4x4 2,4 Ghz increase coverage?

[tiomiguel]

If I want a AP with maximum coverage, in Europe, In some sites say 4x4 radio increase coverage over 2x2 radio.

My question is if this is true or not

In a WLAN with IOT devices far from AP, 4x4 radio increase REAL coverage over 2x2 radio OR NOT?

Many domestic router have 2x2 radio in 2,4 Ghz (but 4x4 radio in 5 Ghz). Only high end routers have 4x4 radio in 2,4 ghz

I owned unifi UAP-AC-LR with 3x3 in 2,4 Ghz but coverage is similar to RT-AX55 or RT-AX57 that are 2x2

For example, XIAOMI AX-3200 with 4x4 at 2,4Ghz would increase REAL coverage?

 

[microchip]

Not true. 4x4 and 2x2 says something about the amount of simultanious streams that can be used, but nothing about coverage.

 

[tiomiguel]

Are you sure?

Some specilaliced sotes soy like this:

Diversity: If there are more antennas than streams (eg: 2×2 client to 4×4 router),the 'extra' antennas can then be used to improve link quality, and increase range. With multipleantennas receiving the same transmitted signal, the receiver can recombine all of the receivedsignals into a better estimate of the true transmitted signal.
FCC documents discuss that the 'maximum' gain when doubling antennas is 10×log(NANT/NSS) dBi, which for a 2×2 client to a 4×4 access point, would result in a diversity gain of 'around' 3 dBi. OR a 4×4 access point to a 1×1 client means a diversity gain 'around' 6 dBi.
This explain why you really do want a 4×4 MIMO router, even though there may only be 1×1 and 2×2 client devices connecting to it!

 

[Tech9]

@microchip is correct. In reality newer more sensitive 2x2 radio may have better range than older 4x4 radio. I’ve seen examples of this multiple times. More spatial streams supported doesn’t translate automatically to better range. The theory applies only when all other variables are the same, but this is not the case.

 

[TheLostSwede]

Range depends a lot more on antenna design and how good the signal is, but you might be able to cover some dead spots with a 4x4 router that a 2x2 router can't catch. Other factors are things like the power amplifiers used, even though you're a lot more limited in Europe than the US.
Technically, you could swap to different antennas and potentially gain some range, especially if you get direction antennas, but then you lose coverage instead.

 

[Tech9]

@tiomiguel has RT-AX57. This model comes with non-removable antennas.

 

[sfx2000]

TLDR - A 4-stream AP _may_ help with range, and speed over a given range compared to a 2 or 3 stream AP... and it could improve things by 25 to 50 percent in optimal conditions...

In the analog domain, the extra radio adds about 3 dB for a three stream AP, and another 1.5 dB or so for the 4th radio.

So it can be up to 4 to 4.5 dB of better Receive at the AP, just at the radio layer...

At the baseband/digital domain - this really depends on the chipset implementation - Broadcom and QC-Atheros do a good job here of combining the inputs and getting a bit of coding gain - let's say another 1-2 dB here as well...

And this is totally outside of the antennas, but doesn't take in to consideration of internal noise...

So it's safe to say that one could see a reasonable 3dB improvement on the receive side going from 2-stream to a 4-stream AP...

Where it gets interesting is on the Tx side - it's because of a quirk in the regulatory specs - we have a value called Effective Isotropic Radiated Power (EIRP) - this take a single theoretical point in space, and how much power can be transmitted from there - and this doesn't make any difference between a single stream radio or an 8-stream device - total amount of radiated power is just that, not per transmit radio...

Going back to the baseband/digital layer - the streams are combined and interleaved - so if a 1-stream or 2-stream client is present, there's going to be some repetition of symbols across the streams, and the client combines them going up into the MAC and the magic of MIMO...

So Tx can see some gain here as well due to digital coding gain due to stream combining and client recovery... this could be up to 3 dB, but we get noise limited due to both external interference and multipath returns (generally caused by both Rayliegh and Rician fading)

 

[collector930]

TLDR - A 4-stream AP _may_ help with range, and speed over a given range compared to a 2 or 3 stream AP... and it could improve things by 25 to 50 percent in optimal conditions...

In the analog domain, the extra radio adds about 3 dB for a three stream AP, and another 1.5 dB or so for the 4th radio.

So it can be up to 4 to 4.5 dB of better Receive at the AP, just at the radio layer...

At the baseband/digital domain - this really depends on the chipset implementation - Broadcom and QC-Atheros do a good job here of combining the inputs and getting a bit of coding gain - let's say another 1-2 dB here as well...

And this is totally outside of the antennas, but doesn't take in to consideration of internal noise...

So it's safe to say that one could see a reasonable 3dB improvement on the receive side going from 2-stream to a 4-stream AP...

Where it gets interesting is on the Tx side - it's because of a quirk in the regulatory specs - we have a value called Effective Isotropic Radiated Power (EIRP) - this take a single theoretical point in space, and how much power can be transmitted from there - and this doesn't make any difference between a single stream radio or an 8-stream device - total amount of radiated power is just that, not per transmit radio...

