Fabrice Bellard, who is quite famous on HN for his open source work, has founded a company for his own 5G implementation. And they are EU-based. [1]
I'm under the impression his work is closer to the "radio protocol" than the infrastructure management behind. Not sure if they are related to Open-RAN though...
Recently Nokia strongly communicates its commitment to Open-RAN. They want to be like Tesla or Toyota (in case of hybrid-drives): so good that competition is unable to keep up even after opening their patents. Once telcos grasp the benfits of open-interfaces infrastructure (eCPRI and stuff), there is no coming back to closed ecosystems.
The "radio protocol" is "open" since GSM, anyone can download standards from 3GPP and implement it accordingly. But the mere amount of knowledge and specialized hardware required to do this, even for single layer like L1, is tremendous. I think this is the real reason why we don't already have open-source implementation of the full stack.
EDIT: an afterthought - maybe the O-RAN is really a chance for open-source here. In the future, once O-RAN is accepted and widely deployed, we could work on implementing the stack piece-by-piece, layer-by-layer, filling the gaps with commercial software/hardware as we go, instead of doing everything at once...
The 3GPP standard leaves a lot of implementation freedom in many areas and there is a huge number of patents on everything. So in practice licensing is a big issue and keeps new entrants at bay.
Nokia thinks it is in good shape because without Huawei there is indeed not many threats. Starting from scratch is hugely costly and takes many years. In any case Nokia has no choice but to be "committed" to Open RAN since that's what telcos want.
IMHO, Open RAN is a push by telcos to commoditize the infrastructure and to avoid being locked in because key interfaces are proprietary.
In the UK where Huawei is on a phase-out, Open RAN is a bet by operators to ensure that they have more than 2 vendor options. They are required right now to use 2 vendors, so they are not solely reliant on any one company for their equipment.
Most operators used 2 radio vendors, in order to keep things competitive when buying more. All have exposure to Huawei to
some extent in their networks.
If there are only 2 permitted credible radio vendors, then pricing on the "new" second vendor is hardly going to be competitive - where else will you go? You can't go with one vendor, and if you did, their pricing would be ramping up as well, as they know you only have one other choice.
Open RAN presents an interesting insurance policy for operators, and Vodafone UK has even announced it will build a couple of thousand sites using less prominent vendors' equipment. This will get them better pricing going forwards from their existing vendors (as there's an onboarded alternative supplier), as well as give them a new option they might wind up preferring.
Right now though, with the proprietary nature of X2 (in the real world - the standards suggest it's not proprietary, but it absolutely is proprietary), you can't deploy interoperable radios, and that keeps your vendor choice limited. If operators can run a procurement and buy a box from one of five companies, that feels a lot better for their shareholders.
Amarisoft and Nexedi also launched a joint company called Rapid.Space last year. It combines cloud and 4G/5G infrastructure operation and is "HyperOpen" = using open-source software, open hardware and open service.
I work at a regional govt in Germany - we want to transition to local cloud - and think this is great product. If these 'NayuOS" can by synced to our own nextCloud installations. Could you help?
Thnx
> Amarisoft LTE and NR network software suit is a unique full software LTE and NR solution ... Our binary licenses can be fixed and bounded to a single hardware, floating on a USB dongle, or floating using a license server.
I suppose it's possible that their product includes some Open Source Software, while still using node-locked licenses. But it doesn't sound like it.
This press release [1] from 2020 makes it sound like Amarisoft is working on OpenRAN.
From reading HN it seems like there needs to be open source radio hardware as well to break the dependency on qualcomm, broadcom and whomever else is supplying chips. At least if the idea is to have something auditable.
From what I understand (someone correct me if I'm wrong) part of the problem is that baseband hardware and firmware requires expensive FCC approvals that's more compatible with the profitability of closed-source intellectual property.
It's not so much that it is more aligned with profitability per se, but the fact that each modification you put in the part of the device controlling the radio also needs said approval before being used to transmit - and FCC has been tightening the enforcement of that. It has been on HN before [0]...
Do we really need control over the radio hardware? Or can we treat them as dumb pipes, and use something like Tor to shield us from whoever can track who is using a radio channel?
