Radio engineers: Why doesn't the AX88U Pro's antenna array outperform the AX86U Pro's?

[sfx2000]

This is probably just BestBuy branding, they had this marketing practice before. BestBuy exclusive like RT-AC2900 or RT-AC86U, then RT-AC1750 or RT-AC66U B1, etc. They also had one exclusive RT-AC1900P or RT-AC68U B2 variant. Just adding more to the Asus model line numbering confusion.

gah - it's the big-box SKU's and price-matching...

BestBuy and Costco have a lot of "private" model numbers unique to them to avoid price matching...

Last time I was able to successfully price-match was at Fry's Electronics, with one of the models on the floor, and what was on Dell's website, and there, they credited me the difference - which is amazing if one remembers Fry's...

 

[jasz]

Hate to break it to you - on all modern SoC's - the ethernet ports and wireless interfaces don't touch the CPU - it's all in the switch subsystem.

CPU comes into play with the bridge interface, along with the control plane (let's configure the switch) otherwise it's an application processor that does some of the upper layer processing...

Sorry if you think I'm calling you out - if you believe so, that's on you...

I'm just here to correct the record - I've been doing this stuff for over 30 years now...

It's fine.. I just want you to explain why a why 2x2 device will bottleneck throughput unless I reset my ONT via SC/ACP, Clearing stats or running an ONT>OLT speed test.

No one seems to have a clue. 4x4 connection can hold higher throughput, even at lower QAM rates. Not a wireless interference issue. More packet related.

All methods can do full throughput when triggered.. Theres just a bottleneck unless I reset ONT or clear connection. RJ45 is intermittent for full throughput to local ISP server.

On this subject, the BE800 is a faster router, especially pushing throughput to AX200 @ full QAM 160mhz.. My ASUS routers ARE worse.

 

[sfx2000]

t's fine.. I just want you to explain why a why 2x2 device will bottleneck throughput unless I reset my ONT via SC/ACP, Clearing stats or running an ONT>OLT speed test.

No one seems to have a clue. 4x4 connection can hold higher throughput, even at lower QAM rates. Not a wireless interference issue

All methods can do full throughput when triggered.. Theres just a bottleneck unless I reset ONT or clear connection. Wir

Fiber and DSL are managed by TR069, so there's a fair amount of device discovery for capability after the ONT/CPE - so if you switch routers, you would be good to power cycle the modem/ONT as well...

There are reasons why I don't do fiber or DSL - and that is the TR069 management spec...

 

[jasz]

Fiber and DSL are managed by TR069, so there's a fair amount of device discovery for capability after the ONT/CPE - so if you switch routers, you would be good to power cycle the modem/ONT as well...

There are reasons why I don't do fiber or DSL - and that is the TR069 management spec...

Nothing seems to matter... I've power cycled around 6 different routers here.. The QCA 1220 platform is the strongest for throughput, when it doesn't even make sense.

Just going to assuming its a routing conflict and I need more RAM and or RISC V on ONT end. Issue happens on wired, just less prone to packet drop?

This is aside from my off sample AX86U outperforming the same and other newer ASUS routers. That was tested in conjunction with a DOCSIS setup. I can't really explain why that one performs to a higher extent, but the BE800 out performs everything here regardless.

 

[sfx2000]

No one seems to have a clue. 4x4 connection can hold higher throughput, even at lower QAM rates. Not a wireless interference issue

Actually a 4*4:4 radio is always that - even when doing 2*2:2 to a client - so there, it's the coding gain at the baseband level... and that coding gain, along with the RF paths, it's reciprocal, e.g. it's uplink and downlink...

A 2*2:2 compared to a 3*3:3 AP - the 3-stream AP will have a 33 percent improvement, a 4-stream AP isn't much more - but it is more because of MIMO diversity...

FWIW - the clients have 2-steam capabilities and the routers doing 4-stream - it's actually the best practical combo with 11ac and later...

 

[jasz]

Actually a 4*4:4 radio is always that - even when doing 2*2:2 to a client - so there, it's the coding gain at the baseband level... and that coding gain, along with the RF paths, it's reciprocal, e.g. it's uplink and downlink...

A 2*2:2 compared to a 3*3:3 AP - the 3-stream AP will have a 33 percent improvement, a 4-stream AP isn't much more - but it is more because of MIMO diversity...

