A Way to Remake BroadbandKevin Fitchard
To say Southern California startup MagnaCom wants to shake up mobile and wire-line communications would be an understatement. The company is proposing that cellular, Wi-Fi, fiber, and wire-line broadband networks do away with one of their fundamental technological underpinnings—an age-old modulation scheme called QAM.
QAM stands for quadrature amplitude modulation, and it’s a means to manipulate signals so they carry information over communications links, whether it’s an episode of Sesame Street coming out of your cable box or a Twitter update popping up on your smartphone.
Improvements to that modulation scheme have led to big enhancements in spectral efficiency, meaning the same signals can carry more information. For instance, you may have heard mention of QAM in reference to mobile technologies: The boost from 16 QAM to 64 QAM was one of the means used to turn HSPA networks into faster HSPA+ networks.
MagnaCom’s founders, Chief Executive Officer Yossi Cohen, who has headed up mobile communications at both Motorola and Broadcom, and Chief Technology Officer Amir Eliaz, a communications scientist who developed MagnaCom’s technology, believe QAM has seen its best days. On Tuesday, they brought their company out of stealth mode with an audacious proposal: an alternative modulation scheme they call Wave Modulation, or WAM.
How WAM works is pretty geeky stuff, and I’ll admit it’s beyond my basic technical understanding. For the engineers out there, here’s MagnaCom’s explanation:
WAM technology uses spectral compression which improves spectral efficiency. The spectral compression enables an increase of the signaling rate (e.g. by 2x) thereby affording the use of lower order alphabet (e.g. a 64 constellation alphabet instead of 4096), which reduces complexity. It provides inherent diversity of time and frequency domains and uses nonlinear signal shaping and adaptive nonlinear model learning. The nonlinearities are handled digitally at the receiver side, allowing a lower-cost and lower-power transmitter design.
The bottom line is that MagnaCom claims it’s created a more efficient modulation scheme, one that can pack more information into the same spectral real estate. This means the same data could be transmitted at higher speeds and lower power, using less spectrum or fewer cables than required today. According to the company, WAM could produce a 10dB gain in a communications system, which could translate into a 400 percent increase in range, a 50 percent decrease in power, or a 50 percent drop in frequency bandwidth over today’s QAM-based systems.
If the claims bear out, this would amount to a tremendous advance in communications technology. MagnaCom has to prove its technology’s merits to the communications industry and then make it commercially viable. The company plans to do the proving at CES next month in Las Vegas, teaming up with programmable logic device maker Altera to build a live WAM demonstration platform at the trade show.
As for making the technology viable, MagnaCom claims that WAM will be easy to implement. Its modulation scheme is purely digital, so it will work over the same analog components, radios, and antennas already in devices and infrastructure, and it’s fully backward-compatible with QAM. WAM would require changes to communications semiconductors, but MagnaCom says its technology would consume 1 square millimeter of silicon space in modern communications chips.
Also from Gigaom
Efficient Twitter Theory (subscription required)