I'm dealing with intel AX200 chipsets (deployed to hundreds of sites only to find out they do not follow intel's own specification for PCIe (it's a 1X board that fails on any socket >1X) and that chipset handles a lot. From AP/Frequency/Channel negotiation one would expect, all the way to WPA3 which one would think to be on the software side, specially because we should trust OS kernel/drivers updates for generics such as WPA to be updated and validated faster than closed source firmware blobs.
Now, I do not know if that is truly done on the firmware or the closed source drivers, because well, the drivers are closed source. But after you accept one binary blob, you might as well threat it all as integral part of the hardware. Who cares where it runs?
i'm getting off topic :) but that is one tiny aspect of a consumer device. One would think that the routing/DPI/etc aspects of 5G hardware for tel Co to be an amplified version of that.
Note: This is being pushed by the major telecom company in the EU
> The Continent's "big four" telcos Deutsche Telekom, Telefónica, Vodafone and Orange on Wednesday published a joint "memorandum of understanding" pledging to prioritize the development of "Open RAN" technology
They want to have more interoperabilities between equipment and more supplier basically :
> Open RAN encompasses the idea of chopping up the 5G supply chain into smaller pieces and imposing standards on equipment and software firms so their products can work together
This makes a lot of sense for telco since it would drastically lower the cost of their deployment and the maintenance of the 5G network.
Right now, if you take your equipments from one supplier, you are basically locked-in with this supplier since they are not compatible with each-other. So if one equipment fail, you need to buy a replacement from the same supplier.
>This makes a lot of sense for telco since it would drastically lower the cost of their deployment and the maintenance of the 5G network.
Actually no.
Integrated solutions are cheaper to make at scale (for how much it's sold it's a different question)
Also integration of multiple vendors into one working solutions has additional costs both for deployment and maintenance/support later (instead of 1 piece of equipment you have 5 boxes with 5 vendors pointing fingers at each other when something fails). Also more points of failure
It'll probably never happen, but we should look into solving the technical debt in our communications infrastructure – even if breaking backward-compatibility
.
Huge issues and vulnerabilities wrt security and privacy that are probably solvable, but were never considered in the first place.
I think there's a growing acceptance that SS7 is a protocol from the past. It's based on a set of assumptions (around telecoms operators being trustworthy) that just never held true.
The problem is that it continues to be the lowest common denominator for a lot of the world to stay connected from a telecoms perspective.
Moving to newer protocols makes a lot of sense, but for much of the developing world I imagine they'll continue to rely on legacy technologies like SS7 for some time to come, as they have that hardware in place etc.
The driver away from legacy equipment in the West will likely be the skills shortage as engineers retire and there's a need to move to newer equipment that can be maintained and understood. There's enough of a shortage of new talent, let alone new talent that understands the old telecoms world way of thinking.
> Moving to newer protocols makes a lot of sense, but for much of the developing world I imagine they'll continue to rely on legacy technologies like SS7 for some time to come, as they have that hardware in place etc.
SS7 is going to be here for a loooong time yet, just like email hasn't died.
> The driver away from legacy equipment in the West
Most of the cell towers that are install in the EU are Software defined. Its far cheaper and future proof to have them programmable.
Most other legacy stuff has been ripped out years ago, its far cheaper to maintain one fibre link than 15 ATM lines.
Older kit eats power, which means for remote cell cites, more fuel. This costs money, so its cheaper to rip out and put newer power efficient kit in.
> The problem is that it continues to be the lowest common denominator for a lot of the world to stay connected from a telecoms perspective.
Not just the developing world. The amount of money needed to rebuild our current telco grids with modern, designed from scratch infrastructure is in the hundreds of billions of euros range - in a time where money is short in supply.
Personally, I believe that even in 2030 there will still be widespread 2G networks as "last fallback" for generations-old IoT devices that simply can't do anything else...
In the US by 2030 both 2G and 3G will be a distant memory.
AT&T shut down its 2G GSM network at the end of 2016 and intends to shut down its 3G network early next year.
Verizon intended to shut down its 2G and 3G CDMA networks at the end of last year but delayed to the end of next year.
T-Mobile US is operating its 2G GSM network entirely within the guard bands of its 3G UMTS service (which means the bandwidth available is extremely limited) and while they haven't specified a set date for shutting down their 2G and 3G service they are no longer investing in expanding it (see the difference between their 3G[1] and 4G[2] coverage maps) and I don't see it lasting significantly longer than AT&T's or Verizon's 2G/3G networks.
