You still haven't explained how you've managed to achieve 54mbps data throughput on a standard 802.11a or 802.11g link.
Or how a ping request manages to go over your sattelite link, and the reply is returned, all in 263ms.
-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: Do you have a Tax-Free Savings Account?
Announcing a Community Wi-Fi Mesh
- Thread starter Iliyan
- Start date
You still haven't explained how you've managed to achieve 54mbps data throughput on a standard 802.11a or 802.11g link.
Or how a ping request manages to go over your sattelite link, and the reply is returned, all in 263ms.
kickass72za
Well-Known Member
heh, dunno why you guys are wasting your time
Xarion its about time you stopped bitching and lets give this a chance.....
I will sponsor your R150 to join aswell.
Ilyian needs some support, hopefully he will let me our usernames on monday, I will swing past his offices.
Fine. Lets test it.
I'll be in JHB on Saturday 20th.
We must be able to:
1) Test your claims of 54mbps from one end of your "mesh" to the other.
2) Confirm that you do infact get 263ms pings.
3) You must make as much information available to us, to verify that the path of the internet connectivity is via a satellite service, NOT purchased from Telkom or Sentech.
4) Its all legal.
5) You have distribution licenses for all the multimedia in your "library"
6) You only utilize 2 channels for your entire mesh (as claimed in a previous post)
You have nothing to lose but your credibility, which is already looking quite sad anyway.
Come on. Prove us wrong.
I'll be in JHB on Saturday 20th.
We must be able to:
1) Test your claims of 54mbps from one end of your "mesh" to the other.
2) Confirm that you do infact get 263ms pings.
3) You must make as much information available to us, to verify that the path of the internet connectivity is via a satellite service, NOT purchased from Telkom or Sentech.
4) Its all legal.
5) You have distribution licenses for all the multimedia in your "library"
6) You only utilize 2 channels for your entire mesh (as claimed in a previous post)
You have nothing to lose but your credibility, which is already looking quite sad anyway.
Come on. Prove us wrong.
Iliyan
Active Member
Hmmm.... kickass72za... you sound very, very serious... Judging you by your early posts, I wrongfully thought you were cynical and poisonous – excuse the silly old-timer.
Why don’t you go and fill-in the join form on the www.wifi-mesh.co.za so we can get this anonymity out of the way and I (the friendly person I am) will make sure that the office processes you application ASAP.
Iliyan
Active Member
I want to tell you about one great feature of our mesh.
It is based on a relatively little known principle of multilateration (the cousin of trilateration) and is designed to deal with the so called noise (we call it interference).
The experts here already know this, but as it is not that obvious, let me tell you that our APs run Linux, 400 MHz CPU, have about 16MB free memory and most importantly, feature -118 db calibrated receivers with continuous FRQ tuning. Also, each of our APs keep records of its own as well as its neighbours, coordinates in a standard form, and with a resolution of zero point one arc second.
I told you before that we developed a mechanism to deal with the noise – I mean, identify the noise, measure its level, and based on that figure, selecting an alternative channel for our community use. What I did not tell you before is how we do that exactly – in techie talk.
Read on for the missing details…
So, one of our techies decided to write a little (but nevertheless very nifty) piece of code which would look for a time when some idle CPU cycles need to be kept busy and when that happens, the AP would then scan the channel, identify the interference, record it’s MAC address as an identifier, measure it’s signal strength as it is received, put all that data into a small packet and push that small packet back to us, the master ship. And it will do that for every source of interference within the 2.4 GHz – 2.483 GHz spectrum. If there is traffic, it will stay at the channel it is assigned at, and when there is no traffic, it will quickly go and scan all the channels within the spectrum.
Here is the most interesting part. You know that we intend to roll out hundreds of thousands of these APs (for the commercially open minded out there, worth hundreds of millions of rands), literary pepper the urban and semi-urban areas so to have uniform coverage, right?
Again, the same crazy eyed techie came up with the idea that if we run this “interference identification” feature, surely if we correlate the data received from three APs against same MAC address in reverse proportion to the signal strengths using the principle of trilateration we should be able to pin-point the exact location of the interference source to within zero point one arc second (for the not so smart, it is less than a meter). Then, the same techie goes on that after first determining its exact location, and as having the coordinates on record gives as the ability to calculate distance, matching that data against the received signal strength trough a calibrated receiver interface, we should be able to determine the EIRP of the transmitting source.