Going back to the baseband/digital layer - the streams are combined and interleaved - so if a 1-stream or 2-stream client is present, there's going to be some repetition of symbols across the streams, and the client combines them going up into the MAC and the magic of MIMO...

So Tx can see some gain here as well due to digital coding gain due to stream combining and client recovery... this could be up to 3 dB, but we get noise limited due to both external interference and multipath returns (generally caused by both Rayliegh and Rician fading)

So I have some data to support 25% to 50% improvement of 4x4 vs 2x2, but for 5Ghz.

I tested an Asus XT9 router for Wi-Fi download speed using 2x2 clients. It has both a 2x2 band (5Ghz-1, for fronthaul) and a 4x4 band (5Ghz-2, for backhaul). I found the further the distance, the higher the improvement in speed for the 4x4 compared to the 2x2. For example, @ 15 feet , the 4x4 was faster by about 25%, @ 25 feet (through 1 wall), the 4x4 was faster by about 35% and @ 35 feet (through 2 walls), the 4x4 was faster by about 50%.

Also, I tested a TP-Link AXE75 (Wi-Fi 6E) router and a Deco BE11000 (Wi-Fi 7) router, both of which have a 2x2 5Ghz band. Their 5Ghz speed was about the same as the 2x2 5Ghz-1 band in the XT9, within 5% to 10%.

Finally, I tested an Asus XD6 router, which has a single 4x4 5Ghz band, and its speed is almost identical to the XT9's 4x4 5Ghz-2 band. Note: All of the clients tested were 2x2 AC/AX devices; also, the XD6 uses an older 4x4 Broadcom chip, BCM43684, compared to the XT9's newer 4x4 Broadcom chip, BCM6715.

My guess is that 4x4 allows better beamforming compared to 2x2, i.e., higher efficiency by using the same total radiated Tx power, but more sharply focused in a particular direction.

 

[Tech9]

I tested an Asus XT9 router for Wi-Fi download speed using 2x2 clients. It has both a 2x2 band (5Ghz-1, for fronthaul) and a 4x4 band (5Ghz-2, for backhaul). I found the further the distance, the higher the improvement in speed for the 4x4 compared to the 2x2. For example, @ 15 feet , the 4x4 was faster by about 25%, @ 25 feet (through 1 wall), the 4x4 was faster by about 35% and @ 35 feet (through 2 walls), the 4x4 was faster by about 50%.

This is not a valid experiment because on XT9 there is no way to test 5GHz-1 and 5GHz-2 radios on the same channel and the radios are different. One of the radios is locked to lower 5GHz band (built-in BCM6756), the other to mid-upper 5GHz band (additional IC, perhaps BCM43684). You have a huge environment factor variable including different permitted power on different channels in different regions.

 

[collector930]

This is not a valid experiment because on XT9 there is no way to test 5GHz-1 and 5GHz-2 radios on the same channel and the radios are different. One of the radios is locked to lower 5GHz band (built-in BCM6756), the other to mid-upper 5GHz band (additional IC, perhaps BCM43684). You have a huge environment factor variable including different permitted power on different channels in different regions.

Yes, it is true that in the XT-9, the 2x2 can only use the lower 5Ghz band (36-64) and the 4x4 can only use the mid-upper 5Ghz band (100-165).

However, here in the US, where I did the tests, the lower non-DFS channels (36-48), which the XT9 2x2 were using during test, are limited to the same 1000mW Tx Power as the upper non-DFS channels (149-165), which the XT9 4x4 were using during test. Also, I did the tests in a house and not an apartment, so I did not have a lot of interference from neighbors.

But in other regions like the EU usually the upper 5Ghz band can have much higher Tx Power than the lower 5Ghz band (i.e, 4000 mW vs 200 mW), for non-DFS channels, so the XT9 test would not be valid in those regions.

In any case, in my testing of different brands of routers with different chipsets, in general I still observed much better performance for 4x4 vs 2x2 for the 5Ghz band at > 20ft (up to 50% higher @35ft thru 2 walls), even for those with much older 4x4 chipsets (Asus XD6 and AX5400 w/ BCM43684) vs newer 2x2 chipsets (XT9 and AXE75 w/ BCM6756, Deco BE11000 w/ Qualcomm NP 620). Wish I had known that before I started shopping for a new Wi-Fi router this year. I was under the impression that a 4x4 router would provide little to no benefit for 2x2 clients.

 

[Tech9]

Perhaps valid for consumer devices built around single AP maximum coverage design idea, but in my own place the multi-AP system was upgraded few times from 2x2 (Cisco) to 3x3 (Ruckus) and now 4x4 (Ruckus) and the difference in coverage area is not that obvious. There is a difference in speeds at distance, but more likely related to BeamFlex the newer APs are using, a form of client independent hardware beamforming using complex antenna arrays. All my APs work at or under 200mW, in another place down to 20mW. So different hardware - different experience.

