Debunk the range myth

[digital10]

I have been researching this and I can not find an answer. A lot of reviews will tell you this router doesn't give enough range, and another router has superior range. After lots of reading and asking, it seems that all Wifi antennas are limited by spec. power(1watt I believe), meaning they should all have equal range. One would think cheap Chinese routers would have less power and brand name(Asus, Netgear) will have the high power/range. This does not seem to be the case.

Any one can explain? I tested some routers and its true, some won't reach as far as the others.

 

[L&LD]

Better hardware (processing power, RF design/implementation, and attention to detail, such as CNC machining of antennae vs. stamped steel cutouts), better QC, and better matching of the specific model/router to the WiFi environment and clients being tested.

Not to mention the quality of the tests being done (was the optimum control channel chosen for each router?, etc.) and the ever-changing WiFi environment that is impossible to be kept at a 'control' level too. Even for testing done 5 minutes apart.

 

[Trip]

Adding to what @L&LD said, there's a lot more that goes into effective range than just antenna dBi and/or transmit power.

First, you've got the wireless standard and spatial stream capability. It's pretty well known that a 4x4 AC chip is going to have better receive gain than a 2x2, which means better range on that aspect alone (all other items held equal). In many instances as well, there's a lot of pre-draft AX gear out there that is generating better throughput over distance than the equivalent AC hardware (as @L&LD will attest to with his AX88U versus the previous AC86U...).

Then you look at the chip brand/model series, as each has a varying degree of pedigree when it comes to effective range for AP/station hardware: Qualcomm is generally considered tops, with Broadcom often in second, MediaTek close or following and others below that.

Another layer would be the amplifier chip being paired with the radio. You can have a great QCA radio, for example, but if the amplifier is faulty or the combo of the two is not a great marriage for whatever reason, then the outcome could actually be worse than an AP with, say, a cheap MediaTek chip with a higher quality amplifier.

Then you look at the implementation in-software: drivers, code base, etc. This is a huge point in the SNB community, where you've got tons of overgrown erector set all-in-one routers with supposedly superior wifi specs that can't hold proper 2.4 or 5Ghz reliability to save their lives. Compare that to seemingly modest, lower-power business-class APs that smoke the pants off them stability-wise, because their distros are being vetted by enterprise QA/QC.

Then you get into the physical design of the antenna array itself and its ability (or inability) to decolorate properly, maintain receive sensitivity, mitigate interference, and/or dynamically change polarity. The best radios (aka Ruckus, high-end Aironet, etc.) can do all of the former, and do it well. The cheap sheetmetal origami and/or screw-on external variety found on 90+% of everything else, consumer hardware included, often cannot.

And that's just a brief overview of how varied the landscape can be. So it's a lot more complex that what get's labeled as "more range!" by a given manufacturer, journalist, etc.

P.S. It's why I, and many others, choose to run enterprise wifi at home. Less questionable everything stated above, so the network runs more like an appliance and less like a toy.

 

[digital10]

Adding to what @L&LD said, there's a lot more that goes into effective range than just antenna dBi and/or transmit power.

First, you've got the wireless standard and spatial stream capability. It's pretty well known that a 4x4 AC chip is going to have better receive gain than a 2x2, which means better range on that aspect alone (all other items held equal). In many instances as well, there's a lot of pre-draft AX gear out there that is generating better throughput over distance than the equivalent AC hardware (as @L&LD will attest to with his AX88U versus the previous AC86U...).

Then you look at the chip brand/model series, as each has a varying degree of pedigree when it comes to effective range for AP/station hardware: Qualcomm is generally considered tops, with Broadcom often in second, MediaTek close or following and others below that.

Another layer would be the amplifier chip being paired with the radio. You can have a great QCA radio, for example, but if the amplifier is faulty or the combo of the two is not a great marriage for whatever reason, then the outcome could actually be worse than an AP with, say, a cheap MediaTek chip with a higher quality amplifier.