Now, you can imagine us standing there with our jaws dropped, but this techie does not stop – he continues with these wild plans of us running the EIRP data against the community mesh member data we have and monitoring three things (1) if the EIRP our member emit does fall within the legal limits so to allow us to warn him if he is to inadvertently breach that limit, (2) if the interference comes from external to us source (say WUG or WISP), again, measure and record it’s EIRP and location and perhaps pass that data onto someone who cares about it but does not have his/her own means of measuring EIRP and (3) offer future location-based services to our community members.
You can all imagine, after hearing these wild plans, there was this long silence in our R&D offices. In fact, we were so impressed with this small, insignificant feature of our mesh that we decided to go ahead and implement it. As you can imagine, the said techie was very pleased with himself, as he was given the opportunity to make some good, have some fun and get paid in the process.
Tell me honestly, does not this all sound super-exciting to you all?
It is based on a relatively little known principle of multilateration (the cousin of trilateration) and is designed to deal with the so called noise (we call it interference).
The experts here already know this, but as it is not that obvious, let me tell you that our APs run Linux, 400 MHz CPU, have about 16MB free memory and most importantly, feature -118 db calibrated receivers with continuous FRQ tuning. Also, each of our APs keep records of its own as well as its neighbours, coordinates in a standard form, and with a resolution of zero point one arc second.
I told you before that we developed a mechanism to deal with the noise – I mean, identify the noise, measure its level, and based on that figure, selecting an alternative channel for our community use. What I did not tell you before is how we do that exactly – in techie talk.
Read on for the missing details…
So, one of our techies decided to write a little (but nevertheless very nifty) piece of code which would look for a time when some idle CPU cycles need to be kept busy and when that happens, the AP would then scan the channel, identify the interference, record it’s MAC address as an identifier, measure it’s signal strength as it is received, put all that data into a small packet and push that small packet back to us, the master ship. And it will do that for every source of interference within the 2.4 GHz – 2.483 GHz spectrum. If there is traffic, it will stay at the channel it is assigned at, and when there is no traffic, it will quickly go and scan all the channels within the spectrum.
Here is the most interesting part. You know that we intend to roll out hundreds of thousands of these APs (for the commercially open minded out there, worth hundreds of millions of rands), literary pepper the urban and semi-urban areas so to have uniform coverage, right?
Again, the same crazy eyed techie came up with the idea that if we run this “interference identification” feature, surely if we correlate the data received from three APs against same MAC address in reverse proportion to the signal strengths using the principle of trilateration we should be able to pin-point the exact location of the interference source to within zero point one arc second (for the not so smart, it is less than a meter). Then, the same techie goes on that after first determining its exact location, and as having the coordinates on record gives as the ability to calculate distance, matching that data against the received signal strength trough a calibrated receiver interface, we should be able to determine the EIRP of the transmitting source.
Now, you can imagine us standing there with our jaws dropped, but this techie does not stop – he continues with these wild plans of us running the EIRP data against the community mesh member data we have and monitoring three things (1) if the EIRP our member emit does fall within the legal limits so to allow us to warn him if he is to inadvertently breach that limit, (2) if the interference comes from external to us source (say WUG or WISP), again, measure and record it’s EIRP and location and perhaps pass that data onto someone who cares about it but does not have his/her own means of measuring EIRP and (3) offer future location-based services to our community members.
You can all imagine, after hearing these wild plans, there was this long silence in our R&D offices. In fact, we were so impressed with this small, insignificant feature of our mesh that we decided to go ahead and implement it. As you can imagine, the said techie was very pleased with himself, as he was given the opportunity to make some good, have some fun and get paid in the process.
Tell me honestly, does not this all sound super-exciting to you all?
