Aggregating both TDD & FDD

[IamMrFamous07]

I'm not sure if anyone posted this article or not but apparently SK telecom and Nokia achieved this in the summer.

 

I know sprint has an abundance of TD LTE spectrum thanks to clear wire and they plan to aggregate those channels

 

Do you think sprint will try to aggregate all their frequencies in the future?

 

http://networks.nokia.com/news-events/press-room/press-releases/lte-throughput-leader-nokia-sets-world-record-with-sk-telecom-of-close-to-4-gbps-using-tdd

[dedub]

I suspect there are too many differences in propagation characteristics between the bands to make much difference.

 

besides the fact, even if it were possible, it would be wasting double or triple the spectrum per user (ie now a user would be using both b26 and b25, instead of only b26 or only b25).

[jroepcke51]

I'm not sure if anyone posted this article or not but apparently SK telecom and Nokia achieved this in the summer.

 

I know sprint has an abundance of TD LTE spectrum thanks to clear wire and they plan to aggregate those channels

 

Do you think sprint will try to aggregate all their frequencies in the future?

 

http://networks.nokia.com/news-events/press-room/press-releases/lte-throughput-leader-nokia-sets-world-record-with-sk-telecom-of-close-to-4-gbps-using-tdd

I can't remember where Robert's post was but due to the depth of Sprints B41 spectrum, there really is no reason for Sprint to aggregate FDD and TDD.  In addition, B41 is suppose to be the capacity and speed which will be aggregated to 40mhz and then to 60mhz.  Finally, there are several markets that have multiple B41 carriers active (chicago??).

[Dkoellerwx]

Do you think sprint will try to aggregate all their frequencies in the future?

No.

 

They may aggregate B25 carriers, or B25 and B26 *maybe*. But there's no need to aggregate B25 and B41, even if there was a way to do it. Which with the current setup, there is not.

[kyle_4thousand]

No.

 

They may aggregate B25 carriers, or B25 and B26 *maybe*. But there's no need to aggregate B25 and B41, even if there was a way to do it. Which with the current setup, there is not.

How would aggregating B25 & 26 work? I would imagine the propagation differences between the bands would make it difficult where B25 doesn't reach.

 

I know Chicago has 2 B25 carriers, would aggregating these two 5x5 carriers make sense? From a capacity standpoint it doesn't make much of a difference but would the improvements in peak speed and (maybe?) better edge of cell performance be worth the trouble?

 

 

Sent from my iPhone 6 on the Now Network

[Dkoellerwx]

How would aggregating B25 & 26 work? I would imagine the propagation differences between the bands would make it difficult where B25 doesn't reach.

 

I know Chicago has 2 B25 carriers, would aggregating these two 5x5 carriers make sense? From a capacity standpoint it doesn't make much of a difference but would the improvements in peak speed and (maybe?) better edge of cell performance be worth the trouble?

 

 

Sent from my iPhone 6 on the Now Network

For B25 and B26, aggregated connections would only be available where you have both signals. You would still be able to connect to only B26 or only B25 if needed. However, aggregating these two carriers isn't likely. Capacity needs to be reserved for those beyond the reach of B25.

 

As far as aggregating two B25 carriers, the peak speed and performance boosts may be worth it. Especially since a lot of people now base their perception of network performance solely on speed.

 

Sent from my Nexus 5

[IamMrFamous07]

I wonder if they will aggregate 600mhz and 800mhz....hmm the possibilities

[Dkoellerwx]

I wonder if they will aggregate 600mhz and 800mhz....hmm the possibilities

If 600 ever even exists.

 

Sent from my Nexus 5

[IamMrFamous07]

If 600 ever even exists.

