Hi Guys, Recently the marina installed 4 wifi somethings. I don't know if they are repeaters or extenders or if they are the same. The model is here;
formatting link
I have one of the >UniFi AP-Outdoor< very near my boat. It's shown in the pdf. This thing has two antennas, is one receive one transmit? Does this thing receive on one channel and transmit on another? My question is how does this system work, especially with four.
Speedtest.com says I have 0.98Mbps download and 0.48Mbps upload, lousy, but it works. Mikek
Neither. My guess(tm) is that they are access points. Is there any CAT5 cables going to them? If so, they're AP's. If no cable, they are repeaters.
That would be the UAP-Outdoor+ or the UAP-Outdoor5 models. The UAP-Outdoor+ is 2.4GHz only. The UAP-Outdoor5 is 5GHz only.
No. It use 2x2 MIMO for 802.11n with one radio per antenna.
No. This one is strictly store and forward.
Lousy. If it's a bunch of repeaters, you have the equivalent of a mesh network. See my previous rants on the evils of repeaters. The only benefit to a repeater/mesh is to eliminate the "expensive" wired backhaul. Everything else about them is detrimental to performance and capacity.
Mesh networks (and repeaters) have been around for a long time. One would think that with all their problems, the technology would just roll over and die by now. Not so. From the pages of Wired magazine: As usual, the article screws up the problems, giving exactly one sentence to the real problem and the rest to technical faith healing and unrealistic expectations: Although the technology is there, routing protocols are currently unable to scale over a few hundred nodes and network coverage is constrained by the limited range of wireless user devices. Incidentally, that's Layer 2 (MAC address) routing, which really should be called switching protocol. 802.11 works on Layer 2 and has little to do on Layer 3 (TCP/IP) where most routing protocols live. Anyway, the switching/routing/whatever protocol is both the problem and the solution. Yet, despite numerous patents and years of active field testing, wireless mesh (and repeaters) are still roughly where they were 10 years ago. They don't scale, they're polluted by broadcasts and management packets, performance sucks, and create plenty of self-interference.
My favorite demonstration is to setup a simple speed test with an access point and a laptop in an isolated area, such as a typical room. After getting benchmarks, I add one or more repeaters to the room. The benchmark speeds will drop, even though the access point and the laptop are still directly associated. That's because the repeaters are STUPID and will repeat everything they hear (with the same common SSID). They have no way to know that the access point and laptop are directly connected, thus wasting air time by repeating everything they hear even if nobody needs to hear the duplicated packets. Even if repeaters worked in the ideal situation, where the access point and the laptop cannot see each other, and the repeater can see both, the maximum speed of the connection is cut in half (or more) because each packet in both directions has to be sent and received twice, thus occupying twice the air time. Add a 2nd repeater to the chain, and the speed again gets cut in half. Think binary progression.
That's consistent with the lousy performance of repeaters and mesh networks. If you fire up a wireless sniffer, capture a few minutes of traffic, and look at the packets, you'll probably see numerous retransmissions of the same data. Ask the marina to add more repeaters so that the system will go into total collapse and maybe someone competent will take a look at the problem and run the necessary wired backhauls.
Incidentally, you might want to sniff the traffic anyway and see if they're setup in the default setup for using 40 MHz of bandwidth on
2.4 GHz instead of the usual 20 MHz.
Also, try unplugging the closest repeater and see if the speed magically improves (assuming that the internet connection isn't constipated or undersized).
If it is just a repeater, why doesn't it have interference, or nulls and peaks?
I works fine most days, however, I tried several times to open your post, but had to disconnect from the marina network and borrow a signal from the friendly neighbors.
It just so happens, I am in control of one repeater. It is mounted on my boat and I supply the power to it. But if I power it down I cannot receive the signal from home base. In reality that might not be true, I am inside of an aluminum box, probably better sealed than Jim's garage. I do have an external 15db gain panel antenna mounted outside, if I turned it around, and used it, I should be able to receive home base. But, that's not a permanent solution, someone would complain if I shut off my repeater. OK, I turned my flat panel to point at the antenna on my boat, which is also inline with home base. Repeater ON----- Ping 35ms DL-0.91Mbps UL- 0.90Mbps Repeater OFF---- Ping 56ms DL-0.95Mbps UL-0.46Mbps
So, my main question which, I ask indirectly, why don't I have nulls and peaks and interference with several transmitters/repeaters. The answer is, I guess I do have nulls and peaks and interference.
I'm guessing that the broadcast is on more than one frequency? I'll look up 2x2 MIMO, that may have the answer I seek. Ah, yes.
I'll ask about the 20Mhz, 40Mhz bandwidth setting.
Any thing I can do to increase the slow speed? It makes videos, at times a pain to watch, some days better than others. Mikek
Because 802.11 protocols insure that only one transmitter is transmitting at a time. There's no other transmitter to interefere with, including other wi-fi systems in the area.
Define "fine" and "most days"? We may have different definitions if 1 mbit/sec download speed is what you can fine.
Huh? Point it at the nearest WIRED access point somewhere in the marina. Pointing it at your own antenna isn't going to do anything.
Basically, no change. I can't tell from here why?
Interference requires that two transmitters be belching RF at the same time. That doesn't happen (in theory). When it does happen, the system uses the CSMA/CA mechanism to back off the timing and try again, this time without collisions.