Last edited:
I will NEVER support any WISP let alone anyone claiming to be a community selling commercial services on ISM!
ja super exciting, going to reach a corner on my suburb imminently
/me yawns
sorry i'm not an expert in wi-fi...oops sorry cant use that term...um i mean wireless. does true mesh (hope i can use that term) use only 1 or multiple channels to work together?
when did you say roll out was in Durban?
encrypted
Senior Member
Router
Hi Iliyan,
I'll be in JHB from next month so I will purchase one of your routers, if the radios can deal with a signal at -118dbm thats a real bonus, I will naturally want all the GPL components on CD please (I'll throw in an extra R10 for the blank etc..).
I have a few questions tho, I don't like buying stuff without knowing what I am getting.
The CPU, 400MHZ, but what is it?
How much memory do they have, you say they run with 16mb free?
Storage.. is it CF, disk on chip or ? and how much storage?
Can I power it over POE? (do I have to hack my own or does it accept power through the ethernet port and does it support 802.13af?)
It stores its co'ords.. so there is a GPS attached right?
What chipset is on your radios?
Another question, if I am on the community as a non internet (my understanding is that is free?) are there any blocked ports or protocols.
Any more detail would also be nice, but not really important.
Thanks.
Hi Iliyan,
I'll be in JHB from next month so I will purchase one of your routers, if the radios can deal with a signal at -118dbm thats a real bonus, I will naturally want all the GPL components on CD please (I'll throw in an extra R10 for the blank etc..).
I have a few questions tho, I don't like buying stuff without knowing what I am getting.
The CPU, 400MHZ, but what is it?
How much memory do they have, you say they run with 16mb free?
Storage.. is it CF, disk on chip or ? and how much storage?
Can I power it over POE? (do I have to hack my own or does it accept power through the ethernet port and does it support 802.13af?)
It stores its co'ords.. so there is a GPS attached right?
What chipset is on your radios?
Another question, if I am on the community as a non internet (my understanding is that is free?) are there any blocked ports or protocols.
Any more detail would also be nice, but not really important.
Thanks.
TheRoDent
Cool Ideas Rep
Where can I signup for one of the -118dbm rx sens wifi devices today please? I'll kill anyone standing in my way...
Moederloos
Honorary Master
This also concerns me. Take this scenario:
Wifi Mesh has (I imagine) one central break out point for Internet.
It is conceivable (and VERY likely), that there may be hundreds of communities dotted around a metropolitan area (with nodes being 75m apart).
Now, I join up, buy some kit and scan - ah a node! I connect to the node, go off and buy my login codes, and then find out that my community is not joined in any way to the break out point.
I do not see a way past this - in fact, in order to work, every community would in some way need a connection to the break out point - but, with a node every 75m or so, one would need a very dense mesh in order for (say) Krugersrop to connect to your offices (Bryanston? Apologies if incorrect there - but the premise is the same).
It is of course likely that you have more than one break out point - although it is doubtful you will have one per community (after all, a "community" may be one lone user). However, should you indeed do this, will it be for the users cost (as with the normal kits)?
Thanks in advance for the time you may need to spend on a reply,
M
Iliyan
Active Member
After seeing such a good reaction to my previous notes, I decided to share with you all some technical details on the way our meshes deal with interference from strong good signals.
It is most ironic that the people who transmit strong (read way, way above the legal limits) signals are the same people who suffer most of interference and they are the same people who complain that the noise is getting on their way. I think that most of the pirates that operate on 2.4 GHz are ignorant to the effect of their actions as they are generally lacking sufficient skills to plan and implement efficient RF communication at this level and I am not talking about commercial skills here… Most of you’ve guessed this already - I am describing WUGs here, and the WISPs…
Let’s look at the formula WUGs usually adopt – 200-400 mw transmitter, low loss coaxial cable (ouch, so expensive), low loss coaxial connectors, high gain but low design quality uni-directional antennas. Using the above equipment, the pirates usually try to cover distances that in a clear day feature well over 150 db path loss. Asked what is their link budged, they usually get irritated as most of them do not know what link budget is or do not know how to calculate one.
So, here is one of the most important parameters in radio communication – signal to noise ratio (SNR) – and the pirates care very little about it. This parameter is extremely important as it plays the ultimate role in determining the in-band quality of the radio link which in turn determines the distance of the said link and eventually the data rate, if the comms was digital.
