LTE interoperability

[wco81]

Arstechnica has a post that Verizon is going to make sure their LTE phones can't be used on other networks.

Besides little spectrum overlap between its network and AT&T's planned LTE network, VZ devices have to fall back to CDMA for 2G and 3G while AT&T LTE devices would presumably have to support UMTS if no LTE network is available.

What is the state of LTE chipsets which are already out and those coming in the next couple of years?

You would think they support multiple bands and all planned LTE spectrum around the world?

Qualcom must surely be planning support for CDMA as well as UMTS in their LTE basebands?

Apple would probably be interested in such a product, so that they don't have to have different designs for VZ from just about every other network in the world.

I thought VZ went with LTE so that they wouldn't have limited device support. But now they have market power and they're going to get support from most manufacturers. They can demand from most manufacturers that phones must only support their LTE and 3G networks, though they do sell some "world" phones to some of its customers.

 

[Arun]

It has little to do with the baseband or RF chips and everything to do with the power amplifier modules. All traditional implementations require one chip per 3G band (sometimes multiple chips are hidden in one module but that's an implementation detail). Just look at the number of HSPA phones which aren't 5-band despite that requiring only a few more dollars versus a 3-band phone. Now consider what happens when you'd need nearly 10 bands to support all LTE operators launching within 12 months, and at least 15 within 3 years. And that adds up to your 5 HSPA bands and a few more CDMA bands.

The other thing is that LTE roaming won't happen for quite some time, and even if it did it'd still be so expensive that you'd best remain on 3G. So Verizon 'world phones' will include the same HSPA bands as today's 'world phones' but the LTE bands will still be far from worldwide.

The only way around this is wide-band PAs which are extremely inefficient in a traditional implementation but can be made even more efficient than traditional ones with technology like Nujira's or (to a noticeably lesser extent) Infineon's polar radios (see: UE2 RF). But that's not ready today and won't be mainstream for quite some time.

 

[wco81]

Wow, 10 to 15 bands?

I heard that in Europe, the operators did not want interoperability for GSM but was required to support it because of law.

I can't imagine the EU would not continue this requirement for LTE, because they're trying to keep the Union together and new technology isolating the member states would run counter to this goal.

Certainly the manufactures and carriers must know that a lot of people require international roaming. Not to mention the advantage of using more common SKUs and components to drive down manufacturing and inventory costs by having the equivalent to today's quadband devices.

 

[rpg.314]

Certainly the manufactures and carriers must know that a lot of people require international roaming. Not to mention the advantage of using more common SKUs and components to drive down manufacturing and inventory costs by having the equivalent to today's quadband devices.

Well, the carriers are almost universally hated for their leech like behavior by their own customers. There's gotta be a reason for that.

 

[Arun]

Wow, 10 to 15 bands?

Yeah, according to Informa there are 8 core bands: 700 FDD (USA), 800 FDD (Europe), 900 FDD (Europe?), 1800 FDD (Europe?), 2100 FDD (Japan & Europe?), 2600 FDD (Europe & USA via Clearwire?), 2300 TDD (India & Misc), and 2600 TDD (China).

However I'd argue you need to consider most 700MHz power amplifier implementations do require different PAs for AT&T and Verizon frequencies, and you should at least add 1700 FDD (AWS) to the core band list, so that's really 10 power amplifiers. And there are plenty of secondary bands that could probably get you to 20 but hopefully the TDD ones at least will be mostly reserved for the receive side (ala what AT&T wants to do with the Mediaflo spectrum) which doesn't require a Power Amplifier.

I heard that in Europe, the operators did not want interoperability for GSM but was required to support it because of law.

I can't imagine the EU would not continue this requirement for LTE, because they're trying to keep the Union together and new technology isolating the member states would run counter to this goal.

I'm not aware of any such law for 3G (although maybe it's not legal to release a phone without the 2100MHz frequency?) but the way that same goal will be achieved for LTE is simply by coordinating the spectrum sales. The EU will mostly use the 800MHz/2600MHz bands with some 1800MHz and 2100MHz as well.

Certainly the manufactures and carriers must know that a lot of people require international roaming.

International roaming on LTE devices works just fine with 3G. If you're paying $LOL/MB anyway, it's not even very useful to have anything better than EDGE.

 

[wco81]

I think there would be interest in voice roaming but for data, there would be people who would want to buy local SIMs to use on unlocked devices.

Or use services like iphonetrip.com or tepwireless.com.

BTW, the way things have gone, this forum should really be called Mobile Technology. Unless the 3DS and Vita really reverse a lot of trends, dedicated handheld gaming devices will be a niche market.

 

[Arun]

I think there would be interest in voice roaming

Voice roaming on LTE? Hehe, couldn't help but laugh, sorry! There will be voice support on LTE eventually (sooner rather than later now) but all operators today are simply reusing their 3G network, and voice really wasn't a priority for the LTE standards body. And quality will not be any different (in theory - in practice it's likely operators will use wideband by default whereas very few operators do so on 3G today).

but for data, there would be people who would want to buy local SIMs to use on unlocked devices.

Agreed, but the bands are so different that I'm skeptical that will happen without a Nujira-like solution. And it is also attractive on a cost and power basis even if you didn't care about the extra bands. I talked with someone from Nujira

BTW, the way things have gone, this forum should really be called Mobile Technology. Unless the 3DS and Vita really reverse a lot of trends, dedicated handheld gaming devices will be a niche market.

Hmm? I think 'handheld' is nearly as generic a word as 'mobile' - I'm not sure how it is limited to handheld consoles. It includes phones and tablets but excludes notebooks and arguably netbooks.

 

[wco81]

What is the base station and antenna design compared to 3G?

Can existing 3G base stations be upgraded easily to LTE?

In the US a lot of communities will block efforts to add more base stations or even change the antenna design, even though a lot of the residents are mobile customers.

So the ability to ramp up a new network as well as keeping legacy networks running indefinitely is in question. Verizon has turned on coverage in major cities but those are probably the low-hanging fruit -- areas with population density and high traffic.

 

[McHuj]

What is the base station and antenna design compared to 3G?

Can existing 3G base stations be upgraded easily to LTE?

You need a new base station for LTE. The physical layer processing and latency requirements posed by LTE are too demanding for an easy upgrade.

The startup I work for is working on an LTE base station, but a picocell. A picocell will offer much smaller coverage areas for fewer users than a typical macrocell base station, but the idea is that you hang picocells anywhere there's access to internet and a powerline (side of buildings, telephone polls, etc). This way a network can be built out without the need to build antennas.

The market is still probably 1-2 years away from someone having an actual productized, picocell. There have been plenty of demo's however.