LTE-U vs. Wi-Fi: What You Need To Know

[sfx2000]

Told folks this was going to happen... LAA or LTE over unlicensed spectrum - WiFi is clearly impacted...

More cell phone data use is negatively affecting Wi-Fi performance, study finds​

Academic Paper and Findings here (warning, PDF link)

Hidden-nodes in coexisting LAA & Wi-Fi: a measurement study of real deployments

 

[thiggins]

Thanks for the links @sfx2000. I suppose we should be thankful that LAA/LTE-U deployment isn't widespread in the U.S. The carriers are too busy with 5G boosterism.

 

[sfx2000]

Thanks for the links @sfx2000. I suppose we should be thankful that LAA/LTE-U deployment isn't widespread in the U.S. The carriers are too busy with 5G boosterism.

ULA/LAA/LTE-U is being deployed in the US, much more than folks know about.

The key problem is that an LAA/LTE-U device may have awareness of WiFi and function properly, other WiFi stations/AP's cannot see the LTE waveforms, hence are unaware - so this impacts everything from the channel estimation (needed for MCS rate determination) to Tx/Rx ops.

It's going to really impact adjacent ESS/BSS's - and this was also observed in WiMax 802.16e/16m studies for unlicensed band operation.

WiFi cannot compete against a scheduled MAC like LTE or WiMax, it will lose every time... not even the improvements made in 11ax/WiFi6

There is a place for the LTe telco operators - here in the US, we have the 3.5GHz CBRS band, where capacity is available and has all the benefits of running in the UNII bands, and none of the downsides for incumbent technologies that are already in the band.

What is CBRS?

Citizens Broadband Radio Service (CBRS) refers to 150 MHz of spectrum in the 3550 MHz to 3700 MHz range (3.5 GHz to 3.7 GHz) that the United States Federal Communications Commission (FCC) has desig | Citizens Broadband Radio Service (CBRS) refers to 150 MHz of spectrum in the 3550 MHz to 3700...

www.fiercewireless.com

Citizens Broadband Radio Service - Wikipedia

en.wikipedia.org

 

[IK Gamex]

Is this only an issue of fairness, as in LTE-U/LAA clients operating on 5 GHz taking priority over WiFi clients, or is there a net efficiency loss as well?

For example, if the LAA traffic for a streaming video on a smartphone is instead offloaded to a WiFi network (like Xfinity Mobile hotspot), wouldn't the overall spectrum impact be the same?
Or are there inefficiencies that come with the LTE-U/LAA traffic vs WiFi traffic, such as the 10ms transmit for LAA vs 4ms transmit for WiFi mentioned in the article?

 

[sfx2000]

For example, if the LAA traffic for a streaming video on a smartphone is instead offloaded to a WiFi network (like Xfinity Mobile hotspot), wouldn't the overall spectrum impact be the same?

If it is offloaded to WiFi, that's ok - the challenge here is that LTE-U/LAA/ULA is LTE used in the channel, and there, wifi doesn't have a prayer of getting any airtime.

Anyways - carriers should stick to the spectrum that they own and charge subscribers rent via their monthly bill at the end of the day.

 

[sfx2000]

Or are there inefficiencies that come with the LTE-U/LAA traffic vs WiFi traffic, such as the 10ms transmit for LAA vs 4ms transmit for WiFi mentioned in the article?

LTE-Advanced is actually more efficient than 802.11ac by a factor of 2 - 11ax is a bit closer, but only with 160MHz channels and 8 streams with OFDMA enabled - since most 11ax is 2 or 4 streams on 80 MHz channels w/o OFMDA in place, efficiency there is similar to 11ac

Again, the carriers really have no reason to use UNII bands when they could more efficiently use the spectrum they already own - devices that support ULA also generally support carrier aggregation in LTE, so make better use of the resources they own.

I hate to sound like that old man - "Get off my lawn!" as I rack the bolt back

 

[sfx2000]

What's worse, a neighbor running a 160 MHz bandwidth Wi-Fi network, eating up every available 5 GHz channel (DFS excluded), or LTE-U sharing one 20 MHz channel?

That neighbor running 160MHz channels - WiFi6 AP's work on the control channel, and use CTS procedures to check if the channel is clear or not before using the secondary channels

(this is one of the reasons why 160MHz channels in 5GHz is always a challenge, in 6GHz/WiFi6e is less so)

LTE-U/LAA/Multiire wil dominate a 20MHz channel, even as an LTE Carrier Aggregate candidate, because it is a scheduled MAC... and it has no legacy requirements that 802.11 has...

That 20MHz LTE-u will burn a hole in the spectrum for LTE...

In any event, it's starting to look like a non-issue here in the US - as more spectrum has been offered/auctioned off in the C-Band, and there's always band 48, which is CBRS, and carriers are starting to notice the opportunity there for private LTE/5G-NR networks there...

 

[coxhaus]

Why don't they use 6 GHz instead of 5 GHz? There is plenty of bandwidth.

 

[sfx2000]

Why don't they use 6 GHz instead of 5 GHz? There is plenty of bandwidth.

Nice to see you @coxhaus - it's been a while.

seriously though, there really isn't any reason for deploying LTE/5G-NR into unlicensed space in ISM/UNII bands - CBRS/band48 is enough, and with the recent band expansions into C-Band and the 600MHz digital dividend...