Tuesday, August 27, 2013

Microwave data application in Dubus Magazine

Dubus is a quarterly magazine for technology, construction, DX and propagation on VHF/UHF and Microwave bands. It is edited in English and German languages by an international group of hams, printed in Germany.

I have blogged a few times before, on how the amateur microwave bands are so seldom used, and are our greatest allocation, as well as potential area for new development.

The problem is for the most part, the microwave bands sit idle, except for a few contests a year. Dubus is mostly what you'd expect, hams building microwave stations for the quest of a contact record.

I have never been that type, to each their own.

I am more interested in building infrastructure or something a number of hams can use. Which is why I like data networks, as I feel it can support a variety of applications while keeping some constant use of the frequencies.

Dubus 4/2006 - Microwave Japan Seiji Fukushima, PH.D, JH6RTO

High-speed IP connection on 5760 MHz by Hiroyuki, JH1UGF and Tsugno, JN1AYV

We report a 63 km high-speed IP connection on 5760MHz using ICOM ID-1's and modified transverters, UTV5600BIIP. The throughput was measured as about 100 kb/s at the application of layer .

Here are the details.

On the 29th of October, 2006, Hiroyuki, JH1UGF moved to Mt. Toogasa (altitude about 900 metres) and Tsugno, JN1AYV stayed at home in Chagasaki, Kanagawa for this experiment. The distance was 63km and both of us were within the line-of-sight. We used the same systems as follows. Our 1295MHz, digital transceiver was ICOM's ID-1 . To up/down-convert frequency, the 5760MHz transverters UTV5600BIIP were employed but some circuits were modified for this experiment. Since fast T/R switching is quite important, we removed the coaxial relay and added a circulator instead, as shown in Fig.1. In addition, two anti-parallel diodes were Inserted at the RX front end in order not to damage the preamplifier. These diodes cannot be seen In Fig, 1 unfortunately . The driver amplifier is MGF0904A and the final amplifier is 2xMGF1302, which outputs about 1W. We used a homebrew 25cm dish antenna with a horn radiator. Its gain was 25dBi.

We started our experiment. The voice QSO on FM was quite successful, RS 59+, as the distance was short enough . Then, we switched to direct digital mode for IP connection but the result was poor . Using a ping command of MS-DOS, Its connection rate was lower than 10%. Considering the multipath effect, we changed the frequency and/or antenna direction, but in vain. Lastly, Tsuguo moved his entire system , which was initially set in his room, to a balcony. Tsuguo had thought the experiment would have been successful even through a glass window since they had been within the line-of-sight. We got 100% ping connection after this trial.

Then we tried an application experiment. Using IP Messenger (IPmsg) software, Hiroyuki loaded an Tsuguo's image file into the Hiroyuki's PC. Look at our evidence of Fig. 2. The front is Hiroyuki and the back is Tusguo in the inset photograph. The 90KB file transfer took 6 or 7 seconds and, then, its application-layer throughput was calculated to be approximately 100kb/s . We learned that any thin obstacle in the path might make the multi-path interference at high-speed data communication. We would like to make the distance longer and to make the throughput higher in future.

What I did pay attention to in this article was that for the fast T/R switching they used circulators. I am not sure why this never occurred to me. I noticed this concept in 2006 Microwave Technology Letter, titled "Switchless Bidirectional Amplifier For Wireless Communication Systems."

So far this is the only microwave data application I have read in Dubus. I find that a bit odd, as there is a huge microwave data radio network in Germany and Europe.

If anyone else knows of good reading in these areas, I'd appreciate a note.

No comments: