How meaningful is dual-band?

[mickwong]

For a dual-band router, it is always true that the 5GHz signal is weaker than 2.4GHz, and especially obvious when you get farther away from it or behind a wall. The actual 5GHz throughput will also be much lower than theoretical value (and worse than 2.4GHz connections), even though the frequencies are less crowded and have less interference. In practical use, my devices seldom connect to 5GHz since the signal is weak compared to 2.4GHz. The only time I can benefit from 5GHz is when I sit next or very close to the router.

Does it really worth the money to buy a dual-band router? Any thoughts?

 

[thiggins]

All depends on your situation. For you, 5 GHz has no value.

 

[PrivateJoker]

5ghz isn't "weaker," they both leave the router (or a client device) with the same power. 5ghz attenuates more quickly over similar distance because the frequency is double the 2.4ghz (an overly simplistic analogy is how AM radio can go over far more lengthy distances in part because it is such a low frequency compared to FM). And literally speaking the 5ghz band is closer to 6ghz than it is to 5ghz.

I posted a research paper the other day showing how different building materials affected 2.4ghz vs 5ghz (in terms of attenuation and reflection) completely differently. The same wall might attenuate 5ghz more, and also reflect 5ghz far more - depending on the client and the entire signal path it might be able to utilize reflected signals.

In my experience, my 3 stream 5ghz clients tend to be able to maintain a higher transfer rate at the extremities of my house than a single stream 2.4ghz client. And medium range distance 2 channel 5ghz clients are about on par with single stream 2.4ghz at same length. Everyone's RF environments are totally different, so your clients, home, and AP position might yield different results.

Also in the US the double digit 5ghz channels are under regulations that can reduce their power so they don't interfere with some aircraft & weather radar systems. I'm not certain if those regulations are automatically imposed based solely on the 802.11d/h regulatory flags or if they only are enforced when nearby RF is sensed by the AP. Nevertheless, I always manually pick a triple digit 5ghz channel and can confirm that at least in a very unscientific test using an Apple Airport Express on a 2 digit & 3 digit 5ghz channel w/ my iPhone 5 single stream N client as the only test device that in the same exact physical position I was able to get significantly better WAN to WLAN speed tests on a higher channel.

If you search on the FCCs site there are a lot of things they are considering implementing like physical location awareness in consumer APs so they could check themselves against databases of registered licensed transmitters (ie weather & aircraft radar) and adjust their channel selection and output power if they knew they were in the geographic vicinity of such equipment.

 

[stevech]

For free space (line of sight) loss
double the frequency - add 6dB more loss.
A 200 ft. line of sight path, at 2.4GHz has 78dB attenuation.
At 5.8GHz, it's 86dB. - a bit more than 6dB as it's more than double.

The 6 or so dB disadvantage is a relatively small percentage of the attenuation due to distance. Some can be offset with the use of higher gain antennas and more sophisticated MIMO processing (called post-detection gain). Antenna gain (for directional antennas) tends to increase with frequency for the same antenna design and size.

Attenuation from materials (drywall, wood floor, masonry) differs too at the two frequencies. One definitive report was done some years back by the National Institute of Standards (NIST) in the US. It's very informative; NIST and M.I.T. collaborated. They carefully characterized a wide variety of construction materials: drywall, brick. several recipes of poured concrete, block walls, wet/dry plywood and hardwood, etc. Their measurements are in "transmission coefficient" such as free space is 1.0, drywall type x at freq. y is, say, 0.8, and so on. In most cases, the translate the numbers to decibels (dB) for us RF types.

4MB PDF doc. (NIST web site closed like most others due to the current political temper tantrums)

NISTIR 6055
NIST Construction Automation Program
Report No. 3
Electromagnetic Signal Attenuation
in Construction Materials

http://www.nist.gov/manuscript-publication-search.cfm?pub_id=909358

For very large-building indoor cellular distributed antenna systems, and for WiFi, in my field we often use the NIST report's number along with AutoCAD drawing files to quickly build a 3D model of indoor coverage, where autoCAD wall type 1 is mapped to a NIST material attenuation, floor type 2, to another, etc. And we see the AutoCAD files where things like floors as concrete poured over a steel pan exist- makes a big difference.

 

[wrugoin]

How meaningful? Well, using inSSIDer, If you look at the available wireless connections in my neighborhood, I see 12 connections. Over 90% of them use the 2.4GHz band. Finding an unused or even uncluttered channel on 2.4 is difficult. The 5GHz band is almost empty. One other home has a connection.

Just from personal experience, my 5GHz band is considerably faster and more reliable in my home than the 2.4GHz band. The signal strength is similar according to inSSIDer. File transfers are much faster (20MB/s vs 1.5MB/s) on the 5GHz band.

Just for reference. I'm using an Asus RT-AC66U router and connecting it with various adapters. My wireless router is on the first level of the home and the computers it is networked with are both on the 2nd level.