Then you look at the implementation in-software: drivers, code base, etc. This is a huge point in the SNB community, where you've got tons of overgrown erector set all-in-one routers with supposedly superior wifi specs that can't hold proper 2.4 or 5Ghz reliability to save their lives. Compare that to seemingly modest, lower-power business-class APs that smoke the pants off them stability-wise, because their distros are being vetted by enterprise QA/QC.

Then you get into the physical design of the antenna array itself and its ability (or inability) to decolorate properly, maintain receive sensitivity, mitigate interference, and/or dynamically change polarity. The best radios (aka Ruckus, high-end Aironet, etc.) can do all of the former, and do it well. The cheap sheetmetal origami and/or screw-on external variety found on 90+% of everything else, consumer hardware included, often cannot.

P.S. It's why I, and many others, choose to run enterprise wifi at home. Less questionable everything stated above, so the network runs more like an appliance and less like a toy.

So you are saying there is no way to know the range of a router unless you actually buy it and test it for yourself?

Does business class AP's give wider range or more stable system? Does it feel any different browsing on your regular consumer router(ex ASUS) or a Ruckus AP?

I thought of installing business class AP's but at $700 per AP, no wireless mesh, and I heard you must have an expensive yearly subscription its just not worth it.I am not sure if the price difference=the added benefit difference. The closest thing I got to was wired Ubiquiti, not sure if overhyped consumer APs or on par with Ruckus and Aruba.

 

[coxhaus]

A consumer wireless mesh is a misnomer. A real mesh system is a redundant system. The only way to have a redundant system with consumer wireless is to place all the wireless AP very close so if 1 fails you can use another. But the way consumer wireless mesh is installed is to push out the wireless as far as possible to extend range which takes away the redundancy.

Business wireless APs are more stable and more flexible than consumer wireless. You can run more units and multiple VLANs across all units like having a guest VLAN across all units and roaming is going to be better.

I like Cisco's low end in their small business wireless like Cisco's WAP581 wireless APs.

 

[Trip]

So you are saying there is no way to know the range of a router unless you actually buy it and test it for yourself?

100% correct if it's a first go-around without having any wifi gear on-site beforehand. If you have a pre-existing router or AP to compare against, then you can probably predict how well the new hardware would do in relation, but even so, the only true way to know with complete certainty is to survey with the new gear on-site.

Does business class AP's give wider range or more stable system? Does it feel any different browsing on your regular consumer router(ex ASUS) or a Ruckus AP?

That's a tough question to give a universal answer to. In general, though, when looking at a like-for-like swap, all other variables held equal, business-class APs actually won't provide as much effective range per single piece of hardware, because they're often designed to operate at lower amplification than your typical consumer all-in-one, whose sole purpose is to blast as much signal (and noise along with it) as possible. That being said, it's the client device and it's weaker radio that is almost always the limiting factor on how much actual throughput and/or connection quality can be attained from that far away. And that's where the often-superior approach of more, lower-power radios comes in -- precisely the paradigm for which most business APs are crafted. Granted, you have stuff like Asus AiMesh in the consumer space, but if distributed wifi is the point to begin with, I'd argue you'd be better off with the likes of even just an entry-level centralized business AP product (ie. TP-Link Omada) for generally better stability, native VLAN support, PoE support, better management, deeper feature set, etc.

Regarding reliability, though, yes, business-class APs will often be more stability-focused in their development, and so often will provide a more reliable experience for the clients. Various brand systems will offer these qualities to varying degrees, but overall, that's the trend when moving from consumer to business and then to enterprise-grade gear, in almost any product segment.

TL;DR on range - For a more complete understanding on how to get the best wifi performance for your needs, see this excellent Duckware article for a solid primer on the subject.

I thought of installing business class AP's but at $700 per AP, no wireless mesh, and I heard you must have an expensive yearly subscription its just not worth it.I am not sure if the price difference=the added benefit difference. The closest thing I got to was wired Ubiquiti, not sure if overhyped consumer APs or on par with Ruckus and Aruba.