Peaks and nulls are a different sorty. You can have them in the form of "frequency selective fading" also known as nulls. 802.11g OFDM does a fair job of eliminating most of those by splitting the traffic into several sub-channels, each of which carries parts of the data. If one sub-channel it trashed by frequency selective fading, the other will not be trashed (because they're on a different frequency) and most of the data gets through. However, if your system reverts to
802.11b (a bad idea), then there is no OFDM and you're stuck with deep fades and big holes.
Might as well turn off 802.11b (useless) and MIMO (who needs
300Mbits/sec performance if you can't even get 1Mbit/sec). Just leave it running on 802.11g only with 20 MHz bandwidth.
Sure. Mostly, it consists of adding WIRED access points, removing repeaters, bandwidth management (throttling), identifying abusers, proper antennas, and verifying that the backhaul isn't overloaded, overused, or undersized. I would not be surprised if you find that one file sharing user is monopolizing the bulk of the available bandwidth.
Yeah, I know the feeling. The commercials make TV not worth watching.
in reading this thread - I think there are a couple of basic questions... You say there are 4 of these around the marina ? and I think you even said one of them was on your boat ? Was that one of the repeaters - or just one of the same type units ?
The question is - are they extending in a line from a single point via leapfrong ? ie ---> Base --> AP1 ---> AP2 --> AP3 ---> AP4 or - are they spread out for coverage - all hoping to repeat from the single base ?
If I can watch a youtube video and it doesn't stop to buffer that's fine. I also think curvacious women are fine. :-) Most days-- I haven't kept a log, but I probably give up on a video once every 4 days. 1Mbps is not good, I agree, at home I get about 9Mbps, although there was a time I was getting 15Mbps. That brings up something else, I address at the end.
That is what I compared, using my flat panel I pointed it at the repeater on my boat, and got the 35ms, 0.91 and 0.90. Then I shut off the repeater. The wired access point is 60 to 80 yards away directly behind the repeater on my boat.They are inline.
About 2 years ago a cable guy stopped and said my house was sing RF very loud. They went into the attic and installed 6 or 8 new connectors and left happy having quieted my house down considerably. A couple weeks ago, two cable guys showed up with handheld receivers and ask if I had a cable buried in my front yard, I didn't and they wondered up to the eve of the house and their receivers indicated more RF. They seemed to trace it up into the attic again. The didn't go into the attic. They said the cable company will be upgrading in a few months and they need to quiet everything down before the up grade. I don't know why their previous fix didn't last. They said my house should be rewired, and splitters would be put on the outside to make it easier in the future. There is a cost, but if I buy the Service Protection Plan for $3 per month, it would be free. So I signed up, better them in the attic during the summer than me. Now weeks later I get notice the $3 Service Protection Plan has increased to $5 and they have not made the appointment to rewire my home. Hrrmph. Mikek
Or you go satellite and tell the cable people to f*ck off. The only thing worse than the cable company is a pager company. (QRM).
Leakage shouldn't be an issue if they used that quad shielded cable especially for cable TV use. I presume a sniffer plane is due in the area and they are trying to clean up their act.
I don't know about a plane, but the first time they heard my house it was from a truck driving down a 4 lane road near my house. It would probably be better to get a life and stop watching TV, but that's not going to happen. Mikek
Satellite internet has some severe limitations in speed and download quotas. It works, but I consider it a last resort. The Ka band stuff seems to be the best (Viasat/WildBlue/Exede).
Yep. I used to be somewhat in the paging business. We were all evil.
All the outside plant is quad shielded in my area. The drop from the pole is double shielded. Still we have leaks. The problem is that trees and branches falling on the cables tend to pull the connectors apart. Once the shield is broken, it leaks, and stays that way until fixed. Another problem is really bad inside wiring using crimp instead of push on connectors. The leaks I've located were mostly in the homes or businesses.
This is at one of my customers. The old stuff (RG-59/u) leaks like a sieve. The new stuff doesn't.
Ok, you have a "mesh". In order for that to work, each "corner" needs to "see" the other corners and the main wired access point in order to play repeater. Since they are farthest away from each other, I suspect that may be a problem. At least you have the altitude. The setup on the Ubiquiti repeaters should show which devices they can hear and at what signal strength and SNR. You might want to login (or break in) and see if your mesh is actually working. If a repeater is isolated from the main wired access point, it will try to repeat through one of the other repeaters, thus forming an unintentional series chain, with the attendant loss in performance.
Incidentally, if the repeaters can all "see" the central wired access point, then there's an easy (not cheap) way to make this work. Change them to access points and use 5Ghz as a backhaul. A Ubiquiti Nanostation (Loco) M5 is about $70. One at each access point, and a central Ubiquiti Bullet M5 with an Omni antenna at the central wired access point. In effect, replace the CAT5 backhaul with a 5GHz wired backhaul. That should work for an added cost of about $600. However, if you're going to mount repeaters on top of sailboat masts, rolling with the waves, maintaining antenna alignment on 5GHz is going to be a major problem. They need to be in fixed locations (light pole, tower, building roof, etc).
There's another problem with putting repeaters at the corners. The useful antenna pattern covers maybe 90 to 120 degrees of horizontal coverage. RF sent to areas outside the harbor is wasted and some antenna gain is usually a good thing. An omnidirection pattern, as provided by the built in antennas of the UAP Outdoor+, is mostly wasted. You might be able to use a simple flat plate reflector behind the repeaters. Better yet, a "sector antenna" with a 90 or 120 degree horizontal beamwidth:
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.