Hacker News new | comments | ask | show | jobs | submit login
America Needs a Nationalized 5G Network (wired.com)
49 points by jonbaer 11 months ago | hide | past | web | favorite | 53 comments

For wireless carriers, the three highest opex costs are 1. Real estate (lease costs, no one owns their cell sites) 2. backhaul, which today means fiber 3. employees

Nationalizing the 5G roll out will not help with any of these costs, the real estate costs are more and more coming from local municipalities or other local entities to place in the right of way which is probably the only realistic option with the densities needed for 5G. The Federal government does not have the power to take those rights of way. This constraint applied really to two of the big items, real estate costs for the cell sites and right of way for the fiber.

The reason countries like China or others like it can more easily deploy a nationalized network is because the central government has total dominion. We don't have that system here so a federal approach will not reduce costs, if anything it will cause them to spiral as local entities will see that as a cash register from the federal side.

And we all understand that a federal work force is not going to be a low cost option for the third component of the equation.

For these reasons I disagree with a federal approach. To my way of thinking, if local entities, cities and counties, want 5G wireless broadband, the best way is to make rights of way available to carriers in a fair and accessible way without trying to over play their hand and jack the costs up which inevitably will just price it out of reach of more and more people.

The capital outlay to deploy these networks is fairly significant and by making a reasonable accommodation for the carriers will help offset that.

The federal government is not a good answer here as much as it may seem like the quick and easy answer. All it takes is a quick review of recent performance of congress to realize how unrealistic that notion is.

The federal government certainly has the ability to use eminent domain to acquire property, particularly with a (specious) national security angle. Nike-Hercules missile sites are an example of what can happen when the Feds are motivated. The interstate highways are another.

This is a fairly obvious play at rebuilding the AT&T monopoly of yore. The Feds would just pay a carrier to run the thing, perhaps by using an existing contract vehicle (maybe first net?).

It might be theoretically possible, or legal, but it would be hugely controversial and would be in the courts for sure. I do not think the FCC could do it with a rule it would take an act of congress which seems pretty unlikely in the current political atmosphere. It is not clear that the federal gov could use eminent domain to force cities and counties to relinquish rights of way, perhaps, but it would be pretty unprecedented in peace time as you point out.

Interestingly, the former bell system was nationalized for a brief period in 1918 and run by the post office (an arrangement that some European countries have today) https://en.wikipedia.org/wiki/History_of_AT%26T But it did not last and was a complete disaster. I am sure a redux of that would end in tears.

In the early 1980's the Bell System had something called ISDN which purported to be something like what we call the internet today. They had a plan to roll it out (64Kbps !) nationally over a 20 year time frame. Had not AT&T been broken up in 1984 I posit that the internet we know today would have never emerged, it was free enterprise and entrepreneurship that made it happen.

> hugely controversial and would be in the courts for sure

Controversial like illegally wiretapping every single citizen in the nation for years? Controversial like the current President? The bar is so low on controversy and legality that it practically doesn't exist.

> It is not clear that the federal gov could use eminent domain

That sounds like an opinion but even if it were true it doesn't mean that the threat of using eminent domain couldn't be used as leverage in negotiations.

For wireless carriers, the three highest opex costs are 1. Real estate (lease costs, no one owns their cell sites) 2. backhaul, which today means fiber 3. employees

It also depends on country and location, but from what I remember, most of it goes to lease.

1. Lease on the Cell Sites. And that is why in certain places, ( or all places ) property pricing has an very large effect on Mobile Network pricing.

2. Retail. Servicing the Customers.

3. Lease for Backhaul. Increasingly Mobile Carrier is either part of a subsidiary of local ISP.

4. Lease for carrier equipment. Well we are only left with a few, Huawei, Ericsson, Nokia, and ZTE. Increasingly these equipment are not part of the carrier asset but services agreement.

All the four things above made up to 70% of the cost structure.