FWIW - the clients have 2-steam capabilities and the routers doing 4-stream - it's actually the best practical combo with 11ac and later...

It's definitely packet related given a straight RJ45 connection is more consistent, but susceptible to said issue.

I just laugh and cry myself to sleep every time I reset a GUI stats setting and instantly get full throughput to servers innately bottlenecking at 1/4th the speed prior. Even on 2x2 WIFI at 256 QAM.

Make it make sense.

Edit: May end up back on DOCSIS with high split deployment. Zero issues there.

 

[sfx2000]

Nothing seems to matter... I've power cycled around 6 different routers here.. The QCA 1220 platform is the strongest for throughput, when it doesn't even make sense.

Which boards to you have - QCA-1220 hasn't arrived in the market yet, but there are EVK's...

You can share the IPQ model, or even better yet, the 65- level which is the actual reference part numbers...

 

[sfx2000]

I just laugh and cry myself to sleep every time I reset a GUI stats setting and instantly get full throughput to servers innately bottlenecking at 1/4th the speed prior. Even on 2x2 WIFI at 256 QAM.

Make it make sense.

Makes no sense - esp your comment on QAM256 - this only works in the lab for 2.4GHz...

 

[jasz]

Which boards to you have - QCA-1220 hasn't arrived in the market yet, but there are EVK's...

You can share the IPQ model, or even better yet, the 65- level which is the actual reference part numbers...

It's a consumer BE800. It's been out for a year....

IPQ9574

 

[jasz]

Makes no sense - esp your comment on QAM256 - this only works in the lab for 2.4GHz...

What do you mean.. im using a 5/6ghz client and dipping into sub 1024 rates.

 

[sfx2000]

It's a consumer BE800. It's been out for a year....

IPQ9574

So a TP-Link device - just note that's basically nominal 11be, as a launch device, it's wave1 - which is ok, but know there are bugs in that device that cannot be fixed...

 

[sfx2000]

What do you mean.. im using a 5/6ghz client and dipping into sub 1024 rates.

2 stream client...

 

[jasz]

2 stream client...

And that matters how? 4x4 vs 2x2 wont change QAM modulation

 

[sfx2000]

And that matters how? 4x4 vs 2x2 wont change QAM modulation

I don't have time to teach you about MIMO...

QAM and MIMO do not have a direct relationship as they're at different layers of the stack...

 

[jasz]

I don't have time to teach you about MIMO...

QAM and MIMO do not have a direct relationship as they're at different layers of the stack...

???? What are you even saying lol

You're making a simple subject overcomplicated for no reason.

If I have a 2x2 device locking into a lower QAM tier, it doesn't matter if its 2x2 or 4x4....

 

[sfx2000]

You're making a simple subject overcomplicated for no reason.

No, it's actually pretty simple - SS is encoded before the modulation at the baseband...

Baseband works with analog RF to determine which modulation is actually used on the Tx side..

Anyways - like I said, if you want me to teach you about MIMO, I'm ok with this, just send me money first, lol...

 

[sfx2000]

If I have a 2x2 device locking into a lower QAM tier, it doesn't matter if its 2x2 or 4x4....

Actually it does - as the AP can use 2 radios for each stream and add 3 dB on the link budget.

No need to get into a pissing contest - you have one perspective, but it's not at a developer level for new devices...

 

[jasz]

I'm glad you want to show off, but I'm pretty certain you knew what I was referring to when it comes to TX lock rates at distance/SNR

Maybe I should have said MCS index instead....happy?

Either way.. We're way off topic.. I just wanted to know why this ONT or back end service impacts WL throughput to the extent it does.

Seems we wont be able to figure it out unless they build a new unit or something gets corrected at CO.

 

[sfx2000]

I'm glad you want to show off, but I'm pretty certain you knew what I was referring to when it comes to TX lock rates at distance/SNR

Maybe I should have said MCS index instead....happy?

MCS in 11n was a good measure of link performance, as each rate had it's own MCS...

With AC and later - MCS gets kind of more aligned, perhaps, in that it takes the number of spatial streams out of that particular metric.

Recall that the link performance itself is generally based not on the AP, but the client itself...

With the Asus devices, there are ways to look at connection stats per client in the WebUI, and if I recall on most, there is also the command line params.