It's called Coded TCP.
Developed by a multi-university team under MIT's Muriel Menard, it's a system for replacing wireless packets with algebra that keeps them from having to be re-sent. Re-sent or dropped packets can reduce the effective data rate of a wireless link by a factor of 10. Eliminating the need to re-send packets essentially makes wireless networks almost as efficient as wired ones and, thus, increases the value of spectrum at a stroke.
This is big news for spectrum holders like AT&T (T), Sprint (S), and Verizon's (VZ) half-owned Verizon Wireless. They can multiply their throughput through a simple firmware upgrade. The system is being licensed through a company called Code-On and could come into standard networking within a few years.
Medard herself did not promise an end of the so-called spectrum crunch to Technology Review, simply stating that "severe inefficiencies should be remedied" before new spectrum is sought. The industry insists demand for wireless services is going to grow by 18 times in the next few years, and the tenfold speed increase promised by coded TCP will still fall short.
But there's a second meaning to this breakthrough. It works on Wi-Fi -- it was first tested on MIT's own Wi-Fi network -- and could also weave data streams from a new cellular standard called LTE (often sold as "4G") with Wi-Fi fairly seamlessly. This would increase the usefulness of Wi-Fi, which is usually closer to a wired base station than a cell tower, and again result in more data flowing without more spectrum being licensed.
So the biggest winners in all this may be the makers of Wi-Fi and LTE equipment, as well as the chipmakers behind that gear. We're talking here of companies like Qualcomm (QCOM), Broadcom (BRCM), and Cisco (CSCO). Then there are the makers of wireless devices -- companies like Apple (AAPL), Microsoft (MSFT), and Google (GOOG) -- that will now find they can actually deliver on promises of moving video seamlessly around homes and offices. Coded TCP could also reduce congestion on hospital and campus Wi-Fi networks.
One of the biggest scams going, for some years, has been the idea that people can easily watch TV on smart devices while connected to a cellular network. It's not easy, it's not seamless, and it's very expensive. You cut those costs by a factor of 10 and suddenly the promises you were making can be delivered upon.
My guess is this will become an industry standard in fairly short order -- meaning by 2015 -- and that the whole wireless industry will get a boost from it. But since the whole industry will be benefiting, the only investment advice available is simply to bet on those companies most closely leveraged to the industry. Companies like Broadcom, the chipmaker, and Apple, the leading device company. Buy these names on weakness and hold them for the long term.