PDA

View Full Version : Question Why UE/MS on 3G run out of battery sooner then MS on 2G?



ejamhas
2011-06-29, 02:48 PM
Are there any logical explanations to the reason mobile devices on 3G are believed to run out of battery sooner than mobile devices on 2G?


Cheers :)

chamanbagga
2011-06-29, 03:29 PM
In 2G, we use GMSK and the chipsets consume less power. The demand on 3G is much higher.

Besides, we have a bigger issue to solve on smartphones. The network timeouts for transition from DCH > FACH > Idle vary from 3-5 seconds right upto 1 minute. As a result, if you are doing frequent email syncs, the phone may be consuming more power.

Sometimes, the 3G coverage is not good as 2G coverage. Normally operators keep the phones in 3G even under low signal strength, which implies that phones consume more battery for transmission.

And ofcourse, if you have 3G enabled, then you will have reselections between 3G and 2G causing additional messaging.

Also see
[url=http://blog.famzah.net/2010/05/24/2g-gprs-vs-3g-umts-connection-battery-usage-on-mobile-phones/]2G GPRS vs. 3G UMTS connection battery usage on mobile phones

fon909
2011-06-29, 03:46 PM
Thanks chamanbagga for covering all the major points, one of the main problems is that of coverage, 3G being an Underlay network coverage foot print is small, plus higher frequency used in 3G (2100 MHz) makes it prone to higher attenuation as a result the signal strength is poor, so UE struggles to keep it self camped on to 3G network as an outcome the battery power is consumed more than while camped on to a 2G network.

Hope this is helpful.

indoor
2011-06-29, 04:05 PM
I was under the impression it was due to the extra processing power required by a 3G receiver compared to a 2G receiver. In idle mode the handset uses more power on 3G decoding the signalling channels than it would listening to the BCCH on 2G.

plannerguy
2011-06-29, 06:15 PM
Hi


some main reasons....

1)In UMTS the Power Contro is 1500 times a minute while in gsm it is 2 times (480ms)
2) Processing capacity reuired is more so power higher.
3) Power amplifier used in gsm is class-C has linear response,
in UMTS it is logarthamic .

Regards
Plannerguy

harrypotter
2011-06-29, 09:05 PM
On the operator side, putting MHA will prolong the UE battery...

sesunguts
2011-06-29, 09:15 PM
Go for 3G Distributed type RF Unit. This will improve uplink noise and reduce UE uplink power to transmit. This also will improve coverage due to no feeder loss.

harrypotter
2011-06-30, 05:11 AM
Thanks chamanbagga for covering all the major points, one of the main problems is that of coverage, 3G being an Underlay network coverage foot print is small, plus higher frequency used in 3G (2100 MHz) makes it prone to higher attenuation as a result the signal strength is poor, so UE struggles to keep it self camped on to 3G network as an outcome the battery power is consumed more than while camped on to a 2G network.

Hope this is helpful.


Is it the same effect with 850 3G?

plannerguy
2011-06-30, 09:23 PM
Hi

one thing more in DSM it is GMSK and QPSK so same happens in phase so have constant AMplitude .

IN UMTS same is having QAM , so there will be ubrupt varitions in Amplitude same cause the amplifier reponse to be non linear.


Regards
Plannerguy

s52d
2011-07-01, 12:09 AM
Are there any logical explanations to the reason mobile devices on 3G are believed to run out of battery sooner than mobile devices on 2G?


Cheers :)

One point: in GSM, receiver needs to listen only 1/8 of time (cca), as well as tramsmitter.
In R99, they work all the times: 8 times more!

We tested some 3g terminals in idle mode: with no traffic, they keep battery 5 days or so.
It is active mode where battery is drained fast, so there is not much we can do in idle mode.

There are few lines of defense:
- common LA to lower signalling
- cell-pch or ura-pch to make it a bit faster
- DTX for modern terminals: HSPA+ shall be better as plan R99

I somehow do not bielive we can get a lot with power generation. Transmitter efficiency
drops with low power, so few dB less transmit power do not add significantly to power
consumption. I might be wrong here, has anyone tested it?

BR
s52d