Now you can check tower location of 5G SA on cellmappet limited to KL Putrajaya Cyberjaya. Hopefully by end 2025 they rollout to other states


Attached thumbnail(s)

It's nationwide I believe, not just limited to KV area only (there is SA towers mapped in Kelantan)

Though mapping it using CellMapper requires bandlocking your phone to NR/5G only... unless you have the specified U Mobile plan with SA access.

This post has been edited by luqman98x: Feb 20 2025, 08:55 AM

Luckily I moved from Umobile last month😅

From the screenshot it's purely N78 only but intraband-CA. Meaning bonding of 2 channels from same band range.

N78 can be very weak because it is operating at 3.5GHz. It cannot cover wide areas unless base stations are extensive.

Looks like there is n28 if you go on Cellmapper.

Based on rigorous and through tests conducted in India for some time now for over a year now with Jio, they are using the same model there with solutions from Nokia and Ericsson. Huawei is not allowed in India market due to a reason you know why





Test was shown at the end of the video standing just right below the base station just metres away.

As you can see, India 5G band utilization is very SIMILAR to that implemented in Malaysia.They are also using NR-5G SA: N28 (10MHz) + N78 (100MHz)

With 5G-SA, you no longer need to be connected with B28 to anchor band with N78. Both can operate individually as they are called "Stand-Alones". Studies have shown that (read the reviews on both videos), combining both N28+N78 yield little improvements when doing CA.

In fact, NR5G N78+N28 combination is not recommended and included in 3GPP standard papers. When you bond both the bands, the smaller bandwidth N28 will tend to drag N78 performance down with it causing performance degradation/slower speeds. With 5G-SA, it's better to have one single band with wide channel or intraband CA than to to interband CA of multiple bands.

This is more towards Nokia and Ericsson which caters for South Asia markets.

Not sure what Huawei will bring to the table to differentiate themselves with the competition. If they want to innovate and stand out, they need to bring something different that improvises over Bharat competition.

India is one operator using 100Mhz+20Mhz, Malaysia currently is 6 operators sharing 100Mhz+20Mhz …

So our speed if all fully using technically will be 5-6x slower than rest of the world …

I also expect n28 to perform not so great as it's being used with DSS with LTE B28.

Spectrum ≠ Speed. Expanding spectrum is one of the way to increase capacity but that is not the only solution. Cell densification can be done.


Source: https://www.ciena.com/insights/articles/Den...-at-a-Time.html

Eventhough 5G usage already reach more than 50% in Malaysia, the median mobile data speed is still increasing and already reach top 11 in global ranking. This is the ranking with limited 5G spectrum that MNO can't even do with competitive 4G networks in decades.


This post has been edited by OfficiallyAhmad: Feb 23 2025, 06:06 AM

Cell densification is equipment vendor wet dream since telco needs to build more sites … that is why vendor like Ericsson hardsell single wholesales to Malaysia government knowing with limited spectrum shared by so many telcos have no choice but to pay Ericsson to build more to get to same level as 100Mhz per operator as in other countries …

Hope UM exit with another telco then can offload abit … but UM has existing LTE bands can enable more CA or even refarm their LTE to 5G since they using Huawei/ZTE, great benefits to Umobile.

The rest of the poor operators like Maxis and CDB will have to suck up to DNB and Ericsson single wholesales network scam … i bet Opensignal and Okla never measure the time our connections stuck in 5G and cant browse which have to disable 5G to resume connection experience …

U Mobile 5G SA
Not really stable though, suspect UMB core network registration got some problem
Some IBS like UTC Pudu have SA Signal but can't register , says timeout
U98 (I sign up just to test SA , need to use min 90 days)
QCI 8 / VoNR

SA still got hanky panky.

Wow. There's even VoNR?

Maxis 5G really worsen up the scenario no point Govt keep self syok about own 5G network it’s time to boost network
Sadly large numbers of users Of big Maxis CD need to stuck and live with it

Gov only want coverage aka the 5G icon, so they can boast about coverage. This is 2G era mobile measurement. However as data user, we all know have 4G icon doesnt mean can good data service/experience - capacity differentiate operator to operator.

My personal experience a lot of time with 5G icon but data wont load, all these those Ookla/Opensignal report wont show. If cant even load then cant measure. Hassle for user to disable 5G. If like this might as well just use 4G.

This inconsistency may cause people just off 5G …

Forgot to mention DNB actually enable MFBI at their 5G Portal , showcasing their Private Network
N77 and N78

[url=https://pictr.com/image/xVzncj]
[/url]

Yes , surprised cuz even U Mobile never mention it.
But seems like U Mobile disable every services when attach to N28 only. Cant register to data services nor voice services. Will fallback to LTE or N78 if in DNB N78 coverage.

Frequently got disconnected on N78 and will get either radio_failure or NR Registration Timeout, like even ur status bar have 5G and connected to SA but the Internet is not working. Have escalated to UMB for check.

Alright bro, if we just go by your own bad experience, what about all the other people having good 5G experiences? Isn't that just as biased? That's why we look at median data from Ookla and Opensignal, it's a much wider view, it takes everyone experience into account. Yeah, the reports aren't perfect but it's way fairer than just basing everything on your personal opinion alone, right?

Ultimately, both the aggregate data and personal experiences should be used in tandem to give a more accurate understanding of the 5G landscape. every user's experience matters, and we shouldn’t just brush it aside for the sake of broader reports that might not always reflect reality on the ground.

If only measure slow and fast samples then median makes sense but whether “nothing can be loaded” is also part of the samples maybe you let us know. If i cant load Speedtest, how they know my problem and capture as part of their data.

Is like rating food tastiness, if food not even serve, how to give my rating? I order 10 meal, 5 not serve, 4 delicious, 1 suck. So, 80% tasty?