Post by thread:[EE] Radio modules

>-----Original Message----- >From: .....piclist-bouncesKILLspam@spam@mit.edu [piclist-bouncesKILLspammit.edu] >On Behalf Of David VanHorn >Sent: 30 July 2007 16:49 >To: Microcontroller discussion list - Public. >Subject: Re: [EE] Radio modules > > >> > I need >10mbps. Bluetooth, Zigbee, and wifi are not suitable. >> The current Bluetooth and Zigbee standard are much slower than >> 10Mbps... > >Hence the disqualification. >Wifi is performing rather dismally at the moment, with an >802.11G link giving about 3mbps of actual throughput. Is this possibly down to signal strength and/or interference issues? 802.11G should give significantly better throughput than that. Would 802.11N be a possibility? What are the range requirements? Is it line of sight? Regards Mike ======================================================================= This e-mail is intended for the person it is addressed to only. The information contained in it may be confidential and/or protected by law. If you are not the intended recipient of this message, you must not make any use of this information, or copy or show it to any person. Please contact us immediately to tell us that you have received this e-mail, and return the original to us. Any use, forwarding, printing or copying of this message is strictly prohibited. No part of this message can be considered a request for goods or services. =======================================================================

On 7/30/07, David VanHorn <.....microbrixKILLspam.....gmail.com> wrote: > > > I need >10mbps. Bluetooth, Zigbee, and wifi are not suitable. > > The current Bluetooth and Zigbee standard are much slower than 10Mbps... > > Hence the disqualification. > Wifi is performing rather dismally at the moment, with an 802.11G link > giving about 3mbps of actual throughput. > > I need to push four cameras worth of data at 640x480, plus a few > kbytes/sec of other sensor data. What codec is the video data in? I'm wondering if you can get more compression on the pre-transmit side, ie an MPEG4 codec? 802.11n might be another option. Alex

Michael Rigby-Jones wrote: {Quote hidden}>> -----Original Message----- >> From: EraseMEpiclist-bouncesspam_OUTTakeThisOuTmit.edu [piclist-bouncesspam_OUTmit.edu] >> On Behalf Of David VanHorn >> Sent: 30 July 2007 16:49 >> To: Microcontroller discussion list - Public. >> Subject: Re: [EE] Radio modules >> >> >> >>>> I need >10mbps. Bluetooth, Zigbee, and wifi are not suitable. >>>> >>> The current Bluetooth and Zigbee standard are much slower than >>> 10Mbps... >>> >> Hence the disqualification. >> Wifi is performing rather dismally at the moment, with an >> 802.11G link giving about 3mbps of actual throughput. >> > > Is this possibly down to signal strength and/or interference issues? 802.11G should give significantly better throughput than that. Would 802.11N be a possibility? > > What are the range requirements? Is it line of sight? > > Regards > > Mike > > While RF is not my main strength, I have a close friend who is an expert, so I learned a few things by osmosis. It seems that the biggest problem with wireless is that the speed can drop precipitously with additional users. In general, everybody knows this, but in practice, strange things happen. I was planning a design in which police cars would dump their video surveillance data by wireless. Police cars were parked in a protected building with a high ceiling. Tests indicated that it would work fine- in general- but overlooked was the fact that all policemen change shifts at the SAME time- so 50+ users would be dumping at the same time. The idea wasn't usable because the upload speed with 50 users was slower than molasses, and the cars had to be gassed up, oil checked as well as other routine things attended to between shifts . We had to use CAT5E and have multiple servers online to get this to work. The wear on the CAT5 jumpers was awful; at least one was replaced daily. The final design was that the sargeant's cars (5-6) were kept as wireless, all the lower-level police used CAT5 jumpers. The servers simply stored data temporarily, since it took hours to properly archive that much data. Computers on the network would slowly copy in the stored data, then process it. Active transactions (tickets, arrests) were clipped out and attached to the paper files of the case. It was fun. --Bob A / /

4 cams aggregated? what about independent streams? Check out WiLife see what they are doing? David VanHorn <@spam@microbrixKILLspamgmail.com> wrote: > > I need >10mbps. Bluetooth, Zigbee, and wifi are not suitable. > The current Bluetooth and Zigbee standard are much slower than 10Mbps... Hence the disqualification. Wifi is performing rather dismally at the moment, with an 802.11G link giving about 3mbps of actual throughput. I need to push four cameras worth of data at 640x480, plus a few kbytes/sec of other sensor data.

