On 6/7/24 00:46, Amin Dayekh wrote:
I agree with you and i never implied that the satellite can carry (as it is now) the capacity of the fiber. Future might unfold new technologies to the contrary.
I'm not sure how. Satellite uses RF technology for transmission. RF is prone to interference from weather and other electromagnetic sources. And compared to fibre, satellite spectrum in the Ka and Ku bands is very limited.
In the 90’s the speeds were in kbps and now i am preparing for Gigabit connectivity for residences.
I understand Starlink are going to be selling 1Gbps for US$75,000/month, with a setup cost of US$1.25 million. Not exactly in the reach for most people, unless you are using that to create a community ISP. Of course, how consistently one will be able to achieve 1Gbps or higher on satellite (never mind 100Mbps) as more customers sign up is yet to be seen.
Remember that Starlink are offering a faster alternative to other satellite providers and services. Their model is not to compete with fibre, because that is a fool's errand.
Think of Starlink as more of a neigbourhood DOCSIS or DSL access network... faster than dial-up (other traditional LEO and GEO satellite services), but slower and less consistent than fibre.
On the other-hand, hypothetically, if the 1:1 solutions are provided to the end market, there is no need for a huge and very costly infrastructure under the sea.
With current 5nm and 3nm DSP technology, we are able to carry between 15T and 25T per fibre pair undersea, over a distance of 7,000km - 11,000km. There is no evidence to suggest satellite will ever get there, especially considering that CMOS process node production is looking to get to 0.7nm by 2034.
I have seen a great decline in demand for fiber and broadband connectivity from local Telcos who have hugely invested over decades for resilient networks in many areas where the satellite services deployed, some ISPs are about to go bankrupt due to low client base and high operational costs, and of-course stability.
You are describing last mile fibre into homes and businesses. That is very different from building submarine fibre networks.
In the Market where I am operating, many large scale organization, universities, to sme’s and residences have disconnected from isp’s and connected to Low Orbit Satellites, despite the Pros and Cons and any logical or Technical explanations that we might Provide, key point in the Market here is Price and Stability. A fiber cut in Lagos will impact my Network in Abuja or Portharcort!
That is not a failing of optical technology. That is just a failing of the local market to advance the technical and commercial benefits of fibre.
Again, you are comparing apples to oranges. Submarine cables carry national traffic; often times the traffic of multiple nations at the same time. It is an aggregation-based model, which justifies the huge upfront and ongoing maintenance costs. The businesses and homes you describe that switch from fibre to satellite are doing so on an individual basis. The two cannot be compared from a scalability and cost-per-bit-per-person standpoint.
Moreover, how do you think the satellite traffic is carried to its destination once it gets to the earth station?
In countries where Public services are degraded such as Power Supply, roads, and etc… operational costs (despite economical hardships, Forex policies and tax) would impose a major increase on final costs of the services.
This is true - but as useful as the Internet is, it's hard to call it a public service. Nevertheless, all manner of enterprise suffer from improperly managed national infrastructure and services. This is not unique to Internet access over fibre.
In the Northern Part of Nigeria for instance 1 STM is priced at 1,290,000Naira per Month, 1USD=1490 today. Satellite is priced at 38000 Naira and Provides speeds up to 180MbPS. Many sme’s are who are neighbors are becoming wisps, they buy a single sat and subscribe to a residential package then share it among themselves.
This is completely acceptable, but like I said, it cannot scale as more customers join the service because of the limited amount of spectrum on satellite, the susceptibility Ka and Ku bands have to weather, and electromagnetic interference. So your 180Mbps at 5AM can easily become 128Kbps at 5:15AM, and you are still paying NGN38,000/month.
The 65000 clients reported by the same entity to be their clients (if i recall the figure well) have been lost from Various Wisps, ISPs and Telcos. Despite the fact that the service is unstable on weather conditions,
Instability due to weather is not a small matter. If you are a business seeking 24/7 uptime, that will matter to you.
the low Latency plays a major role in this competition.
Right - and pound-for-pound, fibre offers lower latency than satellite, which matters at scale, because that low latency can be guaranteed through all manner of weather and electromagnetic conditions.
National Long-hauls are costly and overly priced,
This is not a failing of the optical technology.
Latency to CNN can be 110 to 90. More people will disconnect from the conventional ISP and connect to the Satellite, for many reasons, call it fantasy of new Products , affordability etc. more ISPs will shutdown, the capacity utilized by end users will drop and eventually, Telcos will get to R&D for new solutions other than 5G that proved failure on stability here, and with time and further development of Low Orbit technologies, building more ground-stations etc, the Submarine cables will become a secondary infrastructure to connect with. I might be wrong.
You are wrong.
Telco's will shutdown for other reasons than optical technology "not being fit for purpose". Our own hubris, over the years, will be our downfall - not fibre. But I digress.
Anecdotally, there will be some traffic shifting from terrestrial infrastructure to satellite. But in general, the majority of traffic will be carried over fibre, and over subsea systems. Hanging your business plan on the death of fibre and submarine networks would be what I'd wish on you if you were my competitor.
If things continue this way, we won’t be far from a centralized internet over few key major players globally who can afford space low orbit Satellites.
The Internet has already been partitioned in many ways, between large states like China, and large operators like the content folk. And this is all happening on fibre.
But, partitioning is not quite centralization.
I tend to think of it (sometimes) as a game of thrones, publicly and government funded projects vs privately held entities.
The market will change, there is no doubt. But as it pertains to transport medium, there is no evidence to suggest that satellite is going to take over fibre as a general rule.
Mark.