
But not EVERY competitor will be allowed to compete..
.. if I look at what happened to Huawei.
It all looks wonderful until you scratch off the pretty paint and look underneath.
The National Institute of Standards and Technology (NIST) wants to prevent supply chain challenges and vendor lock-in that might get in the way of US carrier's 5G rollouts. This week the agency, which operates under the US Commerce Department, threw its weight behind the O-RAN Alliance. The industry consortium, which is made …
Huawei's cardinal sin was owning about two thirds of the patents needed to implement a standards based 5G. In the normal course of things this wouldn't be a big deal since the standards process uses a patent pool so that the playing field is level for all participants and royalties are capped at reasonable, predictable and stable levels and shared equitably among the participants. It was was messed up by government action -- when the US took action against Huawei it was at least two years ahead of other players and ready to ship both network and subscriber equipment.
The position that Huawei was is wasn't unusual. I believe that Qualcomm had the same position with 4G. Anyway, the result for us in the US is that we don't have 5G, we have a mismash of incomplete, vendor specific solutions that are more a product of the marketing department than the engineers. This is why customers report disappointing service with 5G, it rarely, if ever, improves performance over 4G. So having held Huawei back we congratulate ourselves on winning the battle. But the war is well and truly lost.
For sure, not every O-RAN vendor is going to be welcome - you can imagine that the triopolistic state which has prevailed for many years will continue to do so (H will be replaced by someone else...). But there is a curious underlying situation with O-RAN in that it is a bandwagon that everyone is jumping on as it is a pseudo-open standard which anyone can create a product for. I imagine that the 3GPP working groups will feel a little insulted that the first standard that's publicly slated to solve all communications woes isn't something that they've toiled over. Having successfully integrated homebrew and commercial RAN with A/A-bis (GSM), Iu/IuH (UMTS), S1 (LTE) and now Nx (NR) which have used the (extremely well written) 3GPP standards, I am astounded at how much ground O-RAN is gaining.
If you have a site on a ship, an aeroplane, a truly rural location or, in fact any non-dark fibre connected site you can forget O-RAN. For stadiums, conference centres, dense urban then O-RAN has real potential value (although I would argue that a well coordinated low power S1/Nx connected LTE/NR site array could do a similar job). Whether the small O-RAN contenders get a chance to play in these environs remains to be seen however. For now, I am happy that I can plug our "nO-RAN" LTE radio into a Starlink terminal and pass traffic through it within 30 seconds and no-touch config.
Is everyone STOOOOOPID! You take some modern AMD Ryzen-7 or Ryzen-9 Chips which already run as high as 4 to 6 GHz in clock speed and you INTERLEAVE at least TWO of the chips at half a Hertz apart to sample a waveform at the require 8 to 12 GHZ for 5G waveforms and do EVERYTHING to demodulate the Wifi/4G/5G and Starlink Signals in software! The CPU chips can input AND output a digital representation of the modulated 5G waveforms on their PICe lanes which the antenna circuits can smooth out using CHEAP analogue circuits to send out into or get from the ether! Each CPU chip is capable of 24 lanes on the AMD Ryzen-7 up to 128 lanes on Threadripper or EPYC chips at anywhere fom 16 Gigabytes Per Second up to 64 Gigabytes per second which can easily represent ANY analogue or digital waveform of ANY modern Wifi, 4G, 5G, 6G or Starlink internet data communications systems!
No need for stupid patents when EVERYTHING cna be done in FREE AND OPEN SOURCE software using common DSP algorithms running on ANY modern AMD Ryzen-5/7/9, Threadripper or EPYC CPU chip using antennas attached directly to the PCIe lanes with some inbetween waveform smoothing alagorithms. The PCIe lane signals on input don't have to be a digital waveform as the incoming analogue waveforms can STILL be digitized into the desires communications standard modulations using common DSP algorithms built right within the PCI-e circuits! That is done using common microcode.
Why pay $500,000 USD for a carrier-class cellular 5G switch/hotspot when EIGHT x.499 motherboards with their PCIe lines directly hooked up to antenna circuits can handle over 512 users of full-duplex 4G/5G/6G I/O per cell at less than $10,000 per cell. AND it's instantly upgradeable to any NEW or revision of a major Comms standard just by updating the flash BIOSes!
There! I just saved you $490,000 USD in cellular 4G/5G/6G hotspot/switch costs!
V