Multipath fading Microwave

Abel.
To increase the output power will not always help. Remeber that if you exceed the maximim input power on your Receiver you will damage it. If you already have a RX level that sits close to the max Rx input level during normal transmission periods and you now change the antennae to larger ones with more gain, or install an amplifier you will damage the receiver. If you have a window to play with additional power, then do it, remeber going one size up with antennae at one end will increase the gain by about 4dB.
I suggest you try the following if your fade margin is already 40 dB:
1. Play with antennae heights to eliminate multipath reflections:
I will first check with a path loss tool what happens if you lower the antennae to below the reflected waves (which seems to me is the problem), or move the antennae above the reflected waves if possible (if the tower height allows). Normally only moving the one end is necessary.
2. Try diffaraction to block the multipath signals:
If you move the one dish down to the point where you get diffarction losses, your reflected waves should be gone – right. Then see what happens if I increase the antennae size to compensate for the loss in power introduced by the diffraction. In other words we have now created a grazing blockage in the path. I reccomend you try it and come back to me.
3. Dead blockage right in front of your antennae (sea link):
I once sorted a link across the sea out by dropping the antennae on the one end so low that a building +/- 5 meters away from the antennae blocked about a third of the bottom half of the antennae. We then installed a amplifier (3db gain)on the one side. We could also increased one antennae to give us the additional power we lost due to the blockage, but had a amplifier availible. So if you do not have a building make a plan, create an obsticle, maybe a mild steel plate covering the bottom third of the antennae as an experiment will do – with the problem you have it might just be a last resort plan. Otherwise go satelite my friend.
4. Passive repeaters:
Passive repeaters over long links are not an option. First the actual repeater must be very close to the one end, no further than 1/8 of the path. The total path lengh must be less than 20Km at 7Ghz. That is my feeling. It will not work over the sea, you will just introduce more multipath fading and losses along the path. We used a passive reflecter once to overcome a mountain ridge about 500 meter away from the tower, and we increased the size of the antennae on the far side from 1.2m to 2.4 meter to compensate for loss in power, the frequency was in the 2Ghz band. The tower was built on the wrong location.