Startups challenged to tackle biggest 5G hurdles – latency and resiliance
The IEEE recently gathered more than a dozen stakeholders at an event – mainly researchers and system and chip vendors – to discuss the work involved in defining the next generation of cellular networking. They share a common timeline for the work under the 3GPP standards body and a set of elements to put in place that includes support for:
- Data services at 20 Gbits/second peak and 1 Gbit/s average;
- Latency as low as 1 millisecond;
- Reception while traveling at up to 500 km/hour;
- Frequencies between 6 and 100 GHz;
- Safety-critical applications such as self-driving cars.
“The faster speeds will come, but the latency and reliability targets will take a lot of work,” said Arogyaswami Paulraj, a wireless expert and engineering research professor emeritus at Stanford in a talk here.
“The biggest challenge with 5G is getting to the reliability so it always works, people talk about the ‘five nines’ concept or may want something even higher,” he said. “Today wireless reliability is a joke in mobile systems, but you can’t afford to let trains or car anti-collision systems go off on their own, so this will be something of a holy grail,” he said.
Today’s cellular systems have latency of about 20-30 milliseconds. Shaving that down to 1ms for a host of applications from virtual reality games to tele-medicine will be another one of the toughest challenges of 5G, Paulraj said.
Supporting new millimeter-wave frequency bands and massive MIMO antenna arrays will also be among the top technical challenges, said Paulraj who helped pioneer the MIMO technology used in today’s LTE and Wi-Fi networks
High frequency bands and arrays of hundreds of antennas will “stress everything we have today… there are lots of open issues, and it will take lots of hard work to clean things up and come up with a reasonable design,” he said.
Paulraj called for startups to take up the challenges. “5G will be a major change and big companies are not good at innovating, they move slowly so there are opportunities for startups,” he said.
He founded Iospan Wireless Inc., in 1998 to commercialize his work in MIMO. It was acquired by Intel in 2003, and the technology was later used in WiMax, LTE and 802.11m networks.
The 71-year-old entrepreneur said he has no plans to form any companies around massive MIMO for 5G, however, he is advising multiple startups. “There’s a long way to go [in building 5G] and I hope we can create some startups to look into these things,” he said.
Indeed, it’s already happening. At least one startup claims it has a new approach that beats the OFDM modulation used in today’s cellular systems and is expected to be used, in a modified form, for 5G.
Cohere Technologies is the brain child of an associate professor of math at the University of Texas at Austin. The professor (Ronny Hadani) applied some math concepts to wireless that resulted in concepts for using Doppler delay for wireless channel estimation to eliminate fading problems experienced with time and frequency modulation schemes and thus improve spectral efficiency, one of the key goals of 5G.
The startup is now preparing a white paper to give some of the first public details of its concept. Paulraj was not familiar with the startup but said channel estimation is one of the key problems for 5G, calling it “an elephant in the room.”
— Rick Merritt, Silicon Valley Bureau Chief, EE Times