Hello Experts.
Imagine you have available 5G bands:
- 3.5 TDD
- 2.6 TDD
- 2.1 FDD
- 700 FDD
And you need to lauch SA.
What would be your preferred strategy for bands and why?
Mine :
3.5 / 2.6 / 2.1 –> NSA + SA
700 → SA
3 Likes
Why did you propose such strategy?
What are the criteria?
If you have 5G Core then why we are proposing NSA?
Thanks guys, but you are missing something…
Device and Smartphones are not supporting from day 1 the Standalone, new security algorithm and SIM card replacement has to be done so you have to keep NSA in your network 
I am asking for Traffic steering and Coverage capacity wise, how would you do that in your way?
2 Likes
I don’t have any issue with SA/NSA hybrid network.
My question is why did you choose 700 MHz as SA only?
700 has no traffic at NSA.
It needs an anchor and x2 interface brings additional delay in NSA.
And goal to use n28 is for long range which is a thing in SA.
If your current network is NSA only and 700 MHz doesn’t have traffic it’s weird.
Generally it should take highest traffic among all layers.
Any specific reason for that?
We don’t have much trafffic in 5G, mainly in cities so 3500 and 2100 is grabbing all traffic.
It’s strange how service providers are jumping into the SA architecture without waiting to get their return of investment on NSA first.
Do you think your low 5g traffic is because of NSA architecture and that’s why the mobile operator is deploying SA?
No, because some phone types are not certified by Vendor, and people don’t enable 5G mode in their phones.
With SA idle mode, there will be more 5G traffic to be grabbed as it will offload 4G data.
I have worked in 5G/4G networks, and the majority of traffic is coming from 4G.
The support for 5G is not fully exploited by users, either the phones don’t support it, turned off for power saving, the packages for 5G are a bit expensive, or simply covergae holes in the 5G network.
I don’t think operators will completely rely on SA 5G any time sooner.
Absolutely, SA can be used for private networks for now not for covering the whole country.
But we are on the track.
Here in Iran also only a few handset are capable of connecting to 5G NSA due to limitations imposed by device manufacturers leading to very low traffic.
If you ask my opionin, 5G NSA service quality and ROI isn’t worth the deployment headaches.
However my opinion may change in the future as we rollout 5G in entire country.
Well, not sure which BW you have but TDD over 60 MHz and FDD 20 MHz even bring such difference.
NSA 2 CC is supported in DL and market is already crawling with the phones supporting CA n1_n78.
On top of that you have LTE
We are deploying 5G NSA in a multi-vendor scenario where anchoring LTE and overlaid 5G come from different vendors.
Many deployment issues due to incompatibility.
It’s n78 with 80 Mhz BW.
I see it’s quite challenging.
I’d never prefer that to be honest.
Even latency might be an issue.
Because we have single RAN BTS meaning 4G and 5G communicating via backplane for cosite, so latency already optimized.
5G SA total throughput < NSA 5G throughput.
Do you have any requirements for latency?
Is your use case sensitive to latency to have strict requirements?
Not for now.
Sensitive application will come wt SA but even in NSA if you have big x2 latency some features will not efficiently work like intersite CA for instance, also impacts NSA total throughput.
For Exploiting all equipment of old technology
Actually it happens since when the user is on 5G-SA the control plane uses resources (PRBs) that while on 5G NSA are used by user plane (PDSCH).
1 Like