Considering Next-Generation Amateur Voice Systems

Considering Next-Generation Amateur Voice Systems
Considering Next-Generation
Amateur Voice Systems
Jon Bloom, KE32
ARRL Laboratory
Next-generation voice systems are a logical outgrowth of high-speed
Replacement of existing systems (repeaters), which is needed
to address the current spectrum-use problems, must be preceded by design of
appropriate multiple-access systems.
Some of the key issues to consider are
analog vs. digital modulation schemes and the types of multiple access
1. The role of computer networking in voice communications
It may at first seem odd to give a paper about next-generation voice sys terns at a
devoted to computer networking.
But this reflects a trend: computer
and more about applications.
networking is becoming less
That is, the
availability of high-speed digital networks evokes new classes of applications, ones
that have little or nothing to do with classical “computer networking.”
For this
conference to consider the subject of voice communications systems is simply a
recognition of that reality.
Most of the communication carried across existing amateur digital systems is text.
But most of the communication amateurs do is by means of speech. This is hardly surprising; most
human communication is by way of speech. So the predominant amateur communication systems in use today
are those that are designed for voice communications. On the VHF and higher frequencies, analog voice
repeaters are the principal means of communication. The underlying premise of this paper is that speech
will continue to be the means of choice for most amateur communication. The issue to be addressed, then,
is in what way the advent of high-speed digital capabilities can and will change the way in which voice
communication is accomplished.
2. Limitations of existing systems
Present VHF voice systems’ use analog FM single-channel-per-carrier (SCPC) transmissions. This is
a technology developed in the period following World War II. Except for implementation details (the use
of solid-state circuitry, for example), this technology has remained essentially unchanged since the 1960’s.
While the fact that a technology is more than twenty years old doesn’t count against it, a twenty-year-old
technology that hasn’t changed perceptibly in that period can be assumed to be mature: little further
improvement can be obtained by refining that technology. If technological improvements are needed in
amateur voice communications, it seems unlikely that analog FM SCPC systems will provide them. Certainly
the experience of other services, such as Land Mobile, which have strong economic incentives for the
improvement of the technology, indicate that a technological plateau has been reached.
’ This paper neglects HF voice circuits. While HF has a role in providing long distance point-to-point
circuits, it has little foreseeable role in a general-purpose voice system.
Are improvements in voice systems needed? Consider these attributes of existing repeater systems:
Poor frequency reuse of tloice channels. Although amateur frequency reuse has yet to make full use of
existing techniques (CTCSS, for example), applying such methods would came only an incremental
improvement. If a large number of (codeless?) licensees are to be accommodated in existing spectrum,
more spectrum efficiency must be obtained.2 Frequency reuse is a key to efficient use of the spectrum,
but existing wide-area repeaters that can tolerate no interfering signals at levels above the receiver noise
floor are near-worst-case systems with regard to frequency reuse.
Idle chnnnels.
In major urban areas of the US, at least one of the available amateur VHF bands, and
often two or three bands, are full of repeaters. That is, in the segments of the band reserved for
repeaters, every channel is occupied. But many of these repeaters sit idle most of the day, coming to
life only during “drive time.” While peak loading performance is an important measure of any
communications system, there are better ways to deal with it than having idle channels for most of the
Encronchment of pncket. To date, packet radio has made use of narrow-bandwidth systems. In areas
where voice repeaters “own” all of the existing repeater allocations, packet is relegated to use of simplex
techniques. At present no efficient simplex channel-access mechanisms are in widespread use. Because
of this, packet users are making use of more channels to achieve lower per-channel loading, or are using
repeater pairs for duplex packet systems. This, plus the push to higher speeds and concomitant higher
bandwidths, is beginning to cause friction between packet and voice users. Packet systems are evolving,
however, which provides an opportunity to address part of the problem in the design of new packet
But at present, voice systems are not evolving. To address the issues noted above, amateurs need to design
and implement new voice systems.
3. Amateur vs. cellular
Amateurs are not the only ones looking to the development of new voice communication systems.
Cellular phone services (and others) are evolving designs to meet their voice communication needs. Since
huge amounts of research and engineering are being devoted to these efforts, it might seem that amateur
system designers should wait and benefit from these efforts by applying the techniques developed for cellular
systems. But there are some striking differences between the amateur environment and the commercial
mobile communications environment. These differences may mean that designs optimal for commercial
service are anything but optimal for amateur service. Some of the key differences:
The commercial systems will handle primarily station-to-station communications. Little if any broadcast
or party-line facilities will be provided; there isn’t much need for it. But the amateurs do need such
services. The ability of amateurs to call CQ is important. Any type of communication that is point-tomultipoint in nature, or in which the initiating user may not know the identity of the station he is calling,
requires a class of service that commercial systems are not being designed to optimize, if indeed they
offer it at all.
’ The oft-mentioned “moving to higher bands” solution is illusory. Microwave bands hold great promise
for point-to-point circuits, but mobile operation becomes difficult at higher frequencies, and mobile
operation comprises a significant part of amateur voice use.