The reason countries like China or others like it can more easily deploy a nationalized network is because the central government has total dominion.

Yes, there are no issues ( well relatively little ) with placing cell tower where they want, rent etc. If it the party direction, no one would (dare to) refuse it. You have cost structure issues? They have China Tower which shares the rural area signals, so the three carrier dont have built each of their own.

For wireless carriers, the three highest opex costs are 1. Real estate (lease costs, no one owns their cell sites) 2. backhaul, which today means fiber 3. employees

Nationalizing the 5G roll out will not help with any of these costs, the real estate costs are more and more coming from local municipalities or other local entities to place in the right of way

Doesn't this situation put a moral hazard on municipal broadband? If a city runs the broadband, why wouldn't they use their government power to keep out competitors and maintain a monopoly?

That type of thing actually already happens, I work in the industry and I have seen it happen. A state run governmental fiber network was put in place and the government bodies prevented other carriers from placing fiber and forced the wireless carriers to use their fiber at a premium.

Which state? I can think of one off the top of my head, but the story doesn’t match up.

interstate commerce clause has lead to federal government passing sweetheart legislation for wireless operators, severely limiting local control over the placement of micro-sites like the kind needed for 5g. https://www.fcc.gov/general/tower-and-antenna-siting

Huge federal overreach.

That's right, but those rule prevent cities from making rules that prevent any use of the airwaves. Nationalizing a 5G broadband network means taking the rights of way away form the cities. That could not be don't without paying for it. That's not going to make things any cheaper, at least any time I see the federal gov pay for things, it does not make it cheaper.

> The Federal government does not have the power to take those rights of way

This wouldn't be a discussion if they didn't.

They do not; not with out paying. That would be a taking. They have to pay to use the city rights of way just like carriers do. Just because the federal gov is doing it does not mean it is free. In China or other authoritarian governments they just say we are putting fiber here and a tower there and if anyone complains they can just put them in prison or whatever they want. In the USA that is not how it works. The federal government would have to buy the right of way just like carriers do. I don't think that will make it cheaper or faster to do.

Having one network instead of four should reduce all those costs, although I agree about the other roadblocks.

I could certainly reduce costs, but that would also mean no pressure to reduce prices or invest in improvements.

This is essentially the situation we have in many markets where there is only one or two ISP choice(s): mediocre (or worse) service and inflated prices.

New Zealand is currently in the process of rolling out fibre connections in almost every urban area of the country (target is 87% of households) and the amazing thing is that a basic fibre connection is now the same price as an ADSL/VDSL connection.

The government put together a technical specification (e.g. minimum speeds, networking standards, API for service providers) and set a series of coverage targets; then they solicited bids from the private sector to build and operate the network in each region. Funding was in the form of an interest-free loan, and the network operators are required to offer wholesale access to any retail provider. The operators are also prohibited from offering their own retail services.

So in the end, once the loans are paid back, the net cost to the taxpayer is likely going to be less than NZ$1.5bn -- that's gigabit fibre to the home for every non-rural household at a cost of less than NZ$1,000 each!

If you are going to use New Zealand an example of what to do, then use Australia as an example of what NOT to do. Previous discussions about it https://news.ycombinator.com/item?id=14320121 https://www.nytimes.com/2017/05/11/world/australia/australia...

I've just recently moved from New Zealand to Australia, and I'm shocked at how terrible my internet is here, it's like stepping back a decade, not just poor connection speeds, but poor connectivity as well, it keeps dropping out.

I would be very, very upset if I were forced to pay 750 USD for a fiber connection that didn't even benefit rural households.

That's how government works. Most government programs don't directly benefit every individual citizen; even though they all get funded from the big pot of 'government spending', which is filled by taxes.

If you think there is a systematic problem, where a disproportionate amount of money goes to urban centers, and a disproportionate amount of money comes from rural areas, then you have something to complain about; but you shouldn't complain about a single project just because it does not benefit 100% of the population.