I'm not sure what wireless product line you were looking at, but the above is, at best, an over-estimation of per-unit price, mesh capability and licensing fees, for the majority of business products, even for a good amount enterprise products. I personally use Ruckus APs, whose best-bang-for-the-buck R510's are ~$250 each on Amazon (albeit gray market), come with rock-solid wired and wireless multi-point mesh, a proper enterprise-class control plane, and run without a discrete controller, for free, with no yearly licensing fees at all, ever. And they're on the higher end of the cost spectrum. You can achieve similar capability with Omada, UniFi, Granstream GWN, Aruba Instant On, the list goes on -- all for less (MSRP at least). Sure, if you want to pay through the nose, you would do so with Cisco Aironet, Juniper Mist, AeroHive with a cloud NG subscription, Aruba with an Aruba Central subscription, etc. but you certainly don't have to, just to get rock-solid, high-speed-everywhere wifi at home. Not even close.

 

[coxhaus]

The Cisco small business wireless APs do not require any extra fees like their PRO line of wireless. You get free firmware updates for the life of the product for free.

The Cisco small business wireless WAP581 APs that I run support up to 16 units per cluster with no controller. This should be big enough to support any home user or small business.

 

[Trip]

Indeed, that's another option. The 581's are nice in that you get AC Wave 2 4x4 5Ghz / 3x3 2.4Ghz, smooth Qualcomm wifi, up to 2.5GbE backhaul and longer-than-average support, probably 2025 to 2027 (as they were released Aug 2017).

 

[digital10]

The Cisco small business wireless WAP581 APs that I run support up to 16 units per cluster with no controller. This should be big enough to support any home user or small business.

When you say "support up to 16 units per cluster with no controller" you mean it can handle 16 connected devices per AP?

Indeed, that's another option. The 581's are nice in that you get AC Wave 2 4x4 5Ghz / 3x3 2.4Ghz, smooth Qualcomm wifi, up to 2.5GbE backhaul and longer-than-average support, probably 2025 to 2027 (as they were released Aug 2017).

Thanks for clarifying. Can those business class APs connect into a wireless mesh with a dedicated backhaul like Netgear Orbi in a daisy chain? Just curious, I know they are built to be wired.

 

[coxhaus]

A cluster of APs so I mean 16 APs working as 1 virtual wireless unit.

One Cisco WAP581 can support 128 connected devices and 64 active talkers.

You can only daisy chain Cisco WAP581 APs 1 deep. Wired is the way to go.

 

[digital10]

A cluster of APs so I mean 16 APs working as 1 virtual wireless unit.

One Cisco WAP581 can support 128 connected devices and 64 active talkers.

You can only daisy chain Cisco WAP581 APs 1 deep. Wired is the way to go.

aha, excuse my ignorance, I thought so long as they are wired via ethernet you can have infinite APs with same SSID. I am bellow noob level when it comes to networking.

 

[Trip]

Thanks for clarifying. Can those business class APs connect into a wireless mesh with a dedicated backhaul like Netgear Orbi in a daisy chain? Just curious, I know they are built to be wired.

Not really, no. Mostly all business/enterprise class APs are dual-band, while consumer mesh products with a dedicated radio for backhaul are almost always tri-band (two 5Ghz radios). As @coxhaus said, Cisco WAP can only mesh out to a single node, and you'll find similar SMB wifi products adopt the same limitation. Higher-end enterprise wireless products will usually offer multi-layer mesh (two or more layers deep), but you're still losing half the original throughput and picking up additive latency per hop. Multi-layer systems combat the effects of this the best they can by usually having more advanced control planes that can do real-time radio re-purposing, channel optimization and split backhaul/fronthaul traffic flow over the same radios, but at the end of the day the limitations are still there, so you're really best off hard-wiring, or only going out to a single hop of mesh, two at the most, if you can help it.