One of the factors driving the design of commercial systems is the need for security. Encryption of voice
signals is highly desirable in a commercial radio environment. Amateurs have an exactly opposite
requirement: amateur systems must be capable of being monitored. More, they should encourage
monitoring in order that self-policing be effective.
The economic base for the construction of amateur systems is significantly different from that for
commercial systems. While commercial operators (presumably) can afford to use a large number of
stations to achieve near-total coverage of a given area, this often is not the case for amateurs. In areas
of hilly terrain or urban, it is difficult to avoid regions of poor coverage, and amateurs are unlikely to
be able to resort to large numbers of cell sites to get around the problem.
While commercial systems are highly reliable, the need for amateur communications is greatest in
precisely those circumstances where commercial systems fail. In such circumstances, a tightly-coupled,
centrally controlled system is likely to be unworkable.
The economic factors and the reliability factors argue for wide-area coverage systems. Monitoring of amateur
communications will be easier in such an environment, as will point-to-multipoint applications. This is
philosophically at odds with the commercial cellular approach. How much of commercial practice will be
transferable to amateur systems is an open question, but it seems likely that the amateur architectures will
differ from the commercial in more than name.
4. Future Amateur Voice System Capabilities
Let us assume that the locations of stations won’t change significantly: repeaters will still be located at
heights that provide wide-area coverage, fixed amateur stations obviously won’t move, and mobiles will be...
mobile. Only the characteristics of the stations will change. How must the characteristics of these stations
change in order to solve the kinds of problems described above, and what useful additional services can be
included in next-generation systems?
To accommodate increased traffic, more channels need to be available? From an economic standpoint,
it makes sense to occupy more channels not by putting up more repeaters, but by making existing repeaters
service more channels (i.e., repeaters will become multiple-access systems). At the same time, a capability
for full-duplex voice communication is highly desirable. Finally, repeaters should be able to operate together
in a network (using the high-speed networks under development now) to increase communications range.
Fixed and mobile stations should include both half- and full-duplex capability. They should be able to
communicate via repeaters or directly, just as they can now. New applications such as store-and-forward
(voice-mail), “call waiting” etc. can be designed into the system. For mobiles, especially, integration of data
and voice capability into one unit would be useful.
3 In this context, “channel” does not (necessarily) refer to a contiguous block of spectrum that carries a
circuit signal. Rather, we use a broader definition of channel: a spectrum resource-a bandwidth for
some period of time-that carries a single circuit.
5. Digital vs. analog systems
One of the key questions in designing future voice systems is whether to use analog FM techniques or
digital techniques between the repeaters and the user stations. For a variety of reasons, it appears that digital
is the way to go. Some of the advantages of digital techniques are:
Error detection and correction capabikty. Digital systems can potentially detect errors in received data.
Error correction can in turn be implemented using FEC or ARQ techniques. This capability could be
most usefully exploited in adaptable systems, wherein degradation of the RF channel would cause the
use of error handling. (Error control would increase the needed bandwidth, so it probably would not
normally be used. Potentially, several steps of increasingly capable error correction could be available,
trading bandwidth for channel performance as the need arises.)
Data transparency.
In analog voice systems, any digital data must be converted to voice-bandwidth
analog signals. The resulting system, optimized for voice, does a poor job of carrying non-voice data.
An all-digital system doesn’t care whether the data being transmitted is voice, image or text. Only the
endpoints are concerned with the real-world representation of the data. This is a much more flexible
system. (An example: present repeater control systems use crude tone-signalling [DTMFJ techniques
to provide control capability. An all-digital system could send many times more control information in
a much shorter time.)
Spectnr m efficiency.
There are two trends of technology that hold great promise in the continued
narrowing of digital voice bandwidth requirements. The first of these is the work being done in speech
encoding (an example of one place where amateur systems can benefit from commercial technology).
Speech encoding at speeds of around 16 kbit/s is a part of current commercial development efforts, and
advanced vocoder designs are expected to produce acceptable speech quality at speeds as low as 4
kbit/s.[l][2] The second technological trend is that of digital signal processing (DSP). At present,
the transmission speeds of mobile systems are limited to a few hundred kbit/s or so, a speed limitation
that is carrier-frequency dependent.[3] The practical number of bits per baud is also limited to two
or three.[4j DSP promises improvements in adaptive equalization that will improve those numbers,
allowing us to squeeze more bits-per-second into a given bandwidth. (Note, though, that even presentday systems are perfectly workable, and can give bandwidths at least as narrow as present analog
Inter jerence tolerance. Although the FM capture effect provides a measure of tolerance to co-channel
interference in analog systems, it is likely that properly designed digital systems will be even more
tolerant, allowing greater frequency reuse. (Note that even the tolerance of FM analog is often not used
in amateur systems. Systems without CTCSS access cannot tolerate any signal that will open the
Time-division multiple access (TDM4). As we will see shortly, TDMA has significant architectural
advantages, and, while an analog TDMA system is possible, a digital TDMA system is far easier to
Secutity. A digital system can easily be designed to require an integral transmitter identification (call
sign) in the data stream. This would provide traceability and debugging advantages.
Newness of technology.
