This is madness. It doesn't make any sense, whatsoever, to sell wireless frequency spectrum, except a quick injection of money into the government's treasury. It only promotes monopolies and pushes new comers out of the market.
In my humble opinion, wireless spectrum should be handled like a resource that can be auctioned at an exchange. That's essentially similar to how electricity is traded in some countries.
Say, you're a telco which needs spectrum in a crowded city, go to the exchange and buy a chunk for the next day. The price will be created by demand and cost of the infrastructure providers. This way, even new comers can auction cheap spectrum in more remote areas to offer service. Also, spectrum will be cheap at night and more expensive during crowded days. By having an exchange, ideally, prices will reflect the actual cost of infrastructure providers, which should be the basis to set retail prices for customers.
So my question is: What kind of principle is behind such an idea where one company could potentially own _all_ spectrum and deny other companies the possibility to offer anything wireless at all?
Your proposal makes very little sense in the existing world. Most equipment is setup for specific frequencies. It can't simply transmit and receive on any frequencies whatsoever. Same with the phones; and even more so with the towers. So, most providers have their networks on at least slightly different frequencies, which are often incompatible.
Additionally, such exchanges certainly wouldn't be free to run, so, there'd be quite a bit of administrative overhead in running them. What did the electricity exchanges in California brought us all? Also, do you really want to have varying phone reception every single day?
Legere's proposal makes a lot of sense -- require that at least half the spectrum is restricted away from AT&T and Verizon, and also require that all bought spectrum must be used within X years or XX months in real networks, or be forfeited without a refund.
Almost all phones nowadays come with a large range of frequency bands within which they can send and receive. Towers themselves are not the problem, AFAIK. The network infrastructure to the towers is problematic, though. If the infrastructure is available, you can put any antenna there.
I'm not aware of California's electricity exchange.
I didn't say Legere's proposal is bad. I'm all for it. But I'm still convinced that actually trading the right for spectrum at its current value makes a lot of sense. If your provider is too greedy to buy enough spectrum at an exchange (but it's okay to spend billions during an auction, I guess) then you can switch providers because, apparently, someone else is buying the spectrum at the exchange.
Another positive side-effect would be that infrastructure and service could be split. It'd make a lot of sense for carriers to make all phones work on all frequencies in order to be flexible which frequencies to use. This in turn would allow consumers to have more flexibility with their carriers without changing phones.
The article describes in perfect detail the shortcomings that are possible with all of these "exchanges" for finite resources.
Also, I think you underestimate what it takes for a tower to support a given frequency. Filters, antennae and lots of other stuff is custom made to specific frequencies. I've been told over at the T-Mobile US reddit that supposedly each frequency requires its own antennae.
Yes, your proposal would likely result in all phones supporting all frequencies (however, aren't you starting with the claim that it's already the case?). However, I think the shortcomings with the extra uncertainty, and the extra layer of having to deal with the exchanges and the extra real-time market analysis, will not help the consumers at all.
The reason I told you that I didn't know about California's electricity crisis is that I didn't have too much time on my hands to figure out what you meant. I'm perfectly capable of figuring things out myself, if I want to :)
No one would deny that there are issues with market manipulations and scamming at exchanges. That's why we have laws for exchanges and exchanges need to monitor the market for illegal activities. There were even one or two incidents in the history of exchanges (in this case stock exchanges) where things went so badly that a complete rollback to the state before those activities needed to happen.
An exchange needs to ensure the legitimacy of trades and it also needs to ensure market liquidity. But that is an issue that's inherent to exchanges. You wouldn't want all exchanges to close down just because California had such a terrible experience, or would you?
Yes, telecommunication equipment is very expensive, if you want to buy it for yourself. However, you also need to see the scale the equipment can handle. During its lifetime, a single antenna can serve millions of phone customers who all pay the telcos each month. In light of that scale, things are rather cheap. Just like a switch for 10GbE is way beyond the price what a normal person would buy for their home network, but a telco stacks them in their data centers because they can serve an awful lot of people with just a ten thousand dollar switch.
However, I do agree that there is uncertainty involved and if something like that would ever happen, a lot of care must be taken. Once everything is set up correctly and legitimacy and liquidity are ensured, the spectrum could be utilized to its full extend. If I want to found a company that wants to create some kind of wireless technology, I can only use the ISM-band. Right now, there's no way I could use anything else if I'm not already one of the top-dogs of telecommunications.
Further more, there is the issue that we actually have a lot of unused spectrum (apart from time square and the likes...). When a telco speaks of the need for spectrum, they talk about those highly crowded regions, but neglect to say that there is a lot of places where they just sit on their spectrum and don't use it to their full extend. In an exchange you could trade those frequency bands for various regions. If, however, a company owns a frequency band for all the USA, they can say: "See, I'm using it" although they only put it to use in the crowded regions.
So, you're basically coming from the perspective that these exchanges are meant to be useful specifically for the newcomers. But then this would never happen, since the whole industry is being lobbied by the bigger two to four players in the US.
Also, are you suggesting that a newcomer would just have a tower or two in a remote region? I guess that may sound interesting, and they could do roaming with T-Mobile US elsewhere, and the whole concept might open up new roads for new services not currently possible, however, I think the required oversight would be overly burdensome, and smaller operators with their own towers, and roaming with T-Mobile elsewhere, already do exist within the current infrastructure -- take a look at Union Wireless from Mountain View, Wyoming, or Viaero Wireless from Nebraska/Kansas/Colorado. Also, I think Verizon Wireless is know to share their licences to some operators in exchange for buildout and mutual roaming agreements.
Additionally, if you invest in equipment to support all of this, I'm pretty sure you'd rather be more certain that you'd be able to have the licence to the frequencies to support it; and not be subject to the overbids from the big players. However, it would certainly be interesting to know the perspective on all of this from operators like Union Wireless and Viaero.
In my humble opinion, wireless spectrum should be handled like a resource that can be auctioned at an exchange. That's essentially similar to how electricity is traded in some countries.
Say, you're a telco which needs spectrum in a crowded city, go to the exchange and buy a chunk for the next day. The price will be created by demand and cost of the infrastructure providers. This way, even new comers can auction cheap spectrum in more remote areas to offer service. Also, spectrum will be cheap at night and more expensive during crowded days. By having an exchange, ideally, prices will reflect the actual cost of infrastructure providers, which should be the basis to set retail prices for customers.
So my question is: What kind of principle is behind such an idea where one company could potentially own _all_ spectrum and deny other companies the possibility to offer anything wireless at all?