Korea 5G coverage 230

Open Signal users reached average speeds of ~230 Mbps on Korea's 5G, consistent with similar networks in other countries. I was surprised when Korean government testing showed much higher speeds, often over 600 Mbps. I normally wouldn't pass on anomalous results like that unless I had an explanation. However, it's an official source and none of the other testing is conclusive, 

Perhaps more interesting is the lower chart below, showing the three telcos felt they had to match each other's deployments. They provide an important lesson to regulators: never subsidize all 5G carriers. Find a way to get one to build and the others will probably follow, if they all have mid-band spectrum.

ZDNet reports, "The government made measurements in 11,000 crowded areas such as department stores, libraries, universities, amusement parks, hospitals, exhibition centres, subways, and terminals."

Read more ...

ziguangCoolpad and China Telecom have brought the price of a decent 5G phone down to US$199. It's 6.5", has three rear cameras and a 4600 mAh battery. It's the first to ship with the Chinese-designed Ziguang Zhanrui Tiger T7510; soon, SMIC in China will be able to manufacture it as well. 

Coolpad is trying to come back in a market where 12 or 13 phone makers are fighting for a market likely only profitable for a handful. Chipmakers are rushing to market with even less expensive 5G chips, like the Qualcomm 690 due by yearend. (Full details below.)

Huawei, Xiaomi, Realme, and Vivo have phones from $214-$260. Pictures below are from sale listings at jd.com

Read more ...

UNISOC T7510 230The new 1388 yuan (US$199) Coolpad X10 5G is shipping with the Tiger T7510 processor from Ziguang Zhanrui, a division of Unigroup. It will be featured by China Telecom. Previous 5G phones have gone on holiday sale below $200 but the lowest regular price was 1498 yuan ($215.) 

Ziguang Zhanrui announced the chip last year. It's apparently made on TSMC's 12 nm process, significantly cheaper than the 7 nm used for most 5G chips. It presumably sacrifices some performance but should be fine for everyday use. 

SMIC in China is spending $billions for additional 12 nm capacity, so the chip could be produced in China in the near future. SMIC will be able to produce some 7 nm chips with DUV lithography, probably by the middle of 2021. As long as the U.S. prevents delivery of the Dutch ASML EUV machine, SMIC and other Chinese fabs will still be behind.

China has a crash project to make its own EUV gear. The common wisdom is it would take a miracle to catch up in less than 5-10 years. China has delivered many miracles in the last decade.


Read more ...

Raluten 2305G software isn't ready yet, but Tareq Amin now manages 5,000 4G cells running on a fully virtualized network that is leading the world. 7,000 more are under contract and 70% of the country will be covered early next year.  

Rakuten is an existence proof of a fully virtual network. Almost everyone agrees the future wireless network looks like this, but Amin is the first to build one. Next, Rakuten wants to sell similar systems to telcos around the world and has hired a U.S. representative to make that happen.

This hasn't been easy. He had to pay "hundreds of millions" for custom chips because the off the shelf hardware wasn't fast enough. Vodafone CTO Scott Petty also has a few working Open-RAN systems, but doesn't expect to bring them into his main network until about 2025.

They said it couldn't be done but they were wrong. 

 Thanks to Sue Marek who pointed me to the data.


In When LTE is faster than 5G, respected journalist  points out that DT's 5G DSS loses 15% of capacity to additional signaling bandwidth. The result, in theory, will be less capacity in the same spectrum than if used purely for LTE. He quotes a DT engineer

"A problem with dynamic frequency release is the waste of resources, because in addition to the 4G signaling channels, 5G signaling channels must also be sent. The signaling effort in a 10 MHz channel is currently around 15 percent, leaving 85 percent of the channel for the transmission of user data. 5G has approximately the same overhead, reducing the available bandwidth for user data to about 70 percent of the channel. "

Max Planken adds

"5G currently has virtually no advantages over DSS. It is all just marketing. If the network operators used pure LTE instead of DSS with LTE / 5G, the maximum achievable speed would be even higher."

Pia Habel of DT contends

"It is a fallacy to conclude that this is 'only marketing'. 5G is the target standard in the long term."

But she does not dispute that the 5G using DSS is slower than 4G LTE would be in the same spectrum. Nor does she suggest that the 5G has a significantly lower latency. 

5G in bands of 2100 MHz and below may ultimately be slightly faster than 4G, perhaps 20%. My source for that figure is Neville Ray, CTO of T-Mobile USA.

Today, 4G is likely to be faster. Half a dozen technical features, especially LAA, are not yet available in 5G. LTE can comfortably aggregate 4 carrier bands, which is not yet possible in 5G.