One of the under-used facets of the amateur service is that experimental
techniques can be tried out there without the need to produce a commercially viable system. Amateurs
can build operational experience with new systems and, in so doing, add to the technology.
Flaibility. Inherent in several of these points, but worth a mention of its own, is the ability of digital
systems to change with the technology. For example, a system designed now would probably use speech
coding rates on the order of 16 kbit/s. In a few years, when 8- or 4=kbit/s coding techniques are
accessible, existing digital systems could incorporate them without complete system redesign. In fact,
such systems could be phased into service along with 16.kbit/s systems transparently to the user.
6. System Architectures
The architectural challenge is to devise systems that ensure that each station transmits in a manner that
ensures reception by the intended recipient without unduly restricting the use of the system by other stations.
Inasmuch as multiple access is required (see section 4), a key design decision will be to determine the best
type of multiple access. Of the three classical techniques: frequency-division multiple access (FDMA), timedivision mdtiple access (TDMA), and code-division multiple access (CDMA, or spread spectrum), we can
select one or more or a combination of techniques.
FDMA is the system used in current cellular phone systems. In these systems, channels are segments
of the band that are wide enough to contain the I3l voice signal. They are no different from amateur
repeaters in this respect. But in cellular systems, channels are assigned on demand. Such a controlled
mechanism is a requirement in any useful multiple access system.4 Cellular systems simply assign a particular
frequency to the calling station for the duration of the call. This seems simple enough, and in an analog
world it’s by far the easiest approach. But it requires that the cell-site (or repeater) have multiple receivers
and transmitters, an expensive proposition for amateurs. If full-duplex communications are required in an
FDMA system, the fixed and mobile units must be capable of receiving and transmitting at the same time;
a duplexer is normally required for this. However, in a digital FDMA system the transmitted bit rates are
lower than in an equivalent TDMA system. This has advantages, not least of which is that bit periods can
be relatively long compared to the delay-spread times.[5]
In a TDMA system, each station takes turns sending a burst of data on a particular frequency. This has
the advantage of requiring only a single receiver and transmitter at the repeater, and allowing full-duplex
communications snns duplexer. Also, by not fixing the channel bandwidth TDh4A provides for more flexible
selection of data rates: if a higher data rate is needed, the bursts of data are simply made longer. This latter
characteristic may override other considerations.
CDMA, or spread spectrum has utterly different characteristics from either FDMA or TDMA. Spread
spectrum operates by spreading the signal out over a large band of frequencies. One method of spread
[email protected] hopping, simply rapidly changes the frequency of the transmitter. The associated receiver
“knows” the sequence of frequencies and follows along. Several such systems operating in the same band of
frequencies can operate with little interference. In direct sequence spread spectrum, a high-speed (quite a
bit higher than the data rate) pseudo-random bit stream is mixed with the signal to “spread” the transmitted
spectrum. The receiver mixes what it receives with the same bit stream, resulting in a copy of the transmitted
signal.’ Adding stations to the band in such a system simply makes the apparent noise level increase, so
interference degrades communication rather than interrupting it. CDMA has the advantage that the signal
can be spread beyond the coherence bandwidth. This reduces the susceptibility of the system to multipath
effects, which is particularly attractive for mobile applications. CDMA has great potential, if little operational
histo$ Amateur efforts have proved promising,[6] but it seems unlikely that CDMA will be the
4 For a demonstration of why control is a requirement, listen to any CSMA simplex packet channel.
145.01 MHz will do in most areas.
’ One of the technical challenges of spread spectrum is to keep the transmitter and receiver bit streams
6 The military has been using spread spectrum for years, but they aren’t too forthcoming about the
technical details.
technique of choice for next-generation amateur systems.7
7. Conclusion
Next-generation voice systems are a logical outgrowth of the push toward high-speed networks, so
designing systems now to accommodate voice data is prudent. Such systems can address some of the
spectrum-management issues facing amateurs today and can provide enhanced voice capabilities as well as
integrated voice and data. The technical challenges of designing and implementing such systems will open
up a world of interesting experimentation and building such as has not been seen in amateur radio in recent
times. Some aspects of the possible approaches have been discussed here, hopefully in a way that encourages
amateurs to begin serious consideration of next-generation voice systems.
Chien, E. S. K., Goodman, D. J. and Russell Sr., J. E., “Cellular Access Digital Network (CADN):
Wireless Access to Networks of the Future,” ZEEE Communications Magazine, June 1987, p 22
Jayant, N. S., “High Quality Coding of Telephone Speech and Wideband Audio,” IEEE International
Conference on Communications ICC ‘90, IEEE, 1990, p 322.1.1
Siew, C. H. and Goodman, D. J., “Packet Data Transmission Over Mobile Radio Channels,” IEEE
Transactions on Vehicular Technolog)? May 1989, p 95
Calhoun, George, Di’taZ Cellular Radio, Norwood MA: Artech House, 1988, p 336
Calhoun, p 366
Kesteloot, Andre, N4ICK, “A Practical Direct-Sequence Spread-Spectrum UHF Link,” QST, May
1989, p 14
’ The author is willing to be convinced otherwise.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF