Are you using NTP or GPS for n78?
On FDD anchoring band, you must have same time reference.
So it won’t work without GPS on FDD if you are using on NR bands.
Thanks. I totally agree with you as it is what practically done because there will be many intra freq and adjacent freq enbs nearby. What I wanted to clear is, theoretically if there are two TDD cells from different bands nearby and there is relative time synchronization error (assume one is receiving gps input with errors , for sake of discussion) both cells will work fine without creating problem for each other.
Is it important for LTE anchor enb and NSA NR enb to be synchronized?
Nope. Thats beauty or want of sync. This will create a that cluster (LoS) goes erratic since UE also get distracted/confused. What you can do is for discussion, just remove the GPS clock in 1-2 sites and observe the -ve effects.
I think I am not properly putting my point.
Lets assume there are only two cells in my lab as neighbor, both are tdd cells from different band.
Do you still think if they have relative time error it will cause problem?
Yes.
Wait for some time frame and see UE will go out of sync from network.
Yes I think it will create problem, in TDD we would need sync.
In TDD networks, it is important that neighboring base stations are phase synchronized to avoid that downlink transmissions from one BTS is interfering the uplink slots of a neighbor BTS.
Ok but why will band 38 interfere with band 40?
Yes, its sync for any bands, eg. how UE knows to do CA if applicable.
How and when UE will know to start and end transmission of TDMA?
Once pss and sss is known, UE will know where the frame and slot boundary is.
How long it knows or stores?
It always knows.
If it cannot, it goes in unsynchronized state and will again start to read pss and sss.
Who feeds sync to pss/sss?
Oscillator in enb, and it needs freq synchronization.
I am talking about time synchronization.
Yes, so keep GPS only for startup and deinstall… what happens then?
It’ s FDD paired spectrum, it’s freq sync. TDD needs time sync.
That is same as in fdd where you don’t need cells to be time synchronized.
If you use gps only for freq sync and supply no time , fdd cell will be up forever.
True…you have tdd also in the same cell.
GPS gives both tdd n fdd. time sync accurate than freq sync.
3G has freq sync coming from RNC, common sync is fed.
Yep I got it, but my question is not about having sync or not.
It is about relative time error between cells.
There was no each site had any gps or sync input.
It takes from the RNC interface.
If its lab…do all negative test you will realise in test cases with error and without error.
May be am missing your exact query.
Look at step by step answer.
It is still not clear.
See:
And:
enb1 = c1+c2 each 20 MHz.
So it’s networkwide config havg TDD cells.
enb2 also have c1+c2.
Either inter or intra CA based on the EARFCN.
enb2 can be immediate neighbour or distantly placed.