> Is this possibly down to signal strength and/or interference issues? Test distance was about 3 meters. Interference is a possibility, but given the band, it's just a fact of life. > 802.11G should give significantly better throughput than that. I would have expected so. I'm not sure where the bottleneck is, but connecting the systems by wire is MUCH faster, and taking the laptop through an external wifi router is also significantly faster. > Would 802.11N be a possibility? Not likely, I haven't seen any small N adaptors, and I'd have to deal with kernel issues. The main system is a PC-104 running Ubuntu. > What are the range requirements? Is it line of sight? Umm.. right. :) We have reasonable expectations of range, and there will be impairments.

On 7/30/07, alan smith <KILLspammicro_eng2KILLspamyahoo.com> wrote: > 4 cams aggregated? what about independent streams? Physically impractical. I barely have room for the one Wi-Fi board now. > Check out WiLife see what they are doing? Probably compression, which isn't acceptable. Image quality is important. We also discussed using four analog transmitters, but finding those with good picture quality isn't very easy either. We have four cameras that give good images (ptgrey fire-eye) and we really don't want to throw away the quality in a crummy radio. In the end, the video needs to end up on a laptop where we can do stitching, filtering, edge/motion detection, etc, so lossy compression isn't an option, and lossless seems to end up taking more processor bandwidth than we have.

On Mon, 2007-07-30 at 11:48 -0400, David VanHorn wrote: > > > I need >10mbps. Bluetooth, Zigbee, and wifi are not suitable. > > The current Bluetooth and Zigbee standard are much slower than 10Mbps... > > Hence the disqualification. > Wifi is performing rather dismally at the moment, with an 802.11G link > giving about 3mbps of actual throughput. Then I'd say there's something wrong with how you're doing your WiFi. My G links go at around 22Mbps. 3Mbps means you're either too far from the AP (in which case getting directional antennas would help), or you're using a transport that can't handle the speed reliably (a misconfigured TCP stack can easily limit you to 3Mbps). TTYL > I need to push four cameras worth of data at 640x480, plus a few > kbytes/sec of other sensor data. Does it have to be digital? Analog video transmitters/receivers are a dime a dozen these days, getting one that allows you to set what channel they transmit on shouldn't be too hard. TTYL

On Mon, 2007-07-30 at 13:31 -0400, David VanHorn wrote: > > Is this possibly down to signal strength and/or interference issues? > > Test distance was about 3 meters. > Interference is a possibility, but given the band, it's just a fact of life. > > > 802.11G should give significantly better throughput than that. > > I would have expected so. I'm not sure where the bottleneck is, but > connecting the systems by wire is MUCH faster, and taking the laptop > through an external wifi router is also significantly faster. 3m? There was something wrong with either you hardware or drivers. I regularly get 22Mbps at distances MUCH more then 3m on my G links. Another possibility, if interference is really a problem is try some 802.11a hardware, very little is on that band. TTYL