 

[Metroguy]

Perhaps valid for consumer devices built around single AP maximum coverage design idea, but in my own place the multi-AP system was upgraded few times from 2x2 (Cisco) to 3x3 (Ruckus) and now 4x4 (Ruckus) and the difference in coverage area is not that obvious. There is a difference in speeds at distance, but more likely related to BeamFlex the newer APs are using, a form of client independent hardware beamforming using complex antenna arrays. All my APs work at or under 200mW, in another place down to 20mW. So different hardware - different experience.

So is 2x2 WiFi router release in 2024 like the Asus RT be58u is better than a Asus RT-88u 4x4?

 

[Tech9]

Only users who have both can compare and tell.

 

[Metroguy]

Yes, it is true that in the XT-9, the 2x2 can only use the lower 5Ghz band (36-64) and the 4x4 can only use the mid-upper 5Ghz band (100-165).

However, here in the US, where I did the tests, the lower non-DFS channels (36-48), which the XT9 2x2 were using during test, are limited to the same 1000mW Tx Power as the upper non-DFS channels (149-165), which the XT9 4x4 were using during test. Also, I did the tests in a house and not an apartment, so I did not have a lot of interference from neighbors.

But in other regions like the EU usually the upper 5Ghz band can have much higher Tx Power than the lower 5Ghz band (i.e, 4000 mW vs 200 mW), for non-DFS channels, so the XT9 test would not be valid in those regions.

In any case, in my testing of different brands of routers with different chipsets, in general I still observed much better performance for 4x4 vs 2x2 for the 5Ghz band at > 20ft (up to 50% higher @35ft thru 2 walls), even for those with much older 4x4 chipsets (Asus XD6 and AX5400 w/ BCM43684) vs newer 2x2 chipsets (XT9 and AXE75 w/ BCM6756, Deco BE11000 w/ Qualcomm NP 620). Wish I had known that before I started shopping for a new Wi-Fi router this year. I was under the impression that a 4x4 router would provide little to no benefit for 2x2 clients.

Will a 2x2 be WiFi 7 router in 2024 be more better than a 4x4 AC router from 2015?

 

[degrub]

only for 2x2 wifi7 clients.

 

[Metroguy]

Only users who have both can compare and tell.

I had 2 routers 3x3 and 4x4 and used them interchangeable once. Asus RT 68u(3x3) and RT 88u(4x4).

Found that they were identical in signal coverage for my area and no differences whatever but it seems that the 3x3 AC 68u , despite its lower hardware soecs is a tad bit faster at loading and snappier than the RT Ac88u. On searching this forums came across a post that 4x4 will introduce more signal noise or congestion into the environment or something along those lines explaination.

 

[Metroguy]

only for 2x2 wifi7 clients.

I have only mostly 2x2 ac devices. So a 2x2 WiFi 7 router with double specs ( quad core and 1gb ram) more than a AC router Wil not benefit AC devices?

 

[collector930]

I would agree with degrub.

For my 2x2 ac devices I tested one 2x2 WiFi 7 router, the Deco BE11000, Hardware Version 2.6 (from Costco), and compared it to my current 4x4 WiFi 6 router, the Asus XD6. The Asus was much faster (30%-50%), despite having a 1.5Ghz 32bit tri-core CPU vs 2.2Ghz 64bit quad-core CPU, though during the speed tests I noticed the CPU utilization was about 0% for the Asus. Basically I found the 2x2 BE11000 was about the same in terms of speed as the 2x2 TP-Link AXE75 (WiFi 6E).

 

[Metroguy]

I would agree with degrub.

For my 2x2 ac devices I tested one 2x2 WiFi 7 router, the Deco BE11000, Hardware Version 2.6 (from Costco), and compared it to my current 4x4 WiFi 6 router, the Asus XD6. The Asus was much faster (30%-50%), despite having a 1.5Ghz 32bit tri-core CPU vs 2.2Ghz 64bit quad-core CPU, though during the speed tests I noticed the CPU utilization was about 0% for the Asus. Basically I found the 2x2 BE11000 was about the same in terms of speed as the 2x2 TP-Link AXE75 (WiFi 6E).

It's a bit flawed when u test different brand routers for comparison with regard to mimo characteristics. For example to tplink or deco may have lower signal/ tx sensitivity compared to Asus (-68 Vs -98). This is where Asus wins most of the time over other brands. According to someone in this forum

U can have a 2x2 antenna with good sensitivity or a 4x4 antennas with lousy sensitivity. The 2x2 can win the 4x4 anyday with a higher power design

 

[sfx2000]

This is where Asus wins most of the time over other brands. According to someone in this forum

Not asus specific...

There are a lot of AP's and Routers that do 4*4:4 - my particular favorite at the moment is the MediaTek Filogic 830 based GL-Inet MT6000