Wifi over 500 km? Impossible?
From What The Wiki?!
Contents |
Preliminaries
Oftelsat, MCCH: http://fundmcch.com.ec (trading as brothers) direct involved as partners together with the IICD (enabling ICT in developing countries) http://IICD.org, initiated the following pilot in Ecuador, Latin-America----- Please look at the end of the article for latest development! End of october.
note: The format we use is not 802.11/wifi. Its the IP traffic over wireless thats important to us. Thats what we intend to do. A wireless long distance gateway to the Internet (com. Karel Fassotte)
The Article
The wireless solutions we see today are a product of western industrial society. They were developed for fast mobile communication in the urban society. Wireless is nothing new. Radio recievers have existed for over 100 years now.
Available solutions are:
G3 mobile phone and Internet. G4, Wifi (802.11 a/b/g) is good for Line-Of-Sight, max. 50 km, with good engineering. Wimax (802.16 a) promises more, Near-Line-Of-Site, maximum 50 km. Packet switching (A)X-25 VHF and UHF networks. Satellite IP. Nearly everywhere.
But what is the MISSING LINK?
Beside satellite connectivity, there is no other solution for the rural, far off regions!
Oftelsat is a small group of radio and computer engineers, working in the south of this planet (Ecuador). Together with the IICD - International Institute for Communication and Development [http://:www.iicd.org] and MCCH [http://www.fundmcch.com.ec ] They're dedicated to finding connectivity solutions for social and economic development in rural environments.
Today Karel Fassotte explained, how he and others solved the issue to bring up a non-expensive voice, fax and data link in Ecuador rural territories using WI-NVIS (Near Vertical Incidence Skywave).
The used Frequency band is 2–10 MHz Shortwave (not microwave).
Radio waves propagate in a straight line of sight (LoS), like a laser. If the signal is blocked or severely attenuated, the link is broken. No connection.
shortwaves 2-30MHz are still very interesting!
Lower frequencies in the 2 to 30MHz. Have special properties.
The radio waves of shortwaves, may be reflected back to the earth at a high altitude, 200km layer, in the so called “IONOSPHERE�.
To travel a long distance, the signal must take off at a low angle from the antenna, -30 degrees or less. This is so it can travel the maximum distance before it first arrives at the Ionosphere. Long gaps before the signal returns to earth – the part in between this and the end of the ground wave is the so-called Skip (or Dead) Zone. The distances that can be reached are some 1,000km and more with Skip Zone gaps. Our goal is 0 to 500km, without gaps!
To travel a local - medium distance, the signal must take off at a HIGH angle from the antenna – typically 60 – 90 degrees! This returns from the Ionosphere at a similar angle, covering 0 – 500km. It thus fills in the Skip (or Dead) Zone – like taking a hose and spraying right up, the water will fall in a cetrain round area.
What is now most important?
-HIGH angle of radiation from antenna -Minimise ground wave, as it will interfere with the returning skywave -Most importantly, CHOOSE THE CORRECT FREQUENCY BAND – go too high in frequency and your signal will pass through straight into space!
Choosing the RIGHT frequency:
The Ionosphere – D, E, F1 & F2 layers
D and to a lesser extent, E layers attenuate and absorb signal Best returns from F2 layer At any one time we need to know the frequency of the F2 layer – The Critical Frequency or foF2 Optimum frequency for NVIS work around 10% below this
NVIS frequency and time:
In practice, highest NVIS frequency can reach 10 MHz band. Lowest can go down down to1.81 MHz band. ‘Higher’ frequency band during day, ‘Middle’ frequencies afternoon/evening, ‘Lower’ frequencies at night. Frequencies also affected by time of year and period of sunspot cycle. For best results, these three different frequency ‘bands’ required.
Also F2 is the critical frequency:
The Critical Frequency is the key to successful NVIS working. The Critical Frequency (or foF2) is the highest frequency at any one time that a signal transmitted vertically will be returned to earth. Anything above this passes into Space. As they are interested in vertical signals for NVIS, then the value of the Critical Frequency (foF2) at any one time is of great importance to them.
But how to find or estimate foF2?
BY PREDICTION:
Real-time web information from Ionosondes Websites offering Critical Frequency predictions: – RAL STIF, IPS Euromaps Software Propagation prediction tables or similar printed material Rule-of-thumb:- ‘higher’ band by day, ‘middle’ band afternoon/evening transition, ‘lower’ band nightime
The most likely bands:
In practice, 8 MHz (35m) usually ‘highest’ band 4 MHz (80m) next lowest 1.81 MHz (160m, ‘Topband’) the lowest 80m and 160m strongly affected during the day by absorption from the D-layer, plus noise at night and varying times of the year Transition frequency around 5 MHz!
e.g. a realtime ionogram:
MUF LUF MOTS:
Predicted signal levels:
The best usable frequencies:
Next Question: What kind of Antenna is here to use?
