LTE - TM3, TM4 and TM8 openloop or closedloop?

[subway]

Dear Experts,

We are doing dual layer BF (TM8) on a 3.5GHz Huawei based network, this is a mostly static network, no mobility, 8T8R, CRS=2.

What is not quite clear is if I set the MIMO mode to openloop or closedloop, will that only affect the the TM3/TM4 switching, or will it affect the PMI/RI feedback for TM8 as well? The documentation is not quite clear about that (or about how to set the PMI/RI feedback mode for BF modes like TM8).

I can see that after the BF TM8 is enabled, there are a lot of clients with TM8 and Rank=1, while previously most of them was on TM3 Rank=2. And I have a feeling that the reason is that TM8 is working in single antenna mode, as there is no PMI/RI feedback.

Any and all help will be appreciated!

[justdream]

Setting Openloop/ Closedloop will impact CQI, PMI, RI
which will impact Throughput

[subway]

Thanks, but this is not really an answer to my question.

Although, I managed to find the answer on my own, here it is:

In a TDD system, normally you don't really need closed loop feedback, as due to channel reciprocity the enodeb can get this info out of sounding reference signals (SRS). With more advanced transport modes like TM9 and above, the mechanism can benefit quite a lot from PMI indication, therefor there is a precoding matrix indication (PMI) feedback-dependent TM9, and PMI feedback independent TM9. Compared with TM7 and TM8, TM9 uses a different weighting mechanism and different reference signals for channel state information (CSI) measurement.

For TM7, TM8, and TM9 PMI feedback independent, the SRS is enough, and the MIMO mode can be configured as open loop.

42241

[khurrambilal01]

Thanks, but this is not really an answer to my question.

Although, I managed to find the answer on my own, here it is:

In a TDD system, normally you don't really need closed loop feedback, as due to channel reciprocity the enodeb can get this info out of sounding reference signals (SRS). With more advanced transport modes like TM9 and above, the mechanism can benefit quite a lot from PMI indication, therefor there is a precoding matrix indication (PMI) feedback-dependent TM9, and PMI feedback independent TM9. Compared with TM7 and TM8, TM9 uses a different weighting mechanism and different reference signals for channel state information (CSI) measurement.

For TM7, TM8, and TM9 PMI feedback independent, the SRS is enough, and the MIMO mode can be configured as open loop.

42241

What if we still activate Close loop MIMO? Will that consume more UL resources considering TDD is already UL limited capacity system.

[justdream]

Dear Subway,
Could you kindly share the final Beamforming setting you are using?

[subway]

What if we still activate Close loop MIMO? Will that consume more UL resources considering TDD is already UL limited capacity system.

Yes, that is correct in general, but there is more to this topic:

What I have seen is that for 4x4 MIMO, and TM9 4 layer the PMI/CQI/RI feedback might not be mandatory, but due to the more detailed feedback, they provide significant gains compared to its open loop counter part. (please note that 4x4 MIMO and TM9 are mutually exclusive).

For example Huawei claims, that 4x4 MIMO (compared to 2x2 MIMO) is 20 to 60% faster in open loop, and 50 to 90% faster in closed loop mode (both in terms of average and cell edge users).

Of course you need to factor in, that closed loop (lest call it feedback dependent) mode works best in near stationary or stationary installations. For high speed UEs the feedback information gets obsolete by the time of scheduling and arrival, so in those systems open loop is a better approach.