On Fri, 30 Jan 2004 [EMAIL PROTECTED] wrote:
> -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > My first question would be has anyone on this list had expierence with > locating a wireless node on a City water tank? Nobut I have sure thought of it! > As a WISP in a rural area, we have one node located on the top of a water > tank. This node has a point to point back to the central node, and an > access point with an omni to provide service to the surrounding valley. How separated in frequency is the PtP link and the local AP? Do you have notch filters to prevent the signal from one bleeding into the other? How much power are you running? > Lately, I have been having some really odd problems which I have been > scrambling to find a solution for. It appears to manifest itself in > the form of service outages for various people, at various times, in > various locations. While one customer is working fine, another is completely > dead in the water, there is no real pattern to any of the service outages. Sounds like a BSSID split, but you say you are using an access point so that indicates infrastructure mode, not ad-hoc. hrm. > Monitoring the point to point proves that it is solid. There is no packet > loss, and no problem pulling data through it. > > The entire infastructure has been rebuilt multiple times, From cable > replacements, to antenna replacements to new bridging and accesspoint > equipment. > > The Omni stands about 10 feet high off of the top of the water tank, > which I would estimate to be a 30 foot diameter circle of steel. We > have tried lowering, and adding height to the omni as well. There are fomulas for antenna height over perfect ground, which is what a circle of steel is. I will try to dig them up. > I would like to know if anyone has any ideas as to why this node would > perform so diferently than my other nodes. I am able to provide a great > service on all of my other locations. This one I have honestly been > fighting with for the past 6 months. > > I can elaborate on anything needed to help analyze the problem. I didn't > want to go too deep into things untill I found if people had an interest. > > Thank you kindly for your time. I appreciate any help. Perhaps if you can run kismet or some other type of diagnostics tool at the sites when they are offline, to see if packets are coming in. Also the water-tank unit -- can you run kismet on that? I imagine not, it's probably some embedded device not linux or bsd. It would be interesting if the level of water in the tank affected your signal. Is there any way to tell? -- general wireless list, a bawug thing <http://www.bawug.org/> [un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless
