MIMO at 500m distance?

I'm trying to establish a high-speed PtP link at around 500m distance using 5G wifi and while I get MCS9 (200Mbps) I can't get NSS2 (two spatial streams) to perform. What does it take to get NSS2 if it's even possible?

Equipment:

With the Alfa antennas at each end I get the following ("baseline"):

        channel 44 (5220 MHz), width: 40 MHz, center1: 5230 MHz
        txpower 23.00 dBm
        signal:         -53 [58, -55] dBm
        signal avg:     -53 [-58, -56] dBm
        beacon signal avg:      -53 dBm
        tx bitrate:     81.0 MBit/s VHT-MCS 4 40MHz VHT-NSS 1
        rx bitrate:     200.0 MBit/s VHT-MCS 9 40MHz short GI VHT-NSS 1

I'm primarily concerned about the RX bitrate (which is download from Internet), the fact that upload is slower is a bit curious, but the other end is likely to have more interference.

Sometimes NSS2 "turns on" but then the data rate drops:

        signal:         -61 [-64, -63] dBm
        signal avg:     -55 [-61, -59] dBm
        beacon signal avg:      -53 dBm
        tx bitrate:     108.0 MBit/s VHT-MCS 3 40MHz VHT-NSS 2
        tx duration:    9098309 us
        rx bitrate:     200.0 MBit/s VHT-MCS 9 40MHz short GI VHT-NSS 1

Here TX went from MCS4 NSS1 to MCS3 NSS2 for a small gain (81 vs 108 Mbps). Sometimes RX drops to MCS5 NSS for also a small gain (200 vs 240 Mbps).

I replaced one of the panel antennas by the two grid antennas mounted stacked vertically about 1.5m apart. I first mounted the two antenna cross-polarized (one vertical, one horizontal) but then saw that the horizontally polarized one got much worse signal strength, so I turned it to also be vertical. I thought different polarization would be one way to get good stream separation but I guess the Alfa panel only has vertical polarization, ugh.

The result with the grid antennas was higher signal strength (around -56dBm) but no real change in bitrate. It also was mostly stuck on MCS9-NSS1. Note that the noise level as reported in the Web interface is between -96dBm and -101dBm in all cases.

I also tried channel 157 and got the allowed 30dBm tx power but aside from getting a signal strength just above -50dBm there was no change in terms of effective bit rate.

Is it possible to get something like MCS7-NSS2 at that distance and what does it take to get there? It can't be a signal strength issue 'cause -50dBm is screaming loud. Do I need to move the grids much further apart? Do I have to get antennas that are properly cross-polarized? Anyone have experience?

You need to fill wifi distance box.
please post

ubus call system board
cat /etc/config/wireless

blanking secrets.

1 Like
root@wcc-ap:~# ubus call system board
{
        "kernel": "5.15.167",
        "hostname": "wcc-ap",
        "system": "Qualcomm Atheros QCA956X ver 1 rev 0",
        "model": "Ubiquiti UniFi AC Mesh",
        "board_name": "ubnt,unifiac-mesh",
        "rootfs_type": "squashfs",
        "release": {
                "distribution": "OpenWrt",
                "version": "23.05.5",
                "revision": "r24106-10cc5fcd00",
                "target": "ath79/generic",
                "description": "OpenWrt 23.05.5 r24106-10cc5fcd00"
        }
}
root@wcc-ap:~# cat /etc/config/wireless

config wifi-device 'radio0'
        option type 'mac80211'
        option path 'pci0000:00/0000:00:00.0'
        option channel 'auto'
        option band '5g'
        option htmode 'VHT40'
        option cell_density '0'
        option country 'US'

config wifi-device 'radio1'
        option type 'mac80211'
        option path 'platform/ahb/18100000.wmac'
        option channel '1'
        option band '2g'
        option htmode 'HT20'
        option cell_density '0'
        option disabled '1'

config wifi-iface 'wifinet1'
        option device 'radio0'
        option mode 'sta'
        option network 'wwan'
        option ssid 'xxx'
        option encryption 'psk2'
        option key 'xxx'

config wifi-iface 'wifinet2'
        option device 'radio1'
        option mode 'sta'
        option network 'wan2'
        option ssid 'xxx'
        option encryption 'psk2'
        option key 'xxx'
        option ifname 'sta24'
        option disabled '1'

This is the client, the AP is managed through Unifi (doesn't run OpenWRT). The link overall is stable, I can pump ~130Mbps through when it's at MCS9-NSS1 (200Mbps). My experience with too low an ACK timeout is a rather unstable connection.

You have to add

option distance '500`

5o both long range radio sections to increase wifi ack tolerances.
ref
https://openwrt.org/docs/guide-user/network/wifi/basic

Makes no difference and:

If distance is too shotr it starts counting beacon losses and packet losses in ethtool -S
Used to be 200m .

ac mesh says 4x4?

tx failed 17 packets out of 517k, beacon losses 0, but ack timeout doesn't apply to beacons since beacons are not acked.
These devices are 2x2.

1 Like

Check
grep _cap /var/run/hostapd*conf

Should not say anything more than 2 antennas or one beam

Other is that one dimentsion is missaligned, 1mm here is one house apart in 1km