>-----Original Message----- >From: spamBeGonepiclist-bouncesspamBeGonemit.edu [TakeThisOuTpiclist-bouncesEraseMEspam_OUTmit.edu] >On Behalf Of David VanHorn >Sent: 30 July 2007 18:31 >To: Microcontroller discussion list - Public. >Subject: Re: [EE] Radio modules > > >> What are the range requirements? Is it line of sight? > >Umm.. right. :) We have reasonable expectations of range, and >there will be impairments. Yeah, I know that sounded a bit silly as you are talking about cameras :D I was actualy thinking it could be an aerial appllication? Cheers Mike ======================================================================= This e-mail is intended for the person it is addressed to only. The information contained in it may be confidential and/or protected by law. If you are not the intended recipient of this message, you must not make any use of this information, or copy or show it to any person. Please contact us immediately to tell us that you have received this e-mail, and return the original to us. Any use, forwarding, printing or copying of this message is strictly prohibited. No part of this message can be considered a request for goods or services. =======================================================================

> Then I'd say there's something wrong with how you're doing your WiFi. My > G links go at around 22Mbps. 3Mbps means you're either too far from the > AP (in which case getting directional antennas would help), or you're > using a transport that can't handle the speed reliably (a misconfigured > TCP stack can easily limit you to 3Mbps). TTYL Could be.. I know I've linked Motorola 850 APs with DD-WRT and had transfer rates up around 6M, probably more but that's as fast as matters, since that's the cap on my comcast link at home. We're only working at a few meters range at the moment, signal strength isn't an issue. > Does it have to be digital? Analog video transmitters/receivers are a > dime a dozen these days, getting one that allows you to set what channel > they transmit on shouldn't be too hard. TTYL In the end, it's all analog :) No, we're not married to digital, but the analog transmitters that I've seen are all excrement. Low bandwidth, not true VSB, basically just cheap AM or FM modulation with no thought to what the passband looks like. In the end, I need to get the video into a laptop for processing, so it has to go digital somewhere. We intend to do some processing before it's transmitted, so that would push the transition point up twoard the cameras.

> I was planning a design in which police cars would dump their video > surveillance data by wireless. Police cars were > parked in a protected building with a high ceiling. Tests indicated that > it would work fine- in general- but overlooked > was the fact that all policemen change shifts at the SAME time- so 50+ > users would be dumping at the same time. > The idea wasn't usable because the upload speed with 50 users was slower > than molasses, and the cars had to be > gassed up, oil checked as well as other routine things attended to > between shifts . We had to use CAT5E > and have multiple servers online to get this to work. The wear on the > CAT5 jumpers was awful; at least one was replaced > daily. > > Did you consider using something a bit tougher than a RJ connector for the download cable?

David VanHorn wrote: >> Then I'd say there's something wrong with how you're doing your WiFi. My >> G links go at around 22Mbps. 3Mbps means you're either too far from the >> AP (in which case getting directional antennas would help), or you're >> using a transport that can't handle the speed reliably (a misconfigured >> TCP stack can easily limit you to 3Mbps). TTYL >> > > Could be.. I know I've linked Motorola 850 APs with DD-WRT and had > transfer rates up around 6M, probably more but that's as fast as > matters, since that's the cap on my comcast link at home. > > We're only working at a few meters range at the moment, signal > strength isn't an issue. Just a wild thought, you aren't too close to the AP are you? Just wondering if your seeing near field effects or some such. Other option is to take a look at a MIMO AP that might bump your signal up without needing to change anything inside your device.

On 7/31/07, Bob Axtell <RemoveMEengineerTakeThisOuTcotse.net> wrote: > I was planning a design in which police cars would dump their video > surveillance data by wireless. Police cars were > parked in a protected building with a high ceiling. Tests indicated that > it would work fine- in general- but overlooked > was the fact that all policemen change shifts at the SAME time- so 50+ > users would be dumping at the same time. > The idea wasn't usable because the upload speed with 50 users was slower > than molasses, I read of a similar problem in a school/university situation where everyone was supposed to download wirelessly the lesson notes at the beginning. it was only a few megs or so, but multiplied by 50 or 100 users... They ended up putting multiple 802.11a access points in the ceiling in a hexagonal/cellular arrangement. 802.11g only has 3 non overlapping channels. 802.11a has a heap more. John