Introduction To The TR-398 Wi-Fi Performance Test

[sfx2000]

The first Wi-Fi performance test standard will change the way we test Wi-Fi products.

Continue reading on SmallNetBuilder.

Looking at the state of CableCo provided CPE (home gateways) - the efforts by Broadband Forum have paid off.

The carrier-provided Customer Premises Equipment (CPE) is a heck of a lot better than it used to be...

WiFi has kept up with the increase in BW provided by DOCSIS 3.1 - most of the newer equipment is WIFi5 Wave2, if not WiFi6 already, and it's interesting to note that at least with CoxHSI and Spectrum down here in San Diego, they have all deprecated 802.11b, running the 2.4GHz radios as default with G/N modes at a minimum, and the WiFi6 as G/N/AX - defaulting to 20MHz channels, and landing on the high/mid/low defaults as 11/6/1.

70 percent of the WIFi here in my neighborhood, based on my surveys is carrier-provided equipment, not off the shelf consumer brands. Interesting, and probably should put fear into the hearts of the OEM's that are not part of the operator ecosystem.

 

[coxhaus]

There is no doubt the ISPs have a chance to displace a lot of consumer gear with these new higher speeds.

 

[thiggins]

With better gear and higher prices approaching those of premium cellphones, the carriers are finding less resistence to their monthly gateway rental fees. Now all they need to do is introduce free upgrade for life programs and retail router sales will really be under the gun.

 

[RMerlin]

With better gear and higher prices approaching those of premium cellphones, the carriers are finding less resistence to their monthly gateway rental fees. Now all they need to do is introduce free upgrade for life programs and retail router sales will really be under the gun.

That's an interesting idea. Imagine if ISPs, tired of their customers getting compromised and flooding their network with malicious traffic, decided to pay manufacturers for special long-term firmware support. Many manufacturers drop support early not due to technical reasons, but merely for business reasons. That might give them an incentive to keep devices supported for a longer term - provided someone pays for it.

Kinda like business customers can still get Windows 7 updates from Microsoft, provided they pay for it.

 

[sfx2000]

That's an interesting idea. Imagine if ISPs, tired of their customers getting compromised and flooding their network with malicious traffic, decided to pay manufacturers for special long-term firmware support. Many manufacturers drop support early not due to technical reasons, but merely for business reasons. That might give them an incentive to keep devices supported for a longer term - provided someone pays for it.

I think some of the operators have caught on that with their provided equipment, a compromise to the customer router/AP is a compromise to their network - at least with a certain class of exploits that is...

I suppose one of the knock-on side-effects of this is reduced functionality - and this isn't just cable, but I would say DSL, Fiber, Fixed Wireless, etc...

Anyways, we're getting off track a bit - my key reason for necro'ing this thread was that the performance of the recent crop of carrier-provided gear has improved significantly since the release of TR-398, and as such, the customer experience with that equipment.

 

[sfx2000]

That's an interesting idea. Imagine if ISPs, tired of their customers getting compromised and flooding their network with malicious traffic, decided to pay manufacturers for special long-term firmware support.

Wanted to add here - it's not that the carriers need to pay extra, it's that if the OEM wants their business, they will have to extend the support lifecycle.

With managed devices, thru both DOCSIS and TR069/CWMP, it is fairly efficient these days to roll updates out into the fleet, and they routinely do this on a regular basis. That is another observation, and one confirmed in sidebar discussions with my contacts in the industry.

 

[Tech9]

70 percent of the WIFi here in my neighborhood, based on my surveys is carrier-provided equipment

Over 70% in my area. I'm sick of Plume pods - the best Wi-Fi a man can get.

Now all they need to do is introduce free upgrade for life programs and retail router sales will really be under the gun.

What is stopping them is the need to install the equipment. Most pay contractors to do it. If the customer stays quiet, they won't move a finger. I have replaced the ISP router 3 times by dropping the old one at the store and installing the new one by myself. Very few people will do it this way.

 

[sfx2000]

I'm sick of Plume pods - the best Wi-Fi a man can get.

Hmmm... from limited first hand experience at a friends house with "Panoramic WIFI" provided by CoxHSI - it wasn't that bad actually - and that was a 4500 sq ft split level home and we were out on the back patio (you should see the back patio, it's awesome)...

With any WiFi access point - location is everything I suppose, but we were working together with several others on a global MS teams call with video etc, and it was fine..

 

[sfx2000]

I think some of the operators have caught on that with their provided equipment, a compromise to the customer router/AP is a compromise to their network - at least with a certain class of exploits that is...

This likelly is a North American view - internationally, we see posts from forum members which clearly show that security is only as good as the operator makes it out to be - I would still be concerned about having Huawei or ZTE equipment on my network...

but they are very common out there, no doubt.

 

[Tech9]

With any WiFi access point - location is everything I suppose,

The pods locations are where the power plugs are. There is no way to optimize anything. If YouTube doesn’t work, they just plug more.

 

[Elio The Italian]

Hi @sfx2000
Thank you for the post.
I believe the TR-398 test suits are the best performance test suit available in the market. I acknowledge that this test suite can be performed using CANDELA equipment. However, I have recently seen an alternative, OCTOBOX, in some articles.
Can anyone name the TR 398 alternative that is performed using OCTOBOX?