FUD Alert: VoLTE Calls Are Battery Killers??

Posted · Add Comment

 Now the Blogosphere is incessantly echoing this story … I first noticed in GigaOM, their story is really Fear Uncertainty and Doubt (FUD) extremist journalism (how’s that?) Their story is linked below.

Voice Calls Over 4G LTE Networks Are Battery Killers

The key quote is:

The results of those tests should give carriers and consumers pause. The average power consumption for a 10-minute CDMA circuit-switched call was 680 milliwatts (mW), while the average consumption for a VoLTE call of the same duration was 1,358 mW. That’s double the power drain. Spirent estimated that on a full charge, its test smartphone could support 502.6 minutes of talk time using CDMA only, but the same charge would deliver just 251.8 minutes of talk time using VoIP on the 4G network.

So as a quick refresher, VoLTE (Voice over LTE) calls are similar to the 3G/2G counterparts with VoCODING and so on, but are carried like VoIP, over LTE packets back to the network, where instead of a circuit switch there is an IMS for calls etc etc…

Some related info from before…

Mobile Wireless HD Voice Today and VoLTE in the Future

 Firstly measuring battery consumption is a brute force thing to do here but it does expose the maturity of VoLTE clients at the moment, so I will give them that. The FUD part of it is that the client on the device they tested is really not like anything on a 2G/3G voice phone today. Most of the grunt work of a 2G/3G call takes place in hardware on the phone, with the phone’s UI being the only significant software piece. Voice calling has been optimized in hardware over the years to consume very little power as there is no need for memory, CPU etc… 

On the other hand, the current crop of VoLTE clients are running the whole VoLTE protocol ‘stack’ (SIP + more) in software plus running the VoCODING at this point in software. The bottom line is today’s VoLTE clients are CPU hogs that do use CPU, memory, UI, etc…but are very UN-OPTIMIZED. Let’s give the developers a year or two to optimize the solution and come back and measure. In the end, there will be no difference in power consumption on average.

Let’s stop getting hysterical!