![]() |
![]() Join NBEMSham
for news on updates to this software suite Requests for assistance and bug reports should be made to the above group You must become a member before posting messages |
| fldigi-3.11.5-win | Programs XP and W2K, includes fldigi-3.10, flarq-4.1 and the correct cygwin1.dll |
| fldigi-3.11.5-winV | Programs for VISTA, includes fldigi-3.10, flarq-4.1 and the correct cygwin1.dll |
Presumably you require a mode that can provide a printed copy of messages sent and received. Otherwise you would probably be using some sort of voice communications rather than NBEMS. NBEMS is the combined use of Flarq, Fldigi, a Single Sideband Transceiver and the appropriate antenna system.
When it is essential that your message be received without any errors NBEMS can provide that facility. Flarq takes the file to be sent and cuts it up into little pieces or blocks. It transmits each block along with a unique check sum for the block. It actually sends out a number of blocks and their corresponding checksums in sequence and then waits for the receiving station to report back on any bad blocks from the sequence that need repeating. The receiving station copies the block and computes it's own check sum for the block. If the two checksums agree, the receiving station knows it has copied the block correctly. If the two checksums disagree, the receiving station knows there are one or more errors in the block and requests that it be resent. Flarq allows you to decide what block size you want to use for file transfers. If conditions are poor, you'd be wise to use a relatively small block size since the presence of only one error in the block will result in the entire block having to be resent. In fact if conditions are really bad, and you decide to use a large block size, you are almost guaranteeing that any block sent will contain one or more errors and have to be resent...and resent... and resent. You get the picture. If conditions are good you can use a relatively large block size since the likelihood of a large block being copied without errors improves as conditions improve. If conditions are good and you decide to use a small block size anyway, then you are wasting a certain amount of time spent in going back and forth when there is little need to do so since the frequency of errors is small. The point of all this is to get across the notion that your choice of block size needs to be based on conditions. If things are absolutely marginal, go with a block size of 16. If conditions are near-perfect, go with a block size of 256. Block size can have a dramatic effect upon throughput. This effect can be of the same order of magnitude as the actual speed with which the characters are actually transmitted. The default block size that is recommended for NBEMS applications is 32. If conditions are poor, drop the block size down to 16. If conditions are good, kick the block size up to 64. As you become more familiar with Fldigi and Flarq and observe file transfers in progress you will become more adept at deciding whether or not you should increase/decrease block size or leave it alone. If you see that lots of blocks are being repeated, consider reducing the block size for your next file transfer under the same conditions. If you see that few blocks are being repeated, consider increasing the block size on your next file transfer. Pay attention to what's going on when file transfers are underway, act accordingly, and you'll maximize your throughput.
While
Flarq has the capability of transferring image files without
errors, such files tend to be relatively large and can take an
inordinate
amount of time to transfer. The image files must first be
converted to an ASCII text format using base-64 encoding. This
usually increases the file size by a factor of 3/2. When
errors in the transferred image can be tolerated, the transfer
time can
be drastically reduced by using the MFSK-pic mode available in fldigi.
The larger the image is (the more pixels) the
longer it takes to send. Images (even those initially in
color)
can be transmitted in one third as much time by transferring them as a
grey scale image. A 200 by 200 pixel image can be transmitted in
40 seconds in grey scale and 120 seconds in color. That same
color image
would take 2 hours and 58 minutes to transfer using ARQ and MFSK-16 for
the modem. On VHF with a good signal path the MFSK-pic image
transfer will result in a very satisfactory result at the receiving
station.
Trying to decide what band and mode to use can be a pretty complex business depending upon such things as the presence or absence of static, the distance over which you wish to communicate, your available resources, the time of day, and the relative importance of error free communications. Other factors such as the deployment of stations you wish to communicate with and their resources also will affect your decisions.
In emcomm applications distances over which communications are required will seldom exceed 200-300 miles. Normally this distance is sufficient to communicate outside the affected area unless you are involved in an extremely widespread disaster. Communications within and just beyond most disaster areas should use VHF/ssb for distances less than 100 miles and HF/ssb for distances up to 300 miles.
A lot of tactical communications take place on VHF FM at modest power levels using vertically polarized antennas at modest heights. Such communications are most often voice communications over relatively short distances ( 1 to 20 miles) and are limited by terrain and line-of-sight considerations. VHF FM communications typically will involve communications among portable/mobile stations and between such stations and one or more fixed base stations. VHF FM is not good for ARQ message transmissions due its reduced range, vertical polarization, and need for receiver capture before signal decoding can occur.
VHF communications requiring longer range, error free file transfers, should use horizontally polarized high gain antennas and Single Sideband transmissions. This is the best and most dependable combination to use for VHF NBEMS.
Whenever possible, use 2 meter, SSB with digital modulation, and horizontally-polarized antennas, for greatest range and dependability. If drift is not a problem, start with PSK63, and switch to DominoEx16, PSK125, or PSK250, depending upon the needed S/N over the chosen path, but switch to DominoEx modes if drift is a problem. MFSK16 be used for VHF unless drift is a problem. It decodes weak signals very well. It can transfer pictures with or without ARQ. If signals are well above the noise or fading, MFSK32 or MFSK64 can be used to speed up picture transfers. If drift is ever a problem, then DominoEX16 can used instead of MFSK16 but note that DominoEX has no picture mode.