China shipped 63 million 5G phones domestically in the first half of the year, including 17 million in June. The phone figure is from CAICT, a government agency with no reason to underestimate. That's a remarkable achievement, more than twice as high as the total in the entire rest of the world. (15-25 million, details soon to follow.) It does not need puffing up. China is on track to easily reach the 150 million plan for yearend 2020.

However, China Mobile reports 70 million "5G contracts" up almost 15 million from May and almost 40 million since March. China Telecom reports 33 million, up ~8 million from May and 21 million from March. China Unicom doesn't report 5G figures but is about half China Telecom. "Contracts" are ~115 million. 4G & 5G contracts are the same price, so apparently many people are choosing "5G" contracts even if they don't have 5G phones.

One side effect of the discrepancy is that some figures for the global number of 5G subscribers are inaccurate. A reputable analyst firm put out a figure of 63.6 million 5G users for March, which was picked up by many. I'm almost sure it included the inflated figure of "contracts," as I pointed out to them. Word had circulated in the industry of the discrepancy before they released the figure, so I was surprised. The adjusted global figure for March should probably be ~45 million, 

I was surprised they would not back up or correct an apparent error. "I make many mistakes," the butler said.

Currently, I'm researching whether I had a major error thinking mid-band 5G would be a major boost to capacity. We're starting to get data suggesting that indoor performance is awful. With indoor a large majority of wireless, if the results are confirmed the capacity boost from mid-band 5G will be much lower than in some analysis reports I've written. Data very welcome.   



Good 5G phones in China cost US$230-260, Monthly service is mostly $13-23, the same price as 4G. 60% of phones sold in China are 5G, a rate going up every month.

5G promotion is ubiquitous, with almost all tech pubs doing several raving articles each week. Most are bs. 5G just isn't good enough to have the claimed impact on industry and the economy.

According to official CAICT figures, below, 63.6 million 5G phones shipped in the first 6 months of 2020, as well as 10-15 million in 2019. (Update July 22: Review of 2019 figures suggest this number, from the telcos, may be inflated) Some are in dealer stock or transit, so the actual number of 5G users at the end of June is 65-70 million. Adding 17 million each month from July to December would take the year-end total to about 170M, three or four times more than the entire rest of the world.

All 5G in China today is mid-band, with typical download speeds of 100-500 Mbps. China Mobile has 160 MHz of mid-band spectrum. The joint network of China Telecom & China Unicom has 200 MHz. Western carriers, except for T-Mobile, have 100 MHz or less, hence lower speeds.

Background: Last June, Minister Miao Wei told the telcos to "accelerate" 5G. This is putting China so far ahead of the West it has already won the (almost meaningless) "race to 5G." It also has been a shot in the arm for Huawei, which increased sales the first six months by 13% despite the U.S. blockade. 

Read more ...

Opensignal US 5G 230Canadian average speeds 2019 open signal 230The average speed of over 10,000 Open Signal tests of U.S. 5G was under 60 Mbps.* The average speed of 4G at Bell Canada last fall was 70 Mbps. At Canadian #2 Telus, it was 75 Mbps. The methodology was essentially the same. The results are consistent with testimony by T-Mobile's CTO Neville Ray; Verizon's Glenn Wellbrock; and just about every engineer building networks. (A CTO has just turned on 5G at a major carrier and tells me off the record their 4G is currently better.)

Ray testified to the FCC that in theory 5G could be 20% faster in low-band and 50% faster in mid-band. For now, 4G is far ahead until 5G adds LAA and other tools already working in 4G. Wellbrock pointed out that in these frequencies, neither 5G NR nor any other software can add much to capacity. 4G is close to the Shannon Limit already. Higher performance without more spectrum or antennas would break the laws of physics. 

Next time a pundit or politician talks of the very high speeds of 5G, ask them what is the actual speed of the 5G being deployed? Be ready with those numbers. 

Open Signal "found no significant improvement in 5G latency over 4G." It will be years before 5G low latency is widely deployed - and 4G will be almost as low by then.

Read more ...

More Articles ...

dave ask

@analysisbranch for latest updates


Welcome  Asia is installing hundreds of thousands of 5G radios and adding 5G subs by the tens of millions. The west is far behind. 200,000,000 in 2020

The demand is there, and most of the technology works. Meanwhile, the hype is unreal. Time for reporting closer to the truth.

I'm Dave Burstein, Editor. I've been reporting telecom since 1999. I love to hear from readers and say thank you when you find an error. daveb@dslprime.com

Also see