminster, CO-based telecommunications startup, has developed and tested a set of algorithms that can efficiently cancel out the signal interference from voice and data streams caused by multiple cell-phone signals, thereby freeing up more capacity.

Currently, in order for an individual signal from a cellular base station to be received by the intended cell phone, the station needs to transmit the signal through the interfering noise of other phone signals. This can use significant amounts of power and bands of frequencies on the communication spectrum -- resources that could be used to increase bandwidth. By clearing interference at each handset, base station resources could be used to improve cellular service.

TensorComm has tested its technology in commercial phones in a major U.S. operator's network, and the company claims that, without using large amounts of battery power, it can increase download rates by up to 80 percent and boost the number of phones a network can carry by 60 percent, resulting in fewer dropped calls.

Interference is a well-known challenge in the wireless communication industry, says John Thomas, chief executive officer of TensorComm. Historically, engineers have assumed that the problem couldn't be solved in an efficient way, he says. "You can come up with algorithms to solve this in the lab," says Thomas, "but the implementation gets so big so fast that [lab solutions are] not practical for a handset." In contrast, he says, TensorComm's technology is small enough and efficient enough to be deployed in smart phones and PDAs.

The signal interference that TensorComm's signal-processing algorithm targets is a result of the basic structure of modern-day wireless signals. In the mid-1980s, a technique called CDMA (code-division multiple access) was found to increase the number of calls sent over a given swath of frequencies. CDMA is signal-processing technology that tags each wireless signal with a code, allowing numerous signals to travel intermingled over the same frequencies, at the same time, to be sorted out by a mobile device's antenna.

Currently, CDMA is the signal-processing method of choice in the United States for Verizon and Sprint. While the technique can cram more calls into a given frequency than previous methods, it produces interference, says Sergio Verdú, professor of electrical engineering at Princeton University. CDMA signals "interfere with each other because they're transmitted at the same time and same frequency," he says. A basic cell phone's antenna tries to pick out the right signal from all the other signals, which appear as background noise. But the interference makes this challenging. It's similar to trying to have a conversation in the midst of a noisy cocktail party, Verdú says.

TensorComm's technology, says Thomas, is able to identify the different types of signal interference that bombard a cell phone's antenna from neighboring cell-phone towers. Then the algorithms, which are built into a tiny piece of silicon installed in a cell phone's modem, are able to selectively block the unnecessary signals, canceling out the interference. The process is akin to figuring out which conversation you want to listen to at a party by sampling snippets of other conversations. From a few words and knowledge of who said them, it's easier to choose a conversation.