[MVMA] The West, The East, and The Center

[Maurice Riggins]

Part III - The West

The MVMA has two high profile 2.4 GHz omnis on the west side; one at Miami Valley Hospital South (MVHS) and one on the KAS tower in north Sugarcreek Township. These neighbor with each other and the KAS omni neighbors with the N8NQH @ N8DCP omni. Unfortunately, as did DARA, they had useless neighbors degrading the LQ of all neighbors.

With walk-out access to the roof, MVHS has proved to be the MVMA flagship site. The Bullet/L-Com Omni there links well with the KAS omni if not the omni at N8DCP. NC8Q added a dished NSM3 at MVHS, which links well with a like node at his home in Kettering. It would be a good site to attempt links further south, which may require access for physical tweaking.

The KAS site has considerable potential. It's in the most populous west side of the mesh, surrounded by 2.4 GHz SISO neighbors. It's the Greene County Winlink site and the logical place for a MVMA Winlink Post Office. Ideally, our high profile omnis would not neighbor with each other, being linked with multiple directional nodes. However, KAS is not a climb-at-will site and it's felt the best we can hope for there is two nodes, an omni and only one single-cable, easy to install and aim NanoStation for linking.

Also unfortunately, like Vandalia in the North, the KAS omni has demonstrated degraded performance since it was installed quite awhile ago. It's a 2.4 GHz Rocket with filters between itself and the Ubiquiti MIMO omni and there is varying speculation as to what the problem may be. According to WA8APB at the last meeting, the MVMA has acquired a Bullet + Filter + L-Com Omni for KAS and Vandalia and a NSM3 for KAS.

To reduce LQ-degrading neighboring, the omnis at MVHS, KAS, and N8DCP were put in a "AREDN-W" SSID. BAM! Immediately, the end-to-end path through them went from totally unusable to a net 1 Mb/s (3 MB file in 28 seconds). That may not sound like much, but it's half the net throughput we get to W8GUC through an internet tunnel to his nearby DARA neighbor, and we know there's still something wrong with the KAS omni. It's a definite improvement.
 
 
Part IV - The East
 
MVMA has only one high profile omni on the east side, the West Xenia water tank omni, which neighbors with the XWARN omni, previously known as Xenia-North. With only two nodes in "AREDN-E" they have good link qualities. If we choose, W Xenia Omni should also be able to neighbor with an omni at the future Trebein/Fairgrounds fire station for still decent three-neighbors' LQs. As at KAS, MVMA is limited to two nodes here, but in this case it's a physical limitation. We can only reduce, but not eliminate omni neighboring.

Although there are no user nodes in Xenia, these omnis could provide good AUXCOMM mesh coverage in the event of another Xenia tornado. The major problem is they are not linked to the KAS site. It's imperative MVMA have a direct non-2.4 GHz link between its east and west omnis.

The MVMA West Xenia NSM5 DtD'd to the omni there was originally aimed toward MVMA's MVHS site. The N8NQH Bellbrook site lies on the line between them and KAS lies about 14º north of that line. This NSM5 was rotated north but needs to be rotated back, even if we have to pay to do that. The existing NSM5 should also be replaced with a NSM3 to link with the one bought for KAS. Installing a NSM5 XW at KAS would require pulling two additional cables--not something we want to ask for.


Part V - The Center
There were five privately owned omnis not involved in this omni LQ improvement exercise. The WA8APB omni recently went down suddenly in foul weather and may or may not be lightning fodder. That has left the KA8ZKR omni orphaned. An attempt to link it to Xenia failed. The KD8UQI omni tunnels to WA8APB-Services. The N8NQH Bellbrook omni and the omni it links with at Xenia South were untouched, keeping the "-Timz" and array nodes all together in the original "AREDN" SSID. These are all nodes not running 3.16.2.0. We can address linking them individually.


Summary

In recent discussions of AREDN topology, we generally agree omnis should be linked with directional nodes rather with one another. We recognize getting to that isn't affordable for the MVMA and we can only reduce but not totally eliminate omni neighboring. We've found much of our omni neighboring to be worse than useless--to actually detract from the link quality of needed omni neighboring. We've found no more than three total omnis neighbored may be our best trade-off for usable omni link quality. We should strive to link sites with a band that has more than one or two 10-20 MHz ham-exclusive channels--that's 3.4 GHz. We've delayed needed remediation for far too long. Let's move forward now.