People for ages have managed to achieve good comms, even under the lowest SNR conditions. Yes, they used SSB modulation, used AF signal processing (sometimes plain filters) and even exploited certain tropospheric signal propagation condition, hoping to time their transmission with the low activity of the largest interference transmitter that there is out there – the Sun. But, their efforts were worth it – with typically SNR down to below 10db (sometimes even bellow ambient noise, negative SNR), they successfully communicated across huge distances.
So - when it comes to communications in the 2.4 GHz band - here we are faced with the seemingly simple question – what would be the formula to configure our links so to achieve (1) legal EIRP (so we don’t break the law), (2) highest link budget (so we can enjoy most reliable links) and (3) highest SNR (so we can enjoy highest data rates)? Ant the challenge lies at the dependencies between the last two points.
Let’s try to address all these points one at a time –
(1) That’s easy – when configuring all components, make sure that one stays under the EIRP limit; if you are planning to install uni-directional antenna, make sure that you reduce the TX power to levels that will allow you to achieve the prescribed EIRP. And don’t believe the urban legends – low EIRP does not necessarily mean short distance – what do you think the EIRP versus distance is to the space communications in the (FRQ similar) C-band?
(2) The link budged is the sum of all gains and losses in a link. Do some maths (I know you hate it) and see what is there to improve. Did you notice the largest contributor to the link budget? That’s right – it is the receiver gain. And as no one regulates the receiver sensitivity – the higher it is, better for your link budget. Also, did you notice that the receiver antenna does not necessarily need to coincide with the transmitter antenna – so go ahead, choose a receiving antenna with the highest possible gain.
(3) Here where the biggest problem lies – as you increase the receive path gain, you increase the level of the good signal, but you also end up increasing the noise - the SNR reduces. That’s not good. That’s what makes our pirate friends complain about noise and low data rates that they achieve after deploying all those powerful transmitters connected to the high gain antennas.
So, how one should solve that problem? I am going to leave the answer to the second instalment, but at the mean time, here is a nice suggestion to the bright and curious – go and read up on the radio comms with the Voyager space probe - check what the SNR was at the time of its last transmission, check how NASA managed to overcome the seemingly impossible way to solve the SNR problem…
I will give you a hint – some have eight or even 16 corners…
It is most ironic that the people who transmit strong (read way, way above the legal limits) signals are the same people who suffer most of interference and they are the same people who complain that the noise is getting on their way. I think that most of the pirates that operate on 2.4 GHz are ignorant to the effect of their actions as they are generally lacking sufficient skills to plan and implement efficient RF communication at this level and I am not talking about commercial skills here… Most of you’ve guessed this already - I am describing WUGs here, and the WISPs…
Let’s look at the formula WUGs usually adopt – 200-400 mw transmitter, low loss coaxial cable (ouch, so expensive), low loss coaxial connectors, high gain but low design quality uni-directional antennas. Using the above equipment, the pirates usually try to cover distances that in a clear day feature well over 150 db path loss. Asked what is their link budged, they usually get irritated as most of them do not know what link budget is or do not know how to calculate one.
So, here is one of the most important parameters in radio communication – signal to noise ratio (SNR) – and the pirates care very little about it. This parameter is extremely important as it plays the ultimate role in determining the in-band quality of the radio link which in turn determines the distance of the said link and eventually the data rate, if the comms was digital.
People for ages have managed to achieve good comms, even under the lowest SNR conditions. Yes, they used SSB modulation, used AF signal processing (sometimes plain filters) and even exploited certain tropospheric signal propagation condition, hoping to time their transmission with the low activity of the largest interference transmitter that there is out there – the Sun. But, their efforts were worth it – with typically SNR down to below 10db (sometimes even bellow ambient noise, negative SNR), they successfully communicated across huge distances.
So - when it comes to communications in the 2.4 GHz band - here we are faced with the seemingly simple question – what would be the formula to configure our links so to achieve (1) legal EIRP (so we don’t break the law), (2) highest link budget (so we can enjoy most reliable links) and (3) highest SNR (so we can enjoy highest data rates)? Ant the challenge lies at the dependencies between the last two points.