Need high angle (60-90°) radiation for NVIS Vertical no use – predominantly low angle Half wave dipole at 0.5 wavelength produces low angle radiation, BUT, if lowered to 0.25 wavelength or below, produces high angle radiation ! Not too low, though – some earth losses. A reflector wire or earth mat can reduce this
The antenna, the key to be successful:
The right angle and direction is very important! It always is, here it is a little bigger! Horizontal dipole radiates at low angle radiation, it must be half (0.5) a wavelength above ground. Lambda 40 = 20meters! In the case of the lower bands such as 80 and 160m, this would be pretty high and big!
Horizontal dipole at ½ Lambda altitude:
The vertical opening depends on antenna height:
- If the height of the dipole is lowered, the angle of radiation becomes higher and the low angle radiation starts to disappear - The optimum amount of high angle radiation is obtained at a quarter- (0.25) wavelength above ground - Going lower than 0.25 causes efficiency loss - In practice 0.25 – 0.15 wavelength heights used for NVIS
High angle radiation “Straight up- Straight down�
What about the antenna bandwidth?
The dipole is essentially a single band antenna There are also a couple of special higher-gain single band NVIS antennas – Dipole with reflector The Shirley The Jamaica
(thought, that the pics from each type is not needed here)
Multiband antennas:
As mentioned earlier, at least three different frequency bands are needed for successful 24 hr NVIS service and so multi or wideband antennas are used. Simple ones include long wire, inverted-L, Shallow (120°) Inverted-Vee Doublet with open feeder, full-wave low (0.15-0.25λ) horizontal loop (reflector could also be used below this) . Other multiband antennas can be used, key is the correct height of the antenna.
Not only the correct angle of transmission TOA but ALE:
ALE : Automatic Link Establishment
ALE scans and tests sets of frequencies – usually in several bands - for a particular path or net until it
finds a frequency that will support communications over the path.
Each radio in an ALE net constantly broadcasts a sounding signal and “listens� for other sounding signals
generated by other net members
Analysis of these signals by processing determines the best frequency for communication at the time and this
frequency is then selected automatically for operations
Here PC-ALE, a software program written by Charles Brain, is the solution (screenshot):
NVIS summary:
Covers 0 – 500Km using High-Angle (50-90°) Skywave Choice of Correct Frequency Band just below the Critical Frequency is most important. Antenna must be horizontal, not vertical. Antenna must be on a correct altitude, between 0.25 and 0.15 of a wavelength above ground (antenna aperture) An NVIS antenna has omnidirectional radiation Multiband antenna (at least three bands) needed for 24hr NVIS coverage
And the digital link???
First some more about mother nature. Shortwave propagation the gift of nature is not easily predictable! It depends on our sun, the solar activity. Particals and radiation hit the upper atmosphere of the earth, charging the gasatoms, giving it the property to reflect certain frequencies. The sun is very dynamic, but has some cycles in her life! 11 year shift, from higher activity to lower activity (right now 2005/2006)
Cycles Solar activity:
There is off cource a day and night cycle Also a seasonal cycle (angle of incedence)
Oftelsat start their pilot in Ecuador (see beginning of the article), and the most important specs were the following:
- Establish a digital link, if possible 24 hours a day.
- First digital voice and audio
- Data transfer, FTP like.
- Gateway to internet
- Remote -control of the transcievers and pc (desktop)
The used COTS-equipment was the following:
Transciever: IC706MKIIG, shortwave, VHF,UHF transciever. Modified to transmit and recieve from 1Mhz. - 470Mhz. Powersupply: a computer (switching) powersupply, 13,8volts, 20 amps max. modified to charge a car battery.
IC706KMIIG
... and more COTS-equipment:
A T2FD broadband antenna. 3-30Mhz. A duoband dipole, 4 and 7Mhz. A desktop P.C. Simple soundcards. A digital voice and data modem ARD9800, capable of 3kbs, using OFDM, multitone. A car with build on antennatuner longwire and whip antenna. Also equiped with an ARD9800 modem and IC-706
Longwire on the car with tuner (scheme):
And here a picture of the mobile unit from beside viewed:
And here (just for fun) a real picture from the mobile unit:
The used software was Digital Radio Mundial multimedia, digital waveform software. They decided not to transmit with a wider bandwith than the 3Khz. Maximum speed 5.6kbs. Q15X25 Packet software (linux) 15 sub-carrier X25. And they want the Q15X25 to run IP traffic and gateway this to Internet. Speed also 3kbs.
