Hacker News Reader

Comment by asciident | Hacker News Reader

Everything I've heard about it indicates that it's a trash technology. Unless you're standing next to an emitter, it's no faster than 4G and potentially even slower. Like you walk one block or behind a wall and it's not good anymore. So you gotta put one of these up on every block. I'm banking on it being a complete dud, like 3D televisions.

View on HN

You have heard wrong - it's amusing that you're so confident in its failure despite having the wrong information.

There are two phases in the 5G NR standard. The first phase relies on some elements of the LTE control plane to coordinate data services. These data services use new numerologies for their OFDM modulation, vastly more flexible resource block, allocation, improved error control coding, lower maximum latency, better channel allocation, and standardized MIMO support - all of which will improve reliability and throughput, especially in crowded channel conditions.

Phase 2 formally introduces millimeter wave signaling - this is what does primarily propagates through line of sight (LOS) channels. Clever beam forming and beamsteering can alleviate this to an extent but these cells are intended for dense urban environments or interiors of large buildings. These will supplement the lower frequency bands when available but 5G NR devices will still use the lowband when mmWave is unavailable. Phase 2 also switches over to an entirely 5G NR backed control plane (Standalone, or 5G NR SA) which has its own reliability and throughput improvements though these will all happen in the lowband since mmWave isn't suited for these tasks.

Do you work in the industry?

Not for the big telecomm's but I design parts of the radios and DSP algorithms implementing the major cellular technologies.

What are your thoughts on the safety concerns?

Not my area of expertise. The main study I'm aware of showing some harmful affects uses an unrealistic amount of power and curiously only showed effects for male mice. I'm hesitant to dismiss it, because effects from heating caused by non-ionizing radiation are well documented but that amount of power is just not present in cellular systems whether that's LTE or 5G NR.

I worry in particular about the towers being in close proximity to schools and housing developments.

Do you own a microwave oven?

More appropriately, do you operate a high frequency radar gun?

It really is the 1980’a again...

Why don't you demand actual reproducible evidence before worrying? Nonionizing radiation has not been shown to be harmful, but self-inflicted psychological stress certainly has.

> only showed effects for male mice.

Interestingly, it might not be the experiment but rather the experimenters themselves. When this study came out, I mentioned it to my girlfriend at the time (a veterinarian), and she pointed to research [1] that suggests a correlation between the presence of male researchers and stress in mice. It's plausible that if male researchers were present, the increased stress response could contribute to greater susceptibility to certain cancers (it was only the male mice in the 5G study that were affected).

I have no idea how the study was done or the genders of the researchers who participated, but it's worth considering that other variables may be at play. Perhaps it should be replicated under an all-female research environment if it wasn't originally.

[1] https://www.nature.com/news/male-researchers-stress-out-rode...

Worry about your blender. Seriously. Big electric motor.

Deploying 5G mm-wave "everywhere" seem like a huge expense.

How will telecomm's justify that ? what do you see as the big use cases ?

it's amusing that you're so confident in its failure despite having the wrong information.

Why would that be "despite"? That would have to be being confident in its failure despite having information that it would be a success. Surely they're confident in its failure because the information they have says it will be bad?

And then your comment becomes "I'm laughing because you're confident in your beliefs" which is content-free smugness, or "I'm laughing at you because you don't know what I know" which is more or less the same, if not worse.

Being confident enough to trash-talk based on hearsay is just being arrogant.

"New numerologies"? This is supposed to inspire me with confidence?

"In the context of 3GPP 5G standardization contributions, the term numerology refers to the configuration of waveform parameters, and different numerologies are considered as OFDM-based sub-frames having different parameters such as subcarrier spacing/symbol time, CP size, etc." (source: https://red-colmena.com/f/JyY1ajY1JmhhZSFpdC91JGcmIyEsYGU/18... which I found with a quick web search for "OFDM numerology")

> Phase 2 formally introduces millimeter wave signaling - this is what does primarily propagates through line of sight (LOS) channels. Clever beam forming and beamsteering can alleviate this to an extent but these cells are intended for dense urban environments or interiors of large buildings.

Thanks for all the info. Why don't we just skip phase 2 and continue using wifi?

Fair point, my thought too. Something tells me "big telco" is too valuable to economies to sidestep with a public service/open mesh network relayed via routers and phones as nodes.

That’s an open technology. How do carriers sell that?

Maybe you'll be interested to hear from actual 5G users before making uninformed claims. I have home wireless broadband in central London (no other good broadband offering in my street) and was upgraded from 4G to 5G last week. I went from an average of 5-10Mbps at peak times and 20Mbps max on 4G, to 150-250Mbps peak time and 350MBps max on 5G, on the same carrier (Three). Enough said.