Eh, if it cost $1250 per household to light up every home in the country, you can bet metro folks would be carping about how much they are paying to cover the much higher per-household costs of rural users.

The most Republican Libertarian TV station on cable is Rural Free Delivery TV named after the United States connecting farmers with free rural delivery. They now hide the history of where the name came from.

But the infrastructure (underground ducts, fibre junction boxes, and the GPON fibre) will last for decades, so it's a very reasonable investment to be making.

Infrastructure costs money, who knew.

$750 for installation to the premise is really, really cheap.

That’s nothing. You probably pay 10x that for highway capital costs.

The benefit of an investment like that is incredible.

> cost of less than NZ$1,000 each!

Yes, of course the net cost per household is low when you count 100% of households. Of course you can only do that if you have a military to force people to buy your product whether they want it or not. Other than that it is a perfectly wonderful business model.

Can you please stop posting generically ideological and/or unsubstantive comments to HN, especially the flamebaity kind? You've been doing it quite a bit lately, and it's the wrong sort of comment for this site.


So basically the wireless companies don't want to pay for all of the real estate they need for the new base stations they have to deploy and so they want to use eminent domain. The free market is great till you have to buy something I guess.

I would propose doing essentially the exact opposite; commodify the entire spectrum using a high-frequency rolling regional auction system, so anyone and their dog can become a wireless carrier. This would be vastly more cost-effective for rural areas, and would offload the effort of organizing fiber leases onto random small (regional) businesses providing bandwidth.

Aren't equiptment costs still prohibitive?

Nope; check out open-source LTE tower projects. Surprisingly cheap these days! Totally within range of individuals, especially for the kind of hardware that only needs to serve rural areas.

With the impending mass cheapening expense of launching satellites via SpaceX and the like, wouldn’t a large array of telecommunication satellites be a more reasonable option for nation wide coverage in a country the size and density of the United States?

Possibly, which is why SpaceX wants to do exactly that[0]. They're launching their first test satellites shortly[1].

[0] https://en.wikipedia.org/wiki/SpaceX_satellite_constellation

[1] https://www.geekwire.com/2018/spacex-gets-set-launch-first-p...


"Between here and the satellite, there are a very large number of nanoseconds"

America needs good and enforced regulations. My country had nationalized network and it was a disaster.

If we’re saying that the free market isn’t capable of delivering the infrastructure that underpins much of our modern economy, then where does that leave the idea of capitalism?

Capitalism is great because it’s been the best tool we’ve had to massively improve quality of life in many different areas of life. That doesn’t mean it’s the only tool we have and/or it should be an end in itself.

The situation where capitalism is not the best tool for the job we should abandon it for that job. That will not only mean we have a better standard of living, but will also mean the idea of capitalism gets stronger because it doesn’t suffer from negative backlash when it fails because it is used in situations where it isn’t the right tool because of ideology.

There is possibly an analogy here to programmers who insist a certain technology be used everywhere, even where it isn’t a good fit, instead of opting for the best tool for the specific application.

"The situation where capitalism is not the best tool for the job we should abandon it for that job." Soooo healthcare, food distribution, housing, transportation, resource extraction, education...

Capitalism isn’t all or nothing and government expenditure on infrastructure isn’t incompatible with capitalism.

There are certain markets (telecom, healthcare, mass transportation) where completely “free markets” don’t exist, because they can’t exist, because the nature of the products and consumers are such that one or both violate the criteria for which a free market can exist (lack of perfect information, no competition, etc.)

The free market can not create a national telecom network because it can not force all people to pay into it, which is necessary in order to have it built in the first place. The only reason we have the internet as we do is because AT&T was mandated by the government to install phone lines to everyone, everywhere, and charge everyone a fee for decades to pay for it. By its very nature there is no competition and no choice. The old Bell System was a government sanctioned and funded monopoly and no free market would have ever created such a network by itself because that’s not what free markets do. They simply allocate capital effectively in markets which are compatible with the criteria necessary for that kind of allocation to be possible.

