Revert to Section 5.4

An issue which may be of concern to system designers in implementing a modulation system for a DTTB service is the possibility of a sudden transition between "perfect service" and "no service" over a very small range of received signal variation. Furthermore this small variation might vary with time of day, propagation conditions, season of the year or other more difficult to predict factors such as aircraft or vehicle flutter or receiving antenna movement in the wind. There are a number of possible approaches to deal with this matter.

Most of the DTTB systems demonstrated so far use non-hierarchical modulation systems designed for fixed reception. They all have a sharp threshold effect in the fringe of the coverage area. From an information theory point of view, the DTTB channel differs from point-to-point communication in that the channel capacities varies with receiver location. The further away the receiver is from the transmitter, the lower the channel capacity. The design of a hierarchical system may improve service to the fringe area. For receivers closer to the transmitter, the channel capacities are not fully exploited in a non-hierarchical system. Hierarchical modulation systems are under study as one possible approach to this problem.

It is thought in some quarters that a multi-resolution coding system may be advantageous for DTTB in as much as it may be able to provide a DTTB performance which degrades gradually as received signal levels are reduced. While the aim is generally supported in principle, it has been argued that, with current source coding, a higher aggregate data rate is needed to achieve this added functionality and that this is undesirable because it increases receiver complexity and may require use of a higher spectral efficiency modulation system (which will have poorer noise performance). This topic is still being studied but here we wish to consider the topic from the viewpoint of its implications for selection of either a single or multi-carrier approach.

The main issue relates to channel data capacity.

For multi-carrier systems, a layered modulation system can be achieved by one or more of the following approaches:

• assigning groups of carriers to different coding layers so that the lower layer(s) have a greater level of error-correction than the upper layer(s);
• assigning groups of carriers to different coding layers and using more rugged modulation formats (e.g. QPSK) for carriers assigned to the lower layer(s) and less rugged codes
  (e.g. 64 QAM) for carriers assigned to the upper layer(s);
• multi-resolution coding where for the lower coding layers groups of states of the modulation constellation are considered as a single modulation state (for example four states of a 64 QAM multi-resolution modulator might be treated as a single state of a lower resolution 16 QAM decoder).

Other approaches to multi-layered modulation systems may also be possible.

For SCM systems using QAM, layered transmission can be achieved by using non-equally spaced constellation modulation and different channel coding.

In a single carrier VSB modulation system, layered modulation might be achieved, at some reduction in total data capacity, by transmitting a mixture of 4 VSB and 8 VSB symbols in a time division multiplex.

A further approach is to use channel repeaters to extend or fill in the coverage in areas where the "perfect service"/"no service" transition occurs. In a DTTB system it might be possible to add repeaters without requiring the use of new transmission frequencies. This is the Single Frequency Network (SFN) concept. In that case, the co-channel signal from the parent transmitter is processed as if it were co-channel interference. Provided the delay is within the system guard interval a seamless transition in coverage can be achieved.

Both SCM systems using adaptive equalizers and MCM systems using guard intervals could support SFNs.

In both cases, the practicality of implementing such SFNs will depend on the levels of wanted to unwanted multipath signals that the receiving equipment can cancel.

To summarize, SCM and MCM are two promising modulation techniques offering comparable performances on a Gaussian noise channel. The better peak-to-average ratio of SCM may reduce the required transmitter output back-off. Channel coding is used to reduce vulnerability to a wide range of impairments. MCM is less sensitive to variation in multipath delay (within the guard interval) and may be a better candidate for single frequency operation.

As discussed above, single-carrier and multi-carrier techniques under consideration in various countries provide a comparable performance in many areas and also some particular advantages and disadvantages. It may then be possible to use either of these techniques to create a common standard which will provide