AirHop, Adapting to “Dense” Wireless of the Future, Develops Self-Organizing Networking Software
One of the recurring themes during the international wireless industry’s conference in San Diego last month was the phenomenal surge in mobile data traffic, and how it is leading to constraints and bottlenecks in existing network infrastructure.
As engineers approach the limit in terms of wringing any more efficiencies out of existing radio bands, Qualcomm CEO Paul Jacobs and other experts are talking about the need to increase network “density” by mixing ever-smaller microcells, picocells, and femotcells in closer proximity with standard macrocell towers. Such changes represent a dramatic change in network design, and would have to be part of the industry’s broader move to 4G wireless technologies. But the advent of more dense networks with overlapping cellular configurations also means increased inter-cellular radio interference among the smaller and more powerful cells that would operate closer together—with some operating entirely within the footprint of larger base stations.
It is a problem, though, that Yan Hui saw coming in late 2007 when he left a San Diego-based research & development group operated by Texas Instruments to found AirHop Communications. The San Diego software developer specializes in SON, or self-organizing networking technology, that is intended to simplify and coordinate the operation of 4G wireless networks while minimizing radio interference and maximizing mobile data rates.
“With 4G networks, we know the structure is going to be totally different,” Hui tells me. Where the high end of data rates in existing 3G wireless networks range from 3 megabits per second to 7.2 megabits a second, Hui says 4G technology is promising 100 megabits per second. Hui says, “The industry recognizes that the only way to get to high data rates is with smaller, dense cells” that will require re-using frequencies and managing the inter-cellular interference.
Hui says the software that AirHop is developing is intended for use in 4G wireless base station hardware, and that the company’s software engineers have been working closely with Texas Instruments and at least three other wireless chipmakers. He describes AirHop’s customers as the component and system venders that make base station equipment for wireless network operators. The company is self-funded, and raised about $1 million in September from individual investors to expand its marketing and business development efforts.
The company, which has 10 employees, estimates its software will reach the market in 2011, although some 4G networks could be deployed earlier. Among AirHop’s selling points is that the multi-tier design of 4G networks will be too complex for conventional installation and set up, so AirHop-equipped cellular hardware will be self-configuring. “It’s called plug and play,” Hui explains. “You turn it on for two hours [while it determines its network requirements] and then it starts working.”
Hui explains that AirHop’s technology also must be capable of managing radio interference from nearby cell sites that is expected to change, a problem that he calls “dynamic interference.” Unlike conventional networks in which cellular base stations more-or-less operate in a steady state, Hui says 4G networks will be continually evolving as more cellular base stations get installed. Because femtocells (the smallest cellular base stations, typically designed for use in a home or small business) will be portable and relatively easy to install, a user’s wireless connectivity could be affected by radio interference from nearby cells in this dense network configuration. It also will change as neighbors and nearby businesses intall their own femtocells. Hui says AirHop’s approach, called evolving SON, addresses dynamic interference by enabling cellular hardware to actively coordinate and manage network traffic among small cells in a neighborhood or business district.
AirHop says that in addition to conventional self-organizing networking capabilities, its evolving-SON technology provides additional capabilities that enable inter-node communication and coordination through a unified, application-level software platform.
“The algorithm is still a traditional communications algorithm used to manage radio resources,” Hui says. “We’ve just applied it to multiple cells… It’s really an optimization problem.”
He adds that AirHop’s technology does not represent an incremental improvement over existing technologies used to build out wireless networks. “We think this idea is very disruptive itself,” says Hui, who also notes the company so far has filed 15 patent applications for its technology. “It’s something that didn’t exist before,” he says. “We looked at the problem very early, and correctly.”