 

Sent from my Nexus 5

Right it's been a mess so far. Even after the "auction" I wonder how long will it take for deployment and being able to use it

[gnoj]

The framework for implementing FDD + TDD CA won't be finalized until 3GPP Rel 12 is frozen this March, so it's still a ways off from real-world deployment. For now, Sprint only has plans for CA in B41. Inter-band aggregation will be much more important for an operator like AT&T, whose spectrum holdings are much more varied and non-contiguous. In most markets, they'll need to aggregate AWS + PCS + WCS to keep up with the peak download speeds Sprint will have with a single 60 MHz B41 chain.

 

While low-band spectrum can be aggregated with the mid- or high-band spectrum, I also believe that operators should leave those bands separate to help ensure usable throughput is maintained for people at the edge of coverage. Aggregation with another low-band (i.e. B5 + 12, or B26 + 600 MHz) would be alright though.

 

As far as aggregating two B25 carriers, the peak speed and performance boosts may be worth it. Especially since a lot of people now base their perception of network performance solely on speed.

 

I agree that this would be a good idea, but can Sprint's equipment be software-upgraded to support CA within B25?

[lilotimz]

 

I agree that this would be a good idea, but can Sprint's equipment be software-upgraded to support CA within B25?

 

Band 25 carrier aggregation is dead and Sprint equipment does not and will not support it. 

[gnoj]

Band 25 carrier aggregation is dead and Sprint equipment does not and will not support it. 

 

It's not entirely "dead" in that 3GPP will support it in Rel 12. It's a question of whether Sprint's existing NV equipment can be upgraded to support it. If it requires a hardware upgrade, then I agree it would be all but dead to Sprint, as they would (justifiably) feel that it would be a waste of time and money to deploy it.

[lilotimz]

It's not entirely "dead" in that 3GPP will support it in Rel 12. It's a question of whether Sprint's existing NV equipment can be upgraded to support it. If it requires a hardware upgrade, then I agree it would be all but dead to Sprint, as they would (justifiably) feel that it would be a waste of time and money to deploy it.

 

No it's dead. According to neal, Sprint does not have the "channelization" to even deploy interband carrier aggregation. 

[gnoj]

No it's dead. According to neal, Sprint does not have the "channelization" to even deploy interband carrier aggregation. 

 

What does he mean?

[WiWavelength]

How would aggregating B25 & 26 work? I would imagine the propagation differences between the bands would make it difficult where B25 doesn't reach.

For B25 and B26, aggregated connections would only be available where you have both signals. You would still be able to connect to only B26 or only B25 if needed. However, aggregating these two carriers isn't likely. Capacity needs to be reserved for those beyond the reach of B25.

 

For carrier aggregation, I typically just refer to the carrier and supplemental carrier, but the 3GPP terms/acronyms are primary component carrier (PCC) and secondary component carrier (SCC).

 

Now, it is important that the PCC have propagation characteristics that are better than or equal to those of the SCC.  Otherwise, the PCC fails before the SCC.  In that case, the entire link fails, and the network must be reacquired.

 

 

http://www.3gpp.org/technologies/keywords-acronyms/101-carrier-aggregation-explained

 

So, apply this to Sprint.  The band 26 carrier would have to be the PCC, while the band 25 carrier would have to be the SCC.  But Sprint does not want that configuration.  Band 26 is not to be primary; it is to be secondary -- used for coverage, not capacity.

 

See the problem?

 

AJ

[gnoj]

Now, it is important that the PCC have propagation characteristics that are better than or equal to the SCC.  Otherwise, the PCC fails before the SCC.  In that case, the entire link fails, and the network must be reacquired.

 

The key then is how long it takes for a network to be "reacquired." If we're talking a couple seconds (like the transition we see in tri-band devices switching between B25/26 FDD and B41 TDD), then that's not much of an impediment. CA could simply be disabled during VoLTE calls when smooth band switching is of paramount importance. If it's any longer than that, then any sort of inter-band CA (even AWS + PCS) could be a problem. 

 

In any case, I agree that low-band (<1 GHz) should not be aggregated with mid- or high-band spectrum. This also presumably wouldn't be a factor to consider with B25 intra-band aggregation.