Talk, Talk And More Talk

A glimpse into the latest research at cell phone companies and universities, and how they’re planning to extend battery life.

popularity

By Ed Sperling
To anyone who owns a cell phone—and there are at least several billion people who claim that distinction these days—it’s not surprising that bad reception lowers battery life.

More bars, while not the most accurate gauge of a signal, are at least an indication that you can extend the time between charges even if you’re watching streaming video. But work is under way at universities and large communications technology companies to add significant battery life to devices from outside the phone, which in turn will have an effect on what is designed into the phone and how it utilizes this new technology.

At the heart of this change is a rethinking of how to best deliver a signal to the handset. Until now, the focus has been on a signal being handed off from one cell tower to the next—and usually the one that’s closest to the device receiving the signal. While that intuitively may sound like the best approach, it often isn’t—particularly if a signal can be shared across multiple towers and then reconstructed.

From a high level, this is something akin to triangulating a signal based upon the receiver’s whereabouts and direction of travel. It requires a rethinking of how signals are received and interpreted, but the amount of energy that can be saved both at the base station and in the handset can be huge. And as an added bonus, there likely will be fewer dropped calls.

Timing is everything
One of the issues that will need to be solved is time delay. Distance adds latency, and while this may not be noticeable to the cell phone user with a single tower, it can be troublesome when multiple towers share the signal. Reconstruction of that signal is possible, but it will require some re-engineering at the base stations and some tweaks in the software and the hardware in the handsets.

Researchers at the University of Technology in Dresden, working with Vodafone, say that while research has been done for urban areas, the distance between towers in rural areas is much greater. This is true in Europe, but it’s even more true in areas with vast stretches of undeveloped land as well as farmland, and in some developing countries where tower locations are irregular.
http://mns.ifn.et.tu-dresden.de/Lists/nPublications/Attachments/754/MCSS11_VKo_SyncRequirementsForCoMPInLargeCells.pdf

But even within urban areas, or in mountainous or hilly regions, interference between multiple towers can cause problems. And the Dresden researchers note that asynchronous timing caused by inaccurate oscillators can cause their own latency issues.

“When you’re designing a protocol, the first consideration is to get as much useful information across a channel per hertz,” said Chris Rowen, CTO at Tensilica. “The move from 2G to 3G to 4G has been about finding denser ways to pack in more bits and increase bandwidth. But bandwidth is expensive, so the first consideration is to get more information across. The second consideration is to make sure there is no excess power spent. That’s why when a handset is close to a cell tower it knows not to have to broadcast at maximum power.”

But power was never actually a consideration inside of base stations when they were first constructed. Like many devices created over the past couple of decades, energy efficiency was an afterthought and often a “nice to have” benefit rather than a “must have.” That thinking has changed significantly in recent years for a couple of reasons. One is that the density inside of SoCs has increased to the point where it can cause myriad physical effects that can disrupt signal integrity. The second is that increasing bandwidth required for the advancing communications standards raises the I/O power needed to drive these chips, which can be pricey.

“In energy-efficient multi-chip base-station designs, I/O power has to be carefully monitored and optimized,” said Mahesh Tirupattur, executive vice president of Analog Bits. “Low-power interconnects for chip-to-chip and wireless and wireline communication is imperative.”

Positive reception
What’s good for the base station will be good for the handset, as well. Cary Chin, director of technical marketing for low-power solutions at Synopsys, has been running a series of tests using both iPhones on both AT&T and Verizon networks to compare signal strength and battery life. http://lp-hp.com/absolute-power/2011/02/

“In the experiments I’ve run, the radio can go from a significant drain to the dominant drain on the device,” said Chin. “There is a high correlation between how fast the battery drains with the strength of the signal. The bottom line is that the radio designers and the people who come up with the RF are looking at changing the standard way of transmission. It’s not just a power implementation, though. It’s also about how you transmit and store that information.”

One noticeable change will likely be the inclusion of multiple antennas, both for transmitting and receiving data. Two antennas are better than one, and presumably many antennas are better than fewer—as long as the data can be effectively reconstructed. This concept, known as multiple-input and multiple-output, or MIMO, dates back to research done at Bell Labs back in the 1970s. The concept has seen limited use in the past, although that is likely to change in the future as shared signals become reality.

“Even if the antennas are just a few centimeters apart in a handset you can get significantly more information,” said Tensilica’s Rowen. “With base stations they can be kilometers apart, so you can provide much more information to the handset. And when you add high bandwidth, you get a large list of time slots and frequencies that you’re listening to, each a few hundred kilohertz in width.”

That also means data can be parsed out and utilized by handsets as needed for specific tasks, with an appropriate level of performance and power for each function. This is particularly important for another reason. Rowen said that as wireless algorithms become more complex, there is more of a challenge in keeping power down.

Conclusions
What’s changing significantly in this type of scenario is the recognition that a communications system is much broader than a single device. It encompasses both receivers and transmitters on a network, and energy efficiency can be best achieved by including both parts of the system in the design.

For the first decade of their existence, the real challenge for mobile phones was simply getting a signal and being able to carry on a voice conversation. The next challenge will be receiving and transmitting a variety of data in the most efficient way, saving cost on the base station side and battery life on the handset side. But getting there will require cooperation across a complex supply chain in which the individual companies don’t necessarily see themselves working on the same problem, from the chips in the base stations and the handsets to the protocols, the algorithms and the software used by each.

The reasons for bringing all of this together, however, are pointed in the same direction. On the base station side there are huge cost savings to be had. On the handset side, competitive pressures will require these changes. And on the consumer side, there is an incessant demand to do more on a handheld device without worrying about the time between battery charges and whether a signal is available.