'A billion devices': Huawei eyes the prize with polar codes

Huawei founder Ren Zhengfei (l) with Erdal Arikan

In fast and furious 21^st century business, it’s rare that a big firm can get its leaders in the same room at the same time.

So something special must have gone down last week to bring together Huawei's chairman, founder, board of directors and all three alternating CEOs.

It was special, but not what you'd expect, unless you were expecting a Turkish maths professor who came up with polar coding.

Huawei's guest of honour, Dr Erdal Arikan, published a ground-breaking paper on polar codes in 2008, enabling a whole new framework for maximising data rates.

Outside China, only 3GPP diehards and information theorists would have heard of it. 

Inside China, polar coding has achieved a kind of minor celebrity, as demonstrated by the presence of the state TV network.

Few might know it’s officially the error correction protocol for the 5G mobile broadband control channel, but everyone can tell you this was the ‘Chinese’ technology proposed by Huawei back in 2016 and that another Chinese firm, Lenovo, scandalously voted against it.

News of this treachery became public in May, just as the nation was learning, thanks to the ZTE shutdown, of its heavy reliance on foreign IP.

The social media outrage brought Lenovo founder Liu Chuanzhi out of retirement. He denied "committing treason” and pointed out that after voting against the Huawei proposal Lenovo had in fact voted for it.

That didn’t mollify everyone - nationalist tabloid Global Times thundered against "underhand moves to derail Chinese businesses."

As we saw with TD-SCDMA, another ‘Chinese’ technology that was invented by foreigners, obscure telecom standards are a big deal in the PRC.

Back at the ceremony, Huawei was giving Arikan the full treatment.

As he arrived, a hundred Huawei staff formed a guard of honour and gave him a standing ovation.

Ren Zhengfei personally presented him with his award, engraved with Baccarat Crystal.

Several dozen employees also received prizes – not bad for a technology that’s been designated for just one fragment of the 5G standard.

It’s not Arikan’s only honour. He was selected for the 2019 Claude E. Shannon Award, named for the father of information theory and routinely described as the most prestigious in the industry. 

Arikan explained at a media briefing that the idea of polar coding had been around since the 1950s.

He says it was purely theoretical and admits he finds it “surprising” that it is now playing a significant role in real networks.

He worked on polar codes for his PhD at MIT, "solving enough problems” to receive his doctorate in 1985.

His supervisor was Robert Gallagher, known for his work on another error correction methodology, turbo codes or Gallagher codes, officially known as low-density parity check (LDPC).

LDPC is already deployed in LTE and, as it happens, was the competing technology proposed by Qualcomm at the notorious 3GPP meeting.

In a classic 3GPP compromise, polar code was designated for the control channel and LDPC was selected for the data channel.

Gallagher's work on turbo codes made him also a recipient of the Claude Shannon Award. He had in fact been a student of the legendary Shannon himself.

Arikan said post-doctorate he was back in Turkey but working on polar code theory in his own time, “just to understand it better.”

He says the less demanding environment of Turkish academia was more conducive to long-term research than the competitive US system. He also took on occasional consulting work that “gave me some engineering insights” into RF transmission.

“It took about 25 years to assemble all the pieces together," leading to the publication of his 2008 paper, he said.

His insights were widely seen at taking data transmission to the threshold of the Shannon limit - the theoretical maximum data rate.

Enter Huawei.

Tong Wen, Huawei fellow and wireless CTO, says the company had been looking for a long-term, high-risk, high-reward technology to bet on.

“It wasn't an easy decision. It would not be a one or two-year result," he said.

It's not the first Huawei technology to win 3GPP approval, but according to Tong it's the most far-reaching.

He and other execs believe polar coding could be deployed right across 5G, but they think in just allocating it to the control channel, 3GPP was showing caution about a new technology.

But they are highly confident about its potential.

"It's going to implement in a billion devices and eventually machines. This is going to be a fundamental technology," says Tong.