Original URL: http://www.theregister.co.uk/2012/12/18/darpa_100g_wireless/

DARPA planning 100 Gbps wireless

Seeks interest in development and demonstrations

By Richard Chirgwin

Posted in Networks, 18th December 2012 00:34 GMT

DARPA (the Defense Advanced Research Projects Agency) has decided that the 200-276 Mbps wireless technology currently used for military communications, known as the Common Data Link, is not going to live forever, and is inviting companies to submit proposals to boost battlefield wireless to an impressive-if-achieved 100 Gbps.

The agency is hosting a “proposers’ day” event in January 2014. It says its Broad Agency Announcement (BAA) will set out the details of its “100G” program, “with the objective to design, build, and test a communications link with fiber-optic-equivalent capacity, long reach, and high availability in airborne-to-airborne and airborne-to-ground configurations that can serve as a deployable data backbone in a military communications network.”

The communications platform would have to match the weight and power specifications of today’s CDL platforms, and provide 200 Km reach for air-to-air links and 100 Km for air-to-ground links.

Since the platform has to operate in all weather conditions, the simplest solution – free-space optics – is ruled out.

DARPA’s media release quotes program manager Dick Ridgeway as saying “Providing fiber-optic-equivalent capacity on a radio frequency carrier will require spectrally efficient use of available RF spectrum … 100G plans to demonstrate how high-order modulation and spatial multiplexing can be synergistically combined to achieve 100 Gigabits per second”.

If, as speculated at ExtremeTech, the 100G program is likely to use Ku-band frequencies, then WiFi’s spectral efficiency of around 11 bits per Hertz doesn’t come even close to 100G’s requirements. The widest transmission allocation in the Ku band in Australia (since that’s the database this writer is familiar with) is 550 Mhz, yielding around 6 Gbps at WiFi rates.

Turning the calculation the other way: at the spectral efficiency of current commercial-grade kit, 100 Gbps would need a radio channel wider than the entire Ku-band, at 9 GHz. Hence the additional attention on spatial multiplexing.

So while there’s no prospect that DARPA is about to render optical fibre obsolete, The Register would anticipate some interesting science to flow from the project. ®