New wifi router from Ignition labs

[stevech]

DFS spectrum under FCC is not of great interest to me - transmitter power control (TPC) is unattractive for a fast-changing channel as we have with handheld devices' WiFi.
DFS is more applicable to the many (proprietary) TDMA point to point systems in 5.4GHz than to CSMA as in 802.11. Such as from Redline Communications.

Mr. Easy need not give credence to any of this from the Dvorak wannabes.

 

[sfx2000]

What is your solution? Most consumers are not able to access DFS channels.

Many will if you set them to auto, and they'll scan in to the DFS space as part of the ongoing OBSS searches (OBSS is a required part of 11n and later)...

Rightfully so however, those same devices, when not in auto, do not allow DFS channels to be selected - stop and think about it - manually selecting a DFS channel can't be done, as it's dynamic, right?

TPC is also kind of nice when done right, as it includes hinting information to the clients as far as link margin, and that can aid with band selection on a single/multiple AP setup...

 

[stevech]

FCC rules I recall said TPC is required. But doing TPC in CSMA is a paradox.

 

[sfx2000]

FCC rules I recall said TPC is required. But doing TPC in CSMA is a paradox.

they operate at different levels of the stack - no problems there...

 

[thiggins]

Many will if you set them to auto

What product(s) are you using that are operating on DFS channels?

 

[sfx2000]

What product(s) are you using that are operating on DFS channels?

Airports at the house, Aruba at the office (the Aruba's are centrally managed dual band 11n)

 

[thiggins]

Airports at the house, Aruba at the office (the Aruba's are centrally managed dual band 11n)

Thanks

 

[Razor512]

I wonder, does anyone have any performance info on how a router reacts when it changes from 1 DFS channel to another?

While performance can be lower if it has to frequently scan for radar and other traffic on the channel, but when it actually decided to switch, is the switch over fast enough to not be noticeable,or will it be like when you manually change the WiFi channel on a router and it has a long noticeable delay.

 

[stevech]

Airports at the house, Aruba at the office (the Aruba's are centrally managed dual band 11n)

Wouldn't DFS/TPC capability be apparent if the product claims 802.11h capability?
In the US, there's Cisco Meraki and as said, Aruba. I may have seen a claim of '11h from a recent Netgear product.
In the EU, ETSI is more adamant about DFS/TPC in the lower ISM bands than is the FCC.

Some older (and current?) managed WiFi systems had/have problems with channel # flailing as algorithms incorrectly tried to deal with traffic load balancing among channels due to intra- and inter-system interference.

 

[sfx2000]

I can confirm that Cisco does, and I believe the old Trapeze gear did as well...

Most of the WLC's these days have anti-flailing mechanisms in place, so a fair amount of management capability is there...

 

[stevech]

I wonder, does anyone have any performance info on how a router reacts when it changes from 1 DFS channel to another?

The practical DFS systems are managed WiFi networks, not consumer routers. These systems have a way to direct the clients to change channels. Same as is done in cellular systems as the base stations direct mobiles when to change base stations/channels.

But as said earlier, many ISM band products that use DFS/TPC are not WiFi; they are TDMA systems, often point to point or point to multipoint private networks with dish antennas. From prior projects I've done, an example is this company whose products are quite band-agile:
http://rdlcom.com/products
and
http://rdlcom.com/products/enterprise-ptp

 

[sfx2000]

Thanks

 

[thiggins]

I've moved the cloud pro/con discussion here. Please continue the discussion there. Please keep posts here related to the Ignition Labs Portal product.

 

[sfx2000]

Thanks - I was going to ask for the thread to be moved, as there was good discussion, but it went off the main topic..

 

[thiggins]

This is IDL's response to the question regarding TPC:

Portal is capable of operating with TPC (Transmit Power Control) but like many consumer routers we have opted to certify our first version of Portal without TPC. TPC is generally required only for outdoor products with very high output power. In the USA, the limit for when TPC is required is an EIRP (Effective Isotropic Radiated Power) greater than 27dBm or 500mW. In the EU, the TPC limit is an EIRP above 23dBm. As point of reference, the typical commercial consumer routers on store shelves today are calibrated between 20dBm and 23dBm, so well below the limit.

Our first version of Portal is going to be certified to operate below the 27dBm TPC limit in USA and Canada because (a) we intended it as an indoor product only; (b) it’s a consumer product so we want it to be compatible with as many consumer client devices as possible (TPC is not universally supported outside of enterprise environments); and (c) we want to keep the cost down… testing, calibrating and certifying TPC is quite expensive. Our second version of Portal, which we are designing for EU, will follow the same principle but with the similar (but different) EU limits. And if and when we decide that we want to build an enterprise version of Portal we can implement TPC quite easily with the same design.

Having said that, even with lower EIRP level restricted to indoor limits, we designed Portal to have great range and coverage (enough to cover a typical size US home) by using four MIMO streams (the latest state-of-the art), solid 360 degree antenna design, and of course our unique spectrum boosting and active traffic avoidance technology.

As background, here’s the excerpt from FCC Section 15.407 for TPC
"U-NII devices operating in the 5.25-5.35 GHz band and the 5.47-5.725 GHz band shall employ a TPC mechanism. The U-NII device is required to have the capability to operate at least 6 dB
below the mean EIRP value of 30 dBm. A TPC mechanism is not required for systems with an e.i.r.p of less than 500 mW.”
And here’s the excerpt from ETSI Section 4.4.2.2

 

[stevech]

Are they saying that the FCC requires TPC in 2.4 and 5.8GHz ISM (rather than 5.4GHz)?
It is/used to be that EIRP limits were a function of antenna directionality (and thus gain). Higher EIRP as antenna gain increases and beamwidth decreases.
The principle was that you can interfere less as beamwidth lessens.
(I never understood these rules because they didn't stipulate horizontal vs. vertical beamwidth, i.e., it's easy to get 12dBi gain in an antenna that is omni-directional in the traditional sense. The contractor to FCC that wrote that had a mindset of dishes and yagis and not stacked-dipole verticals - easily done in 2.4/5GHz.)

These regulations differ greatly around the globe.