I suppose the same place we were left when figuring out roads, water mains, water mains connections, telephone and electricity initially, emergency rescue, etc?

Better off?

We have a mixed market economy, like we've always had. It says nothing about anything.

Capitalism is one of those things that only works in theory. One only has to look at the countless deaths and massive wealth inequality caused by it to realize it's unfeasible.

We tried this. It was a national broadband initiative in the late 90s. Telecom companies were given BILLIONS of dollars to build fiber across the country. They took it and did exactly zero.

If we are dumb enough to do it this way again (paying the wolves to secure the hen house), we better have good checks and balances on it.

Separating concerns prevents these problems. There should be one company responsible for the infrastructure and other companies competing to provide service to customers.

Deregulation of the Texas energy market should be the model. In Texas, I get an electricity plan that is 100% wind power at a third of the cost of electricity in California. They even gave me a free digital thermostat and installed it for free. The service is reliable and gets restored quickly if there's a natural disaster.

If only telephone service and internet access worked this well.

But I agree, if you give billions of dollars to customer-facing companies to improve the infrastructure, they'll squander it.

You aren't really getting 100% wind power since the grid doesn't work like that. You are, however, promoting more wind energy into the grid.

Yep, and that was a disgrace for everyone involved, and screwed the consumer. There's two pretty big indicators that we're doing it wrong. First, with few exceptions, USA pays among the most for internet data, at middling speed. Secondly, those providers are consistently ranked among the least liked by customers. Why this doesn't scream monopoly/oligopoly needing intervention is beyond me. I personally believe these companies are so disliked that even the biggest pro capitalists would not even care if it were nationalized.

The ISPs, their lobbyists, the politicians whose campaigns they donate to, and the revolving door of government regulators who were and will be employed by said ISPs definitely care.

The proposal is to build public infrastructure and lease it to private firms, not to pay private firms with public money to build and own infrastructure.

When people talk about this "nationalized" 5G network, it's very likely that it would be outsourced to a single prime contractor (probably either AT&T or Verizon). The government has actually done this before with FirstNet.

Which also has potential for abuse and failure

A nationalized 5G network will be essential for high-bandwidth applications like V2V across level 5 autonomous vehicles. As has been done in the past for large scale projects like GPS, LEDs, photovoltaics, and nuclear (just to name a few!) which private companies wouldn't be able to eat the long-term capex against. These technologies are also licensed either for free or very cheaply to companies in order to benefit the public at large since we've already paid for the R&D with tax dollars.

In order to prepare for the future of the digital economy at large, bandwidth shouldn't be an artificial limitation of it. Once all the associated RFCs for 5G are standardized, a rollout of this scale would most definitely pay for itself and more over the network's useful lifetime.

I expect that l5 autonomous vehicles will be able to function without any internet connection.

I'd be afraid if my car required cell service to function.

While there isn't an official RFC yet since the protocol is still in the RFP stages[0], the amount of new cars on the road will require roughly about the spectrum allocation as what's needed for cellphones right now. If we tried to cram all of this into existing 5GHz range of lower frequency spectrum, it's simply not enough. 5G has both a larger spectrum allocation and throughput of data available to it.

Edit: The cost of putting in a lidar system for your car won't be too expensive (probably <$1000 when done across a manufacturer's whole fleet); but think about having to pay for another cellular line on your phone bill at current telecom prices. It would be prohibitively expensive to the end users for mass deployment.

[0] https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/readiness-of...

That ship sailed. V2V is a thing that will start showing up in 2023 in the US. It’s a mandate that is being accelerated by federal rule making as the gov was dissatisfied by the advancements being made.

The tech will eventually do things like communicate road conditions, allow lane prioritization, speed limits, etc, even for human driven cars.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact