| Using Fringe Multipath... Wow, I may have gotten a little...verbose...
Anyway, I'm looking for antenna suggestions that can perform to some degree in a environment with a crazy amount of mutipath. There, now you don't need to read the rest of this post...unless you're a masochist...or want to yell at me for not doing my home work. But if its the latter then "the dog ate it".
*Sigh* Where to begin...
I've been using a EVDO device on the fringe of Sprint's coverage in my area for a few years now. When I first plugged it in the connection manager said 1xRTT but couldn't make a connection. I bought a small whip and an adapter, moved it around the room but still couldn't connect. I was kind of familiar with a few common types of wifi antenna so I got a yagi and 35ft of LMR-400. Aiming was a pain, the connection manager kept flashing EVDO Rev.A but I still couldn't connect. Most the time it just seemed like faint spikes of signal coming were coming from..well everywhere. I eventually found 2-3 tiny points on the horizon that I could connect to, but after 10-20 minutes the device would overheat and power cycle. I was about ready to return what I could and be done with it all.
Then I read Jim's blog post about the use and utility of a 2.4ghz grid. After considering that my only real fallbacks for home internet was going to be either Hughsnet or an old POTS line that would barely hold a 26.2k connection(Gee, thanks verizon), needless to say I tried the grid. I Installed a 1/2" spacer and placed the grid under the second story roofline where the yagi had done its best, and after a few days of constant tweaking I found the heading/inclination that seemed to work best. It wasn't completely miraculous (I think the meter only showed -106), but I was able to connect most of the time and the device would only occasionally overheat and conk out. Still considering my other options, I decided it would do. There were low bitrates and the odd day when I couldn't connect at all but the only real unbearable pains were the aim (the grid seemed to get misaligned by strong winds, flying branches, continental drift, plus I think the local squirrels would play badminton up there on sundays) and wild ping spikes (every few seconds/minutes I'd get a lag of usually around 1000-3000ms, not totally detrimental but enough to create a pretty constant series of page timeouts and broken downloads).
About six months ago I got tired of the constant tweaking and reloading, and since I figure I'll have to change over to LTE at some point I thought I had better do something about my improving my signal beforehand. So I got 50ft of tower sections...but chickened out and only put up about 35ft (I fear lightning...and cable attenuation). Still, the grid was now over the second story roof and..according to an old USGS map..just higher then most of the nearby terrain elevations. And after all that...I could really tell no difference. Not that I was expecting much but still, disheartening. So I checked another map and found that the closest tower was on the farside of a nearby town, unfortunately the nearby town was on the far side of 5 miles of gently rolling forest. So I spun the grid around 180 degrees and pointed it directly at the tower 5.3 miles away, I also unboxed a wilson sleek that I'd gotten for roadtrips. And after all that I must say I saw the perhaps sharpest change that I've seen so far, a second bar on the device, faster bitrates, and a RSSI of -96. Unfortunately the latency spikes didn't go anywhere and the connection seems if anything even more temperamental. It seems to literally change with the wind(tonight I'm getting 20k, tomorrow, maybe a meg).
Anyway, I was talking with someone in the front yard a couple of weeks ago when they pointed out the grid (now peaking over the highest point of the roof) and asked if I was having any success with my little project, I was about to tell them that I was just going to have to relegate myself to the signal I'm getting. Then I looked down and noticed their phone they had just set on the hood of their car showed 3 bars, after staring with what I'm sure was no small amount incredulity at where "Sprint" & "EVDO" were listed on the screen I asked if I could borrow their phone for a minute. I then waved the phone around the yard like a dowsing wand, finding 1-2 bars of EVDO just about everywhere outside. I handed the phone back and asked them if they had signal inside the house they said no they had to wait til they were outside to use it. So,I started asking my more technical friends where did I go wrong. Eventually someone had a dumbfoundingly simple answer, beamwidth, he said that it sounded like 90% of my signal was multipath bounces and that the cellphone seemed to work simply because its built-in antenna didn't care which direction the signal was coming from.
I can honestly say I had not even considered that usable signal could come from multiple directions at the same time. Maybe I spent too much of my youth tinkering with the big dish in the backyard but, in my mind focus was a good thing as a usable signal originated from a infinitesimally small point in the distance and any thing not coming directly from that point would only cause interference. I thought I was trying to find a beam from a laser pointer, but assuming he's right I've really been trying to catch snowflakes in a blizzard. He suggested getting a panel and pointing it at the tower, but it seems like several of the companies that made them don't any more in favor of these short little log-periodic...things. Most of them have comparable gains and even higher beamwidths though(especially on the horizontal) when compared to the panels. Then again following that same logic, why not a omnidirectional?, I mean the two strongest signals were about 180 degrees apart... Hold it, isn't the same beam coming from opposite directions what causes nulling? Maybe I should upgrade to LTE sooner rather then later.. I've been dreading it due to the higher modulation and smaller spectrum "slices". But it would allow me to put up a second antenna, and wasn't there a part of the MIMO specification that was suppose to dramatically improve performance in multipath rich environments?
But then if I...maybe I could...But..Arrrg! my head...
Maybe I'm starting to over think.
If someone could be so kind as to do some saging, or perhaps sprinkle a little enlightenment, even drop some knowledge..
I would be ever so grateful. |