HF communications over distances of 100 to 300 miles are ideally suited to use Near Vertical Incidence Skywave (NVIS) propagation. NVIS propagation is less influenced by terrain features than VHF propagation. Mountains or large hills in between two stations employing NVIS propagation do not pose a problem. NVIS is also used to describe a communications strategy. The idea of this strategy is to maximize one's signal within the desired communications zone and to minimize it outside of this zone. At the same time, an NVIS strategy tries to minimize the reception of signals and atmospheric noise from outside the communications zone. In order to do this all stations within the communications zone can take a number of steps to benefit from the use of an NVIS strategy. Stations should make use of relatively low antennas that direct most of their radiation upwards (nearly vertical). A half wave dipole at about one eighth of a wavelength above ground is an easy to deploy NVIS antenna. A quarter wave vertical or a very high horizontal dipole normally is NOT good for NBEMS usage. DX is NOT the object here. In fact, DX is undesirable within an NVIS context. Stations within the communications zone should use relatively low power (no more than 100 watts) in order to minimize the effects of multi-path propagation. Multi-path propagation occurs when the receiving station gets essentially the same information arriving at slightly different points in time because the information has traveled over two or more paths of different length. Under good NVIS conditions it is possible for two stations to communicate very well with power as low as 5 or 10 watts on 40 and 80 meters. In addition to reducing multi-path effects, low power helps to conserve batteries which is often a factor to be considered in emergency communications. NBEMS has been tested repeatedly over a 200 mile path on 80 meters at night during the height of the summer storm season in South Florida ( thunderstorm capital of the USA) and has worked extremely well running 50 watts or less at both ends of the circuit. NVIS antennas were used for these tests.
The frequency below which NVIS propagation can take place varies throughout the day and seasons. The Australian government maintains a very useful website where real-time information regarding the maximum useable frequency for NVIS propagation for North America is available.
http://www.ips.gov.au/HF_Systems/4/3
Another site that provides real-time information on the maximum useable NVIS frequency is
http://www.spacew.com/www/fof2.html
Both sites can be studied from time to time in order to get an idea of how the maximum NVIS frequency varies over the course of time in your area. The maximum useable NVIS frequency becomes less at the higher latitudes. In British Columbia, for example, the maximum NVIS frequency at any one point in time may be 4 Mhz while at the same time it may be 7 Mhz in Texas. At this particular time 80 meters might work for NVIS in British Columbia when 40 would not. At the same time in Texas both 80 and 40 meters might be available for NVIS communications. Clearly referring to a website over the internet for up-to-the-minute NVIS data might not be an available option in an emergency situation. For this reason it is important for emcomm operators who expect to be using HF communications to get some experience on bands that might be useful for NVIS applications...typically 80 and 40 meters. The simple rule is 40 meters during the day and 80 meters at night. Things get more complicated in the vicinity of sunrise and sunset. There is no simple way to predict what the best time is to switch from 40 to 80 (something you would normally do in the late afternoon or from 80 to 40 (something you would normally do in the morning.) You really have to listen to what is going on to make these kinds of decisions. It also helps a good deal if you get some hands-on experience on 40 and 80 so that you become familiar with how these bands typically behave over time. If you are on 40 meters in the afternoon and your NVIS circuits start getting flaky, it's time to consider moving to 80. Certainly when you start hearing foreign broadcast stations and DX on 40 you KNOW it's time to head for 80. The decision is usually a little easier in the morning. 80 meters will usually be good for a few hours after sunrise before absorption makes it increasingly difficult. When things do get flaky on 80 in the morning, it's time to switch to 40. Obviously if you've switched from one band to another and things have gotten worse, go back to the band where things were better! Sometimes you just have to make do with the best thing that is available - even if it isn't that good. To simplify matters, and for greatest dependability, use VHF instead of HF whenever possible.
MFSK16 and MFSK32 are the recommended modes for use on HF. DominoEX, PSK, and Thor and MFSK can all be used in conjunction with Flarq to effect error-free file transfers, but only MFSK is recommended for use on 40 and 80 meters. The MFSK modes are relatively robust modes and work better than many other modes where signals are buried in the noise and static - conditions that frequently occur on 40 and 80 meters, particularly in the lower latitudes and during the summer and fall seasons. MFKS32 is a faster modes, somewhat less robust than MFSK16, and can be used when signals levels are high with little static discharge noise. If MFSK16 fails to prove robust enough for conditions then the mode recommended for NBEMS use is MFSK8. This mode is too slow to realistically be used for file transfers but has proven to be more robust than MFSK16. It occupies slightly more bandwidth and runs at a half the speed of MFSK16 .
While the MFSK modes are the recommended modes for HF they do require accuracy in tuning and frequency stability. If frequency stability and tuning present problems then DominoEX and/or Thor modes may be used as alternatives to the MFSK modes. Thor is a Forward Error Correcting mode with Incremental Multiple Shift Keyed tones. It is similar to DominoEX, but adds the robustness of FEC as found in MFSK. THOR can be used to advantage on both HF and VHF. The fldigi implementation of THOR (and DominoEX) does not require accurate tuning of the received signal. The receiver audio frequency tracking can be anywhere within +/- 100 Hz of the transmitted signal and decoding will not be adversely effected. No AFC is required for this mode and has been disabled in fldigi for this mode. Visual, audible and written descriptions of THOR and other modes can be observed here: Sights and Sounds. This synopsis points out the strengths and weaknesses of the various digital modes available in fldigi. MFSK16 enjoys a 2 dB advantage over Thor16 and therefore will give better performance when signal conditions are marginal. It is also twice as fast as as Thor16. But the equipment on both ends of the communications path must be very stable and tuned to within 3 Hz of the actual received signal in order to use any of the MFSK modes. Please note that there is NO single mode that is best for all propagation conditions.