4 ms·
You do have to wonder at what point Apple decides they don't need Verizon, T-Mobile et. al. Maybe it's still a long ways off, but if you've gone through the eff
by poulsbohemian 4y ago
You do have to wonder at what point Apple decides they don't need Verizon, T-Mobile et. al. Maybe it's still a long ways off, but if you've gone through the effort of building relationships, hardware, software, etc to weave together a satellite network, why not just keep going down that path?
- xattt 4y agoMy guess for Apple’s long game is is to turn the Find My network into a distributed network for comms beyond location data. One of the incentives to participate may be free or reduced-cost satellite data.
- martinald 4y agoYou're vastly underestimating the bandwidth required for cellular service. A starlink sat "cell" can do ~20gbit/sec (though I actually think real world performance will be massively lower). This covers hundreds of kilometres. A single cell tower with 5G/4G is not far off doing that, per sector. And these cell towers can cover as little as 500m or less in dense urban areas. The problem is there are very real physics you come up against with this. We are really tapping out efficiency gains these days so the only option is more, higher frequency spectrum and much denser cell networks.
- jsjohnst 4y ago> A single cell tower with 5G/4G is not far off doing that, per sector. Assuming you mean LTE when you say 4G, the per sector bandwidth is customarily an order of magnitude less than 20gbps. Even in the case of 5G, with sub-6Ghz anyway, you rarely will see above 15gbps. There’s only so many bits per second you can squeeze per megahertz and below 6Ghz the allocations there just isn’t enough available frequency. That’s why cell towers have gotten much closer together.
- martinald 4y agoWhere did I mention only sub 6GHz? With mmWave towers are comfortably pushing around 20gig. 4G (LTE) and 5G have similar bits/Hz loadings anyway. And towers are getting closer together. You reduce transmit/receive power and place the towers closer, which results in similar coverage but less interference between cells (with reduced power output).
- jsjohnst 4y ago> Where did I mention only sub 6GHz? When you mentioned 4G in the same statement. The highest I’ve ever seen for a LTE tower in the US (I’ve been in 49 states and checked via field test equipment in multiple areas of each state, yes, I’m that nerd) was under 5gbps. LTE sectors more commonly are in the 1-2gbps range. As to 5G, sub 6Ghz it’s possible to get close to 20gbps air rate, but you need an very wide channel to do so (have to do the math, but I think >100mhz of spectrum) and I’m not aware of any US carriers doing so (but they could be). If we are talking the rare unicorn mmWave, the one with less than 1% of all US cell towers supporting, sure, 20gbps is common. Twice that isn’t unheard of either. Long story short, we are violently agreeing both before and now. It’s just nuances we are both dinging the other for. Cheers!
- poulsbohemian 4y ago>You're vastly underestimating the bandwidth required for cellular service. No, I’m really not, that’s why I put the caveat about maybe not today, but big rivers have a way of starting as a small trickle.
- fomine3 4y agoEven if radio tech improved so much, big fewer cellular towers can provide better service cheaper than satellites.
- zamadatix 4y agoIf anything I see increased convergence between connectivity methods not any one technology supplanting another. Making your service plan able to connect the user from Wi-Fi + Celluar + Satellite when they are indoors, in a dense area, or far away will be able to deliver a far greater experience than locking into any single technology both in coverage and performance. We already see this kind of thing already with T-Mobile + Starlink, Home internet providers bundling phone plans and allowing you to roam to other networks, traditional cellphone providers hopping on board Wi-Fi calling support, and Apple + Globalstar.