Let’s try to address all these points one at a time –
(1) That’s easy – when configuring all components, make sure that one stays under the EIRP limit; if you are planning to install uni-directional antenna, make sure that you reduce the TX power to levels that will allow you to achieve the prescribed EIRP. And don’t believe the urban legends – low EIRP does not necessarily mean short distance – what do you think the EIRP versus distance is to the space communications in the (FRQ similar) C-band?
(2) The link budged is the sum of all gains and losses in a link. Do some maths (I know you hate it) and see what is there to improve. Did you notice the largest contributor to the link budget? That’s right – it is the receiver gain. And as no one regulates the receiver sensitivity – the higher it is, better for your link budget. Also, did you notice that the receiver antenna does not necessarily need to coincide with the transmitter antenna – so go ahead, choose a receiving antenna with the highest possible gain.
(3) Here where the biggest problem lies – as you increase the receive path gain, you increase the level of the good signal, but you also end up increasing the noise - the SNR reduces. That’s not good. That’s what makes our pirate friends complain about noise and low data rates that they achieve after deploying all those powerful transmitters connected to the high gain antennas.
So, how one should solve that problem? I am going to leave the answer to the second instalment, but at the mean time, here is a nice suggestion to the bright and curious – go and read up on the radio comms with the Voyager space probe - check what the SNR was at the time of its last transmission, check how NASA managed to overcome the seemingly impossible way to solve the SNR problem…
I will give you a hint – some have eight or even 16 corners…
RichardP
Banned
As long as you believe what you are saying its fine. Everyone is entitled to believe the tooth fairy and unicorns.
You should stop getting your sales and technical details from chappie-wrappers, its not good for you.
You make ill-educated and wild allegations that are totally unsubstantiated and uncalled for.
200-400mW is a total LIE and YOU know it... you are now grasping at straws to re-gain some of your credibility. The technology might be fantastic, but the bottom line is "You are using ISM band for making money **FOR YOURSELF** , and screw anyone else who uses the same frequency band"
The topic of the argument started out with "Get your own frequency and leave ISM band alone". Maybe go ask a VC company to buy you a license or bribe someone to fastrack your application. (there might be a reason ICASA dont want to allocate 2.4Ghz to a commercial company)
You fail to realise that the WUGs are 100% free, with no commercial interest and NO huge funding from anyone. its 100% paid for out of the pockets of the enthusiasts, as that is their motivation.. .. connecting the South african people -- NOT selling movies and STB's --
You are quite right about SNR, signal to Noise ratio just in-case you dont know what the acronym means. NOISE is not controllable, and its one of the MAJOR factors in a wireless link. If you want to tell me that you are using beam-forming and active antennas for noise reduction, I would have to ask how that gets squeesed in to a rubber duck.
BTW, I do radio planning currently and have also used TEMS and such tools, I have written commerical programs (That you most probably USED once) to do radio coverage predictions.. . do you even know what modulation Wifi used ? ODFM/QPSK/QAM? these modulation schemes change the link budget margins. How many bits/hz, constellation pattern.
http://en.wikipedia.org/wiki/Image:NASA_ECC_Codes-imperfection.png
Your NASA link, did you also see the speeds at THE SNR? 10KByes/Sec - download the internet in 2 minutes. but I fail to see how you installing 802.11g kit and NASA's Radio room are related.
Last edited:
Biggest crap I have read in a while... Wish the guy actually knew something instead of just posting crap like this. I have been planning 6+ months for my network and securing funds. And what a battle.. And this guy comes along and claims crap like this?
He says yes, I would fly up and test with duffy these ridicules claims, and although I doubt these claims, if proven wrong, I would personaly make certain to follow the same network planning..
Do we look like idiots???? Please, go and waste somebody's else time!
He says yes, I would fly up and test with duffy these ridicules claims, and although I doubt these claims, if proven wrong, I would personaly make certain to follow the same network planning..
Do we look like idiots???? Please, go and waste somebody's else time!
Pitbull
Verboten
So anyone tested it yet ?
Moederloos
Honorary Master
I am willing to give it a go - just do not know how I will get to the break-out point(s) and if there are any nodes near me.
I am waiting for Iliyan to answer my questions in this regard (previous page) though.
I am waiting for Iliyan to answer my questions in this regard (previous page) though.