IP camera over wireless

I have to add one IP camera to existing IP video recorder.

This camera should be on the same LAN as the recorder but the main thing is it has to be connected over wireless because its cca 100 meter away from recorder.

I was thinking on using nanostation m5 but it's too expensive. Is there some other solution ?

GM

Reply to
gm
Loading thread data ...

It might be worth measuring it to see. The length limit for twisted- pair Ethernet is supposed to be exactly 100 meters, so maybe a really long cable will do.

If you are trying to do Power over Ethernet, it might not work so well at 100 meters. You may need to use a PoE injector closer to the camera.

If it really is more than 100 meters, you could do something like this, assuming you have power available in the middle of the run somewhere:

camera, 60 meter cable, cheap 5-port switch, 60 meter cable, recorder

You could also use a 30 meter and 90 meter cable, or 40 and 80, or whatever, if a good location for the 5-port switch is not exactly in the middle.

Or, use a media converter at the camera and the IP video recorder to switch to fiber. You'll need power at the camera but you probably have that anyway.

Matt Roberds

Reply to
mroberds

It's not clear what combination of power and range you need, but there's lots of all-in-one WiFi videocams. Just put a WiFi hotspot in range of the camera, where it can be easily wired to your Ethernet.

Reply to
whit3rd

You can run ethernet farther than the official limit of 100 meters. My record is 250 meters over CAT5e but at only 10 mbits/sec HDX (half duplex). Depending on your unspecified choice of IP camera, that's kinda marginal. My guess(tm) is you can proably do the same distance with 100baseT but again HDX to avoid NEXT (near end crosstalk).

Since this is a design group, I was wondering if you were interested in designing a wireless bridge instead of purchasing one.

What's your budget? There are cheapo wireless bridges available, mostly for 2.4GHz. For example, any box that runs DD-WRT can play bridge: Ubiquiti Nanostation Loco M5 are good radios for the purpose, but do cost about $75/ea plus the PoE power injector and mounting contrivance:

Since you haven't bothered to supply any information as to what currently have to work with, and the local RF environment, I can't offer any useful suggestions. However, some common alternatives are:

MoCa (ethernet over Cable TV coax)

Fiber optic cable and fiberethernet media converters.

HomePlug (ethernet over power lines)

HomePNA (ethernet over POTS phone lines)

Good luck...

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

Yes, the 110m limit for most Ethernet standards is for the minimum quality cable, the maximum speed, and any choice of device supplier. With better quality cable, reduced speed, and matching suppliers at each end, you can do a great deal more than 110m.

Another possibility is to put a little PoE-powered switch in the middle as a repeater.

Reply to
David Brown

When I install a camera system, I usually try to convince the customer to let me run CAT5e for the cameras and computahs instead of doing wireless. However, because of the cost of installing the wiring, they usually opt for the cheaper wi-fi solution. No problem. I just install it with wireless, wait for the inevitable problems, and then install the necessary wiring. Most IP cameras come with both wi-fi and ethernet. Usually, the problem is interference from other such systems, or from the other IP cameras on the premises. Excessive bandwidth consumption by the other IP cameras will make things worse. Motion sensing and adjustable resolution are a big help, but a mess of cameras still generates quite a bit of wi-fi traffic. Trying to use a laptop, tablet, or wireless device in a typical 8 camera environment is futile. Extra credit for a remote monitor to suck additional bandwidth.

The OP seems limited to only one camera, so bandwidth consumption and interference might not be an issue. Still, it should be considered especially if it is possible to use ethernet wiring.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

There is no reason you cannot use 802.11n for those building mount cameras everyone uses - they don't eat that much.

802.11n. Pretty much only that. Not a, not b, not g....

I don't mean a GoPro. I mean a real one. I forget the maker, but they are a massive pain in the neck to configure and use. They're built to screw into a wall.

But lose line of sight, have cosite interference problems or just plain-old interference problems and you may not like it as much.

And it'll be non-cheap enough to make cables look attractive.

--
Les Cargill
Reply to
Les Cargill

I beg to differ. Perhaps you should test that assertion.

Find an 802.11n flavored wireless router and set it for 802.11n mode only. Your choice for 2.4 or 5GHz. Connect with a laptop or tablet that claims to support 802.11n. Stream something and start walking away. How far did you get before it quit? How far away would the camera be located? I just did this test with a Linksys EA2700 router and a customers Toshiba Chromebook 2 on 2.4GHz using iPerf3. I started dropping packets at about 20ft, became noticeably erratic at about 25ft and was dead at 35ft. Streaming slowed down as the range increased, mostly from retries and retransmissions. Much depends on the number and quality of the reflections necessary for 802.11n spatial diversity.

With wi-fi, you can always trade range for speed. Very roughly, you the distance doubles if you go 1/4th the speed. That's not really true because the protocols vary with the speed, which also has an effect, but it's a tolerable rule of thumb.