The link maintenance was realized with PC-ALE software, mil.std. 188-141 with kind permission of Charles Brain (the author). Because of lack of interest in the experimenters world, he did not have the linux sourcecode anymore. With this software, a multimedia Laptop and an inexpensive transceiver, it is possible to successfully enter an ALE network.
So what happended in real life with this stuff?
First tests, taken on 7 and 8juli 2005. They first established a digital voicelink with el Mirador, close to the town of Quinide in Esmeraldas, at some 100km distance After that they've established a digital link between Esmeraldas and Guyaquil. It works! They aimed at 24 hours service capacity. This was not imediately achieved. Only parts of the night en day. This was due to the wroung transmission angle TOA.. The antenna was to low.
'Digital voice
The digital voice unit worked fine. The quality cannot be compared with the analoge signal. What a difference! The mobile unit could recieve well but transmission quality was not enough. The whip antenna, 2meters long, did not a good job.
'In the last trails we did end of juli, we repaired the horizontal whip that was burned and could stay connected, mobile, in the whole of Ecuador, the south of colombia and the north of Peru, more than 1000km. We were driving our mobile from the city of Riobamba in the middle of the country to Quito, the capital. We connected to every station in the HC-Network (amateur radio emergency network)!!! The mobile unit'
The longwire antenna worked very fine. The car when parked could setup digital voice communications within 10 minutes. Recieving was possible, mobile and over hundreds of kilometers.(see above for update)
The digital data transmission
Non IP data were succesfully transmitted over 350 km. Speed: 3kbs troughput. DRM transmissions, voice and data were also succesfull 3.5kbs. Q15X25 packet at 3kbs, did not function, because of some bugs in the software. ALE did not work well, because of interfacing problems.
So what happened next
Now end of august 2005,we got the PC-ALE running and linking. The problem we had was related to a hardware problem in the interfacing. We resolved this and all is working perfect. The mobile unit is capable of participating mobile in the network
But in view with all this problems to solve in the future, they have still an link over 350 km up and running quite well.
THE FUTURE
It is surely not a mass solution for bigger cities etc. Their goal is to create a radiomodem that is capable of 64Kbs in a multislotted system, using more than 1 radiochannel at a time (10). The plan is to use a SDR (software defined radio) Time critical aplications are not always possible! Several digital voice channels are possible. DSP will be a natural choice for the radio. Easily reproduced and updated, with the same hardware. This development has to replace the Icom 706 A transparant IP gateway, can easily provide sufficient bandwith for smaller, rural, far off telecenters, These centers are well served with this wireless solution. Operational costs are low. Next phase will be an ip gateway to the intranet/internet. Planned for the end of 2005.
(Update Now august 2005, we are already implementing these services in the link).
Third phase is the implementation of a SDR multiple carier system and system integration under Linux. Planned for 2006.
CONCLUSIONS
NVIS, Near Vertical Incedent Skywave, automated shortwave IP data transmissions, are feasinable at costs less than satelite. Operation has always been dificult because of the many factors involved, maintaining a link. Now with possibilities of digital techniques, systems can be automated and transparent for the user, f.i. an ethernet gateway. Bandwith that may be obtained are some 1bs /hz, using more sofisticated modulation and demodulation (adaptive sequential coding) can optimize link efficiency. 3bs/hz should be obtainable.
Digital mobile NVIS, within a cell structure of more dan 2000km diameter, is available (with good antenna engineering).
What do OFTELSAT want???
- The job we are doing has to be done.
- We are looking for reenforcement!
- Involve more brains
- Keep it an open enviroment, GNU
- We are interested in integrated solutions for development and communication
If you're interested feel free to contact Karel as follows: We notice a lot of people reading the page, but only few people, participate. karel fassotte karel@oftelsat.com karel.fassotte@gmail.com
Latest results 22-10-05 We still have some problems with the link. Its not open 24 hours a day. We have the ALE working, but the antenna system is not optimal. The signals are recieved in Atlanta US well, but in Esmeraldas not always. The vertical antenna seems to be to low now, < 50 degrees. MCCH has asked us to implement 2 more stations in the province of Esmeraldas. Installation should be finished in december 2005. (karel fassotte)
Please see also the pages of our partners: [http://www.IICD.org ] IICD - The International Institute for Communication and Development (IICD) assists developing countries to realise locally owned sustainable development MCCH [http://www.fundmcch.com.ec ] organisation of cacao farmers in Ecuador,
(eof, and i need a new pair of fingers. pls excuse my bad english, and the sometimes bad build sentences =) ...)
olly
References
http://inventors.about.com/library/weekly/aa121599a.htm
Categories: Published | Wifi | 25 July 2008 | July 2008