 

[Razor512]

In my environment, the 5GHz works much better than 2.4GHz. (while the 5GHz does get a bit weak when in areas like the back yard, while the 2.4GHz will be stronger. 2.4GHz loses some of it's reliability and becomes slow (might be due to the 150+ access points in the area).

5GHz used to be completely clear (scanning the 5GHz would show nothing other than my network when in the house, now there is a 5GHz time warner access point (both 2.4GHz and 5GHz). I wish my router had the range of that time warner access point. (it is nearly 600 feet away and I can connect to it from the house)

Why cant they do this with home routers?

To keep good performance throughout the house and in the yard, I am using 2 decent routers that can work on the 5GHz band. (while disabling the 20/40 MHz coexistence makes the 2.4GHz faster, it still does not match the performance of the 5GHz band)

 

[stevech]

How meaningful? Well, using inSSIDer, If you look at the available wireless connections in my neighborhood, I see 12 connections. Over 90% of them use the 2.4GHz band. Finding an unused or even uncluttered channel on 2.4 is difficult. The 5GHz band is almost empty. One other home has a connection.

Indeed, 5.8GHz is less used. So far.

What we need is a freeware tool to dwell on each SSID detected and report how "busy" that SSID is. Do this in the evening busy hours.

It's not how many SSIDs you detect - it's if any are quite busy with data flows, and on/near your chosen channel. A neighbor streaming HD video (not Netflix) frequently- would be displayed with a tool as above. And you'd simply change channels. Such tools exist, but are costly; I've yet to see a free one that reports channel utilization as 0 to 100% by SSID.

 

[PrivateJoker]

I wish my router had the range of that time warner access point. (it is nearly 600 feet away and I can connect to it from the house)

Why cant they do this with home routers?

Go look back in that thread where you first mentioned that and find the link I posted to another AP that TWC uses for public hotspots (I could only find specs on the Ruckus unit, not any Cisco equip, but they use both), it's $5k!

Give me $5k, and an I beam to mount it from, and an 8 channel Docsis 3.0 back haul, and I will produce for you an amazing access point.

 

[stevech]

Cisco and Tropos went after this Cable DOCSIS as backhaul, and WiFi on DOCSIS backhaul for public areas like parks.

This came a bit after Earthlink's investments in muni-WiFi as it was called - in Philadelphia, Houston, San Francisco, and a couple of others. Then Earthlink conceded that the business model for revenue just didn't work, and cancelled all their projects. They were using Tropos mesh network WiFi with dual radios, and Motorola's Canopy for multi-route backhaul (capacity injection they called it).

The WiFi access links were about 200mW into a 5.6dBi antenna. Lots of forward power. Strong signal. But of course, they could penetrate the walls into the residences, but not so on the reverse link from the 30mW clients. Along comes Rukus and another, to provide bridges/repeaters, to go in a window and repeat signals from the wee clients. All of this collapsed of its own weight. The premise was that Federal Funding for eliminating the "digital divide" in low income areas would make this lucrative. But it didn't go that way. Then in its death throws, cities tried to re-purpose the WiFi systems to serve the needs of public works, cops, etc. Some of that remains.

Engineering coverage with a giant mesh network, citywide, county wide, was quite a challenge. Of course, some complained that the systems branded as City of xxx WiFi, jammed out too much of the spectrum. True.

The WiFi via cable DOCSIS is a good idea, but small in scale by comparison.

 

[Razor512]

Go look back in that thread where you first mentioned that and find the link I posted to another AP that TWC uses for public hotspots (I could only find specs on the Ruckus unit, not any Cisco equip, but they use both), it's $5k!

Give me $5k, and an I beam to mount it from, and an 8 channel Docsis 3.0 back haul, and I will produce for you an amazing access point.

yep, they just need to bring that price tag down to less than $200 and get rid of the modem part and just make it a dedicated router, or access point.

The thing I don't get is why aren't these companies using these improvements on their home consumer routers? like in the last post, the one near me is a cisco and likely also very expensive, but cisco also makes (or used to) routers for home users.

Why cant they just walk over to the department that made those access points that these ISP's are mounting outdoors, and just take th wireless module and stuff it into the home/ consumer routers

no need for the weather sealing or power over ethernet or any enterprise features, just the wifi radio and antenna.

If they were to do just that, they can still easily maintain a sub $200 price point and offer good wifi range without all of the R&D cost since they will be using a design from a product that they already have, and is proven to offer good range.

 

[sm00thpapa]

Same room upstairs and my daughters laptop using inSSIDer gets -40 signal on the 2.4 GHz band and -70 signal on the 5 GHz band. But it has no effect on her using her laptop. Downloads and uploads are equally as fast.

 

[stevech]

Same room upstairs and my daughters laptop using inSSIDer gets -40 signal on the 2.4 GHz band and -70 signal on the 5 GHz band. But it has no effect on her using her laptop. Downloads and uploads are equally as fast.

-70dBm will probably yield WiFi speeds well in excess of your ISP's speed provided.