Following the T-Mobile acquisition, the entire Sprint legacy network (2G, 3G, and 4G LTE) is going to be shut down by the end of this year with all customers being transferred to the T-Mobile network.
Dish Wireless's new cellular network (which is being built out as part of the legal settlement that permitted the T-Mobile/Sprint merger) will be entirely 5G from the start with no 2G, 3G, or 4G service offered.
Do you think that bandwidth heavy apps and sites should chip in? If we didn't use so much bandwidth we could easily stay with 3G? So it seems like because companies built these video sites and now everyone wants to have access to them in their pocket, why should telecom companies invest in it?
In addition this article is also strongly opinionated, charged with emotions, feels like stitched together without second reading IMHO. Like:
> It would allow operators to procure [...] with different players to piece together a 5G network, breaking the market power of “end-to-end” vendors like Ericsson and Nokia.
And later:
> the O-RAN Alliance. It's a standard-setting body that includes [...] leading vendors Ericsson and Nokia.
First, it doesn't work like that. They will still offer end-to-end deals, even with ORAN, it's just that winning conditions change. Second, why would companies support standardization effort that is supposedly intended to "harm" them? :D
> The operators, now barred by governments from using Huawei in several European markets, see Open RAN as a fix to what they consider a duopoly in the vendor market that allows Ericsson and Nokia to charge higher prices for 5G equipment.
I LOL'ed. From what I know, Ericsson and Nokia does not charge higher prices for 5G equipment due to duopoly, because that would be called price collusion and the journalists don't have proof to back it up. Also, Ericsson and Nokia are fighting each other for every piece of market share, I don't see how pumping up the prices would help here.
> Second, why would companies support standardization effort that is supposedly intended to "harm" them? :D
IMHO because decent regulations say so. AT&T has to comply with splitting up but did AT&T go extinct? Even though the situation isn't explicitly the same, the concept that government regulations are changing the market and it's direct effects are fairly similar. In the case above, Ericsson and Nokia are players that the government want to see exist in the market but for this specific aspect of the industry, they want to see a reduced role.
IMHO, it's government trying to work better on a very important topic (for the nation, for the people, and for the global market place [more competition for Huawei).
FYI, if anyone is wondering how the US Government spending fits in this puzzle, remember that when Motorola broke itself up, they sold their mobile infrastructure business to a Nokia/Siemens joint venture, which still has thousands of engineers working in the Chicago suburbs.
Most, if not all, of this money is going to be spent in the USA within the American tech sector, even if the company name is European.
I see, I was wondering why you said that, indeed the linked article[0] says million, this article typod. Yeah nearly a trillion dollars for any technology invested on in one swoop would be quite jaw dropping (this comment may not age well if inflation makes trillion the new billion).
Telephone networks are a dead business. You look at European telecom stocks and they just keep going down year after year. Wont be long before they're all bankrupts and will get nationalized again.
Yes, if there is no future growth in the market. Companies need to figure out how to expand their total addressable market or enter new lines of business.
I'm not sure this follows. There's plenty of business that are sustainable with flat growth. Especially utilities once they reach market saturation.
Just as a quick example, the supermarket down the street from me is neither expanding their market or entering a new line of business. There is no risk of a Walmart opening next door given the location. They just seem to be doing just fine selling the groceries people need each week.
Yes, sometimes market conditions change, and the companies need to react. Maybe people prefer buying avocados instead of carrots this season, and they need to adjust their stock. But outside that, there's no need for them to radically expand to stay sustainable.
imo this is the difference between investors think and consumer/worker mentality. the latter group is far bigger and totally fine with a sustainable bussiness w/o any growth but then again 'money talks'.
Yes, I would say you are correct. For small/local businesses a sustainable business which continues to produce a reasonable profit is perfectly acceptable. For publicly traded companies, the expectations are very different and analysts almost don’t care what you’re doing today, as long as the near-future will produce growth.
IMO, this would be a bad move. Let them die. Voice and text over the internet is here to stay. In the last two years I didn't a single phone call thanks to WhatsApp, Signal and Instagram.
Reliance Jio, a major player in India, is working on a OpenRAN based solution too [1].
Airtel, one of it's rivals is considering a similar approach , potentially joining in the efforts.
Governments do not understand the economics behind internet connectivity, they are just treating it as a way to make money, for example selling spectrum licenses to the highest bidder.
