by Charles Brabham, N5PVL
The Amateur Multicast Protocol (AMP) was designed to allow a single AMP server to transfer information to an unlimited number of client ( recieving ) stations simultaneously.
No single station can provide coverage to the entire world though, or even the entire United States. So to obtain nationwide coverage, multicast stations at a variety of locations must work in concert as part of a national network.
In the following illustration, a theoretical coverage area for a single AMP server located in Oklahoma City, OK is shown.
Finally, this image shows national coverage achieved from six AMP servers located in strategic spots around the country, with a considerable amount of overlap in a number of areas.
These images are in no way intended to depict actual propagation or coverage areas, but serve to illustrate the concept of networking AMP stations in order to achieve nationwide coverage, with areas of overlapping coverage.
In the real world, HF propagation from a single location does not stay in a convenient blob-shaped area, 24 hours a day. Instead the area of coverage for a particular frequency moves, shifts, expands and contracts throughout the day, the month, and the year as the sunspot cycle progresses.
The following series of images show hourly HF propagation shifts for a hypothetical station located in San Francisco, CA on a single day, March 1st 2006.
These images were taken from the Australian Government IPS Radio and Space Services website at http://www.ips.gov.au/HF_Systems
By following the propagation shifts for the lime green color for example, you can follow the hourly coverage area for a AMP station transmitting on 10 MHz.
Note that at the date these charts were created, ( March 1st, 2006 ) 20 meters looks like a far better choice than 30 meters for overall coverage. As the sunspot cycle progresses though, this picture will change and 30 meters will outperform 20 meters by a wide margin at the bottom of the cycle.
It is for this reason that 30 meters may be the best overall location for a permanent AMP "watering hole", as locating it there will give us some measure of independence from the ups and downs of the sunspot cycle.
It may not be possible to achieve nationwide coverage with six AMP stations, but then again, it may not take very many more than six to provde coverage if they have more than one transmitter each, operating on different frequencies.
Alternatively (and most likely) we can achieve national coverage by having a larger number of stations, coordinated by frequency so as to keep overlap on any particular band at reasonable levels. This can most easily be done by recommending a standard distance between AMP stations operating on the same band.
A number of programs for PSK31 have included a capability to receive and decode ten or more PSK streams at one time. This works out because PSK31 signals are narrow and operators congregate in specific "watering hole" frequencies recommended on the different bands. On twenty meters for example, the watering hole frequency is 14.070 USB. It is not unusual at all to see ten or more PSK31 QSOs in progress there, all at the same time and well within the passband of a PSK program so they can all be decoded at once.
The advantages to adopting this system for HF Multicast use are obvious. - It allows amateurs to tune to a single frequency and recieve data from a number of HF Multicast stations at the same time. This acts like a speed multiplier from the recipient's point of view, significantly increasing the rate at which they recieve data from the Multicast network.
HF Multicast client software that will decode ten or more PSK streams at a time has not yet been written, but the technique is widespread enough to reasonably expect software with that capability to appear. So even though we can only recieve and decode one PSK Multicast stream at a time now, I believe there is good reason to establish and populate HF Multicast PSK watering holes now, in anticipation of multi-stream multicast client software.
In any case, establishing an HF Multicast watering hole for each band will make it easy to find Multicast data and prevent friction with amateurs operating other modes. These Multicast watering hole frequencies will be listed here at USPacket as they are developed. At this writing the activity is on 30 meters around 10.134 MHz LSB with occasional test transmissions on 14.073 MHz USB. Three PSK modes, PSK31, QPSK31 and PSK63 are currently being tested, but PSK63 gets the most air-time right now and is what you are most likely to encounter.