Doc. #23-A
1. Introduction.
The Digital Voice Working Group held a teleconference on July 7, 2002 to discuss:
2. Dayton Digital Voice Forum.
The forum was well attended for a Sunday morning at Dayton. It reinforced the idea that many hams are anxious to try digital voice. Some excellent presentations gave details of the state of the art and of future possibilities. Among those in attendance were TTF member Paul Rinaldo, W4RI; TAPR's Brad Noblet, WA8WDQ and Steve Bible, N7HPR; and AOR's Yoshi Nishimura and his associate. The speakers were Doug Smith, KF6DX; Gary Barbour, AC4DL; Charles Brain, G4GUO and Andy Talbot, G4JNT; Harold Reasoner, K5SXK; and Cédric Demeure of Thales Communications.
N7HPR was kind enough to record the audio from the forum. The presentation slides and audio are posted on the TAPR Web site at www.tapr.org/tapr/dv/index.html, as are those of the WA8WDQ presentation during the TAPR forum. A link to that URL was placed on the ARRL TIS digital voice page.
The Group feels the forum was successful in publicizing digital voice. Pre-forum announcements ran in the W5YI Report and on many popular Web sites, including those of the RSGB in the UK and WIA in Australia.
3. Alinco Product Review-- Feedback.
The Group was pleased to see that its recommendations about the use of the mean-opinion score (MOS) system were adopted by the TTF and used by the QST product-review staff. The Alinco DJ-596 review revealed useful information about the performance of the digital voice system (CVSD) and modem (GMSK) it uses. Samples of digital voice as transmitted and received were placed on the ARRL Web site so that readers could hear what it sounded like.
Based on the average MOS received by the Alinco system (2.5, or significantly less than toll quality), the Group cannot recommend it for general amateur use. We reserve a higher level of recognition for the perceptual voice codecs mentioned below, all of which receive scores significantly above 3.0.
Significant feedback was received on the two QST articles about digital voice in the January and February 2002 issues. Many readers queried us about the TAPR digital voice kit and were steered to the TAPR Web site. Visits to the TAPR digital voice page now total in the thousands. Some correspondence was from folks using the Kachina remote control system and others told of their current APCO25 operations on the ham bands. We now know of at least four APCO25-compatible repeater systems in the U.S. on Amateur Radio.
4. TAPR Digital Voice Kit.
This kit was not very successful because Motorola discontinued the MC56002EVM on which it was based around the time the kit reached beta test. In addition, the AMBE-1000 digital voice chip it uses is scheduled for obsolescence. You will read below about work to implement the software modem on another platform and about work to redesign a kit around the newer AMBE-2020.
5. High-Speed Software Modems for Digital Voice.
Group member Gary Barbour, AC4DL, reports success in operating a 3600 baud modem on an Analog Devices AD2181 development kit. His software code exactly emulates the system originally developed on the MC56002EVM by Group member Charles Brain, G4GUO.
The Group views this as the attainment of a significant milestone. Any modem sustaining up to 2400 bits/s of error-corrected data throughput on HF is a very significant development for Amateur Radio. Gary is about ready to have a printed circuit board made for the modem. Bench tests are under way and field tests are in the planning stages.
6. Progress With Other Digital Voice Systems.
Final integration of the Thales digital audio system is now taking place at the Thales labs in Paris, France. They are using a Ten-Tec Jupiter and a Kenwood TS-430S for initial Amateur Radio tests. Mr. Demeure reports that preliminary field trials can begin within 30 days.
The Thales system is a subset of the ITU standard for digital audio broadcasting that was adopted last year. It occupies 3 kHz of bandwidth and uses an orthogonal frequency-division multiplexing (OFDM) arrangement that yields remarkably good quality on difficult HF paths. Listeners give it an average MOS of 4.25 over the full range of propagation conditions.
One very attractive feature is that the system runs on a fast PC using a standard sound card. The software has many advanced properties, such as "real-time" impulse-response and spectral displays. Should Thales reach agreement to distribute the software, it could be a boon to the popularity of digital voice over Amateur Radio.
AOR Japan announced at Dayton their Model 9800 digital voice/image/data modem. They report it is currently working only in digital voice mode but they expect to release it by the end of 2002. It uses the new AMBE-2020 voice codec and a 36-tone OFDM modem. NTSC standard video is supported and adaptive JPEG compression is employed. Reed-Solomon coding is combined with ARQ/FEC for error-detection and correction. The Group has undertaken work to see what must be done for compatibility between the AOR system and the G4GUO/TAPR system. The systems are not that different.
Group member John Gibbs, KC7YXD, informs us that ICOM is proceeding with its DSTAR system. Along with some other technologies mentioned here, DSTAR was covered in QST earlier this year. John has agreed to demonstrate the system at the Digital Communications Conference in Denver later this year. DSTAR features digital repeaters and gateways to TCP/IP networks and the Internet.
Dennis Silage, K3DS of Temple University reports that, with some of his graduate students, he is undertaking to upgrade the G4GUO/TAPR system to the new AMBE-2020 voice codec. The project is evidently just getting under way and the Group looks forward to being able to report on its progress soon.
7. Support for Digital Voice Standards.
The Group sees most of the digital voice technologies being tried on Amateur Radio today as viable. The two that have emerged preeminently are the G4GUO/TAPR system and the APCO25 system, by virtue of their actual use. We expect to see the Thales system gain acceptance because it is a subset of an international standard, because of its ease of use and its performance. The AOR and K3DS systems are expected to build on the work of G4GUO while maintaining compatibility with his original system. DSTAR seems to have a lot going for it and while it may not be compatible with any of the others, it should get an equal chance in the marketplace.
We continue to embrace the idea that there is room for more than one digital voice system in Amateur Radio right now. Rather than dictate a single standard, we choose to allow things to evolve as experimenters do their bit. Users will ultimately decide what is best.
8. Recommendations and Thanks.
Having no further recommendations at this time, the Group reiterates some of its previous recommendations:
A. Direct the ARRL lab to evaluate the APCO 25 radios that they have had.
B. Inquire of the IARU about the uniformity of rules and regulations pertaining to digital voice in various countries. Also ask about DAB and what impact it may have on Amateur Radio. The Working Group does not have the resources for this.
The Group thanks President Haynie, 1st Vice President Harrison, the TTF and the ARRL Board of directors for establishing Software-Defined Radio and High-Speed Multimedia Working Groups. We have already had fruitful discussions with both. We wish to especially thank Joel Harrision, W5ZN, for his tireless work heading the TTF over the past few years.
Respectfully submitted,
ARRL Digital Voice Working Group:
Gary Barbour, AC4DL
George Bednekoff, AC5WO
Charles Brain, G4GUO
John Gibbs, KC7YXD
Jesse Morris, KC5GTK
Doug Smith, KF6DX, Chair
Mike Tracy, KC1SX, HQ Liaison