For those of you interested in a little of the history leading up to this point, particularly from a European perspective, you might find this of interest from me, written in 2019. A number of events lead up to our present dilemma.
https://blog.eutopian.io/huawei-5g/
to make code appear to no longer use unsafe functions, and make auditing harder. [2]
- When critical vulnerabilities were found (arising from poor code quality in user-facing protocols, and an old operating system), they were fixed, but the fix introduced another major issue into the product in question. [1]
- Reliance on out-of-support RTOS in products, and no alignment of own product lifespans to external support lifespans, and no identification of this issue by themselves. [1]
- Previously, 70 full copies of 4 different OpenSSL versions were found in products (and 304 partial copies), some of which dated back to 2006 with multiple disclosed vulnerabilities in each, showing no dependency management at all, and no management of vulnerabilities in dependencies. [2]
Does it matter if they did? At this point, everyone has to assume the CCP has the ability to take complete control of any Chinese company it wants to (see Ant Group). I’m not sure why they even bother with the “monopoly” or “corruption” pretenses at this point.
That's fine, but then you have to assume the same of the CIA and 101 other countries/agencies. So you have to design and build every part of all your systems nationally down to the copper wire pretty much and only after you've vetted all the people involved.
People need to stop with this US/China equivalence rubbish.
US is an open democracy albeit imperfect, independent judiciary, free press, free association, culture of openness and transparency. China has none of those.
That matters when you're asking which side to trust.
While it’s true that various US and foreign agencies can (and already are) tap into communication networks, that is a very different scenario than what you see in China, i.e. the state essentially taking control of private companies whenever they want.
We know the US used the NSA to gain an economic edge in negotiations with the EU. It even shared the information commercially didn't it?
I don't think the US would hesitate to interfere for a moment if it wanted. The EU is already under US sanctions for building a gas pipeline with Russia. There is the Iran situation and the wider Middle East and North Africa messes and Turkey. Plenty of places for a disagreement to form...
The EU has no major beef with China (I would prefer if they did, there is a lot to oppose about the PRC). China also has a much more isolationist/localist foreign policy so they're less likely to act and less likely to come into conflict with the EU.
1. EU is not being sanctioned for the Gazprom pipeline but rather EU companies which is a pretty important distinction.
2. EU does have major issues with China. They abhor it's substantial human rights abuses and geopolitical ambitions and have said on many occasions. Germany even has its navy warships on the way to the South China Sea for war gaming with APAC allies.
They are just willing to put all of that aside for better trading relations.
And the Germans joined to cleanse themselves of genocide and apply for readmission to the human race. While the French only joined because they wanted to protect their inefficient farmers from commercial competition.
The British joined to screw the French by splitting them off from the Germans that and of course their foreign policy objective for at least the last five hundred years which has being to create a disunited Europe.
I don't think everyone here is familiar with "Yes, Minister", so it would probably be best if you quote it.
For anyone else interested, the parent paraphrased a scene [0] from the legendary British sitcom "Yes, Minister" (~early 80s) where the civil servant Humphrey explains, in the most cynical manner possible, how the EU (at the time EEC) came to be.
You're most likely right, fortunately I don't possess quite the level of cynicism that my comment might have suggested. I do hope there was no one who believed my comment was made in anything but jest (at the expense of it's parent commenter), that would be disappointing.
I was present on a talk about the EU given by an UK official back in 2014. At that time the person didn't even knew how many members the EU had (Croatia was just admitted) and they were pretty open that they are not really interested to know. When I've been to the UK, two years earlier, a bus driver that I talked to was afraid that eastern europeans will come and take his job. I perfectly well know the UK relationship with the EU, but trust me when I tell you that giving some numbers does not reflect reality. Maybe because the UK considered itself a winner in WWII and wasn't ruined from it like France and Germany, or maybe because of Soros and black Wednesday, or for whatever reason,but Brittan was never a cornerstone of the EU the way France and Germany are, which is the thing the GP was trying to imply and I definitely disagree.
Alternatively:
Came late, when realised they were missing out, got special treatment,stayed for 30 years, made a mess, threatened to divorse, it's been 5 years and they still can't shut the door.
I'm under the impression his work is closer to the "radio protocol" than the infrastructure management behind. Not sure if they are related to Open-RAN though...
[1] https://www.amarisoft.com/about-us/