GoPro does it wrong. The GoPro camera is not a wireless client, but instead is a wireless access point. If you want to use more than one GoPro to stream video to a local wireless router, you have to go through some rather messy configuration. For GoPro, this makes sense so that you can connect directly to the camera with their wireless remote or iphone/android apps.

They're all a pain to configure. Best of the breed methinks is Axis.

That's much of what I was ranting about. Wi-Fi doesn't scale very well. The problem is really one of what I call "air time". The "air" is really a shared medium, where only one device can transmit at a time on a given channel. Stuff too many cameras and devices into a small area and everything slows to a crawl due to lack of "air time". (I could go on and on about this, including my requisite denunciation of wireless mesh networks, but not now). The problem with wireless cameras, wireless video streaming, and similar streaming devices is that they occupy too much "air time" for other users. One common service call is complains about erratic and slow wi-fi when the streaming stuff is running. It's usually wireless cameras or wireless media servers. Wiring with CAT5e usually solves the problem, unless it's interference from the neighbors wireless streamers.

Yep. Incidentally, my favorite wiring scheme is CAT5e or CAT6 everywhere, and use 4 wires out of the cable for DC power and video through a video balun. One CAT5e cable can handle 2 cameras, or for fairly short runs, 3 cameras.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

This is a lot more involved than a laptop with a wireless nubbin and a consumer AP. Line of sight is maintained, a clear band is used, high horsepower AP/clients....

--
Les Cargill
Reply to
Les Cargill

Well, yes. Wi-Fi is a mess and I'll plead guilt to using too many buzzwords and barely explaining what they imply.

What I'm objecting to is your recommendation "There is no reason you cannot use 802.11n for...". 802.11n is not going to help or do anything useful.

Having the capability to do really fast wireless connections using

802.11n doesn't make much sense when it will only work out to about 8 meters (25ft) at these speeds. Depending on how one reads the original question, he's either 100 meters away, or has a cable route of 100 meters with less distance. Line of sight, maybe. What will probably happen with 802.11n at 100 meters is that the wireless devices will associate (i.e. connect) at some lesser speed. It won't go down to 802.11 slowest speed (1 Mbit/sec) because the 802.11n spec bans these slower speeds if 802.11n is enabled. 802.11n only can go as slow as 6.5 Mbits/sec connection speed (using the standard 20 MHz bandwidth and 800 nsec guard interval): using one stream and BPSK, but at that point, one might was well be running 802.11g which also has a 6 Mbit/sec slowest speed. Note that my previous test obviously didn't slow down to 6.5 Mbit/sec or the range would have been somewhat longer. If you sniff the traffic or monitor the wireless access point, I can see what speed it's using. In real life situations, with real live interference, it's going to be slow. The slower it goes, the more "air time" is required for a given amount of data takes to be sent, and therefore the less "air time" will be available for other wireless devices.

So, what's the OP to do? Probably run 100 meters of CAT5e and be done with it.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

It's been a long day :), so I want to stress first and foremost that a cable is a much better idea. It just is.

We keep a cable reel handy so if the wireless goes meshugginah.

But!

But you should only need about 10-20-40 megabit for say, sixish cameras.

And I *saw* it work out to 100-200 yards. I did not do detailed analysis of the guardband after I saw it work.

I think I sorta drew you into thinking that 400 MBit figured into this somehow - oh no, not at all. It is that the better .n chipsets ( on many Wifi things mainly the bridge can be appallingly and unecessarily bad ) work better.

It's more about the underlying stuff more than the waveform.

Sure. You'll most likely get tennish, maybe 20 MBPs (in terms *of service type*, but not in terms of throughput), of which only a few to a few dozen are needed to run a few cameras.

Generally, it was better, but mysterious are the ways of wireless.

But you gotta have a whale of an install - an *expensive* install. And when it doesn't work, then what?

Very much so. Oh, and we were definitely using 40 MHz bandwidth, and had zero interference ( so far ).

No, because the modulation schemes behave differently in the face of TCP traffic. I was as surprised as anyone when this was discovered. g trainwrecked worse than n.

You'd think OFDM would be worse than the g scheme, but it's not. Plus, many g modems are 10 to 20 mBit internal bandwidth - including some expensive ones.

And this may have just been luck of the draw - there may be g radios that work great.

I really found that watching the interface stats worked better than sniffing. Pulling the interface speed/bytes via SNMP works too.

Indeed.

I couldn't agree more.

--
Les Cargill
Reply to
Les Cargill

On a sunny day (Wed, 29 Apr 2015 08:02:22 -0700) it happened Jeff Liebermann wrote in :

My main concern is hacking. Any WiFi is easily hacked (it happened to me), and if your camera has the gateway set and an email address you are in trouble. So every camera here is now wired. No RF exposure either.

Reply to
Jan Panteltje

DIY cantenna

--
umop apisdn
Reply to
Jasen Betts

Possibly it could run of camera power instead of POE, many 10/100 switches want 6V power, unlike cameras which want 12V, but you could use a DC-DC converter.

--
umop apisdn
Reply to
Jasen Betts

ElectronDepot 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.