How did you manage to get upgraded?

It's available commercially. http://www.three.co.uk/5g

Is it? I don’t really see how that contradicts what op was saying.

But does it matter for you? I know it would not for me. I can certainly find counterexamples, but for me the last leg bandwidth is seldom the bottleneck. My 2c.

Back in the day I didn't think upgrading from a 56K PSTN modem to DSL broadband would matter that much to me. I was naively thinking in terms of downloading things a bit faster, the classic futurologist mistake of seeing things through today's frame of reference.

What happened instead is that it enabled Youtube, Skype, Facetime, and that did change the world.

Maybe it just works fast because there still are only a very limited number of users? And will speeds and latencies drop as the airspace gets more saturated?

I believe 5G handles congestion much better. But even if the speed drops, it will still be an order of magnitude faster than 4G. Absolutely worth it.

In short - some but not really. 5G NR allows for extremely efficient resource block allocation and can handle ~10x more concurrent users than 4G. The latency should not degrade as long as the carriers don't do weird things with traffic allocation - the standard allows for ~4ms worst case latency.

How is real-world throughput across the internet?

I have a gigabit pipe that works just fine on a speed test, but most real-world usage doesn’t go that fast because of bottlenecks elsewhere.

I have 150Mbps right now (checked on fast.com) with 4G in Central Denver. It’s not about the number before the G, but whether the carriers are building enough infrastructure to support it.

It is definitely about the number before the G. With perfect infrastructure investment like what your carrier seems to have done with 4G in Denver, 5G could reach 1-2GBps. In other words, real world imperfect 5G is ten times better than 4G with all stars aligned.

It doesn't look like like your old "4G" was legitimate full performance LTE. Cell carriers lie and demur about what performance they actually provide. You can't trust the number to mean anything.

I don't know much about the specifics behind 5G but if asciident's claim is true then towers on every block should perform very well under high load. What kind of performance are you referring to? Throughput? Latency? Availability? Reliability?

It was definitely a busy area and dozens of Mbps are more typical on 4G. My point is that real world 5G is an order of magnitude faster than real world 4G, much like 4G was in comparison to 3G.

Your 5G, and your 4G. That doesn't mean it is the 5G that is the difference, unless you did an exhaustive root cause investigation of the poor performance of the 4G.

Backhaul matters as well. If you do carrier comparisons at scale, it becomes obvious that the connectivity to the towers varies dramatically, as do network conditions, quality of proxies on the carrier network and other factors.

A fancy new 5G tower also has underutilized connectivity to it.

The population density in London is an order of magnitude higher than in Denver. That has an obvious impact on air congestion, which is one of the primary limiting factors on cellular network performance.

That sounded high, so I checked:

London Density 14,670/sq mi (5,666/km2) (https://en.wikipedia.org/wiki/London)

Denver Density 4,519.94/sq mi (1,745.15/km2) (https://en.wikipedia.org/wiki/Denver)

London has a population density 3.25x that of Denver.

Why exaggerate when the facts are on your side? Or more specifically, why make an explicit numeric claim without confirming that it's correct?

"order of magnitude" is literally the least explicit you can be without just saying "lots bigger". Why be overly pedantic?

I think the problem may be that "order of magnitude more" means different things to different people. Colloquially, it does get used to mean "a lot more", and some people use it because it sounds more scientific. But in a more correct sense, it means "closer to 10x than 1x or 100x": https://en.wikipedia.org/wiki/Order_of_magnitude#Definition.

I personally dislike the phrase and try to avoid it. But if one is going to use it, I think it should be used correctly.

(I agree with jdietrich that the definition of "city" matters)

That depends very much on your definition of "London" and "Denver". Your first figure refers to the population density of Greater London, a vast area covering much of what used to be the Home Counties. Your second figure refers solely to Denver City rather than the broader Denver-Aurora-Lakewood metropolitan statistical area, much of which is functionally suburban Denver.

I don't think it's particularly useful to attempt to pin down a fixed definition of the boundaries of either city, particularly one as amorphous as London. Suffice it to say that London is densely populated, Inner London is even more densely populated and rush hour on the tube is genuinely insane; Londoners regularly encounter densities that simply don't occur in most US cities outside of a football stadium and have a subsequent need for far higher throughput than 4G can deliver.

"Maybe you'll be interested to hear from actual 5G users instead making uninformed claims."

Claims from "what I've heard" /are/ informed claims; your anecdotal information doesn't automatically beat other people's anecdotal information.

I appreciate your perspective but I did base my comments on what actual 5G users told me. Here's a video review that captures about what I said:

https://www.youtube.com/watch?v=_CTUs_2hq6Y

But I will add your data point to my opinion of 5G.

Reply on news.ycombinator.com