Can somebody help me understand what might be happening with my new PTP link?
When we put it up, the slave showed an RSSI of around 50 and the master showed an RSSI of about 75 at a distance of 7mi. Now a few days later, the slave is showing an RSSI of between 72 and 81 which fluctuates wildly every second. The link constantly downshifts to MCS1/0 and has tons of packet errors. The spectrum analysis shows a clean frequency range on both sides and the master seems to be dialed in appropriately; we haven't checked the slave yet.
I'm still unsure how to extract a guess about which side of the link problems like this stem from from when looking at performance statistics. SNR differences? RSSI differences? I have attached PDFs of the performance statistics of both ends.
It might be helpful to see the Monitor->Wireless stats from each end, and the Configuration->Radio page.
That said, if it actually started at -50dB and is now at -72 to -80 dB then something is causing the signal to be weaker - not interference, which would decrease the SNR but wouldn't impact the received signal level. Two main possibilities - either you have the radio on ACS and it is now on a channel with more restricted transmit power than where it started, or something has physically changed (like a radio moved), or something has moved into the LoS like a tree branch sagging down in front of a tree-mounted radio. (we see that regularly with SMs, but since this is a PtP link you're talking about I guess that's less likely as a mounting location)
As for narrowing things down to one end, if (for example)the uplink RSSI is much weaker than the downlink RSSI, that often suggests that the transmit power on the slave is reduced (automatically via coordination with the master, or manually set to a lower transmit power or a higher antenna gain). The SNR is just the difference between the working signal and the noise floor so assuming RSSI is about equal each direction, if the SNR is noticeably lower on one end, that tells you the receiving radio at that end is seeing an interfering signal of some sort. (or a stronger interfering signal than the other end of the link) Since this is a Force200 you can't sync it or change the antenna, so the only recourse for a too-strong interfering signal is to choose a clearer channel.
Generally with a well-aligned link you should only see a significant difference between RSSI uplink and downlink if there's a significant difference in transmit power between uplink and downlink.
Can somebody help me understand what might be happening with my new PTP link?
When we put it up, the slave showed an RSSI of around 50 and the master showed an RSSI of about 75 at a distance of 7mi. Now a few days later, the slave is showing an RSSI of between 72 and 81 which fluctuates wildly every second. The link constantly downshifts to MCS1/0 and has tons of packet errors. The spectrum analysis shows a clean frequency range on both sides and the master seems to be dialed in appropriately; we haven't checked the slave yet.
I'm still unsure how to extract a guess about which side of the link problems like this stem from from when looking at performance statistics. SNR differences? RSSI differences? I have attached PDFs of the performance statistics of both ends.
Are you sure you have 100% LOS? If not, the tops of swaying trees may cause the rapid RSSI fluctuations. That should and would effect RSSI on both ends, not one.
In my opinion..... there does seem to be some interference present due to some of the information contained in the pagesyou did share.
Constructive and/or destructive radio signals (interference) on the same frequency, but with phase changes, due to reflections/refraction can cause the rapid fluctuations in RSSI levels you are seeing. It can also cause the retrans and error drops you are seeing. If one end can not "hear" what it needs to "hear" due to interference it replies back to sending end, "I did not hear you, please say again",which triggers a retrans. We are seeing ALL of this on both ends for your link. Interference on 1 end can cause retrans on both ends.
What is particularly interesting is on both ends, your downlink (master to slave) MCS does not add up to 100% on either end. It adds up to about 90%. 10% of the time there are modulation rates that are not accounted for. At the same time your uplink (slave to master) adds up to 100%.
The fluctuation in RSSI at the slave end and missing percentages of MCS levels on your downlink -seem- to point to me that there is some unkown interference on the slave end. This is going by the limited information that was posted. This may only be part of your proble though.
I would like to see wireless performance, edetect, and spec-an graphs(you will need to turn opposite end of link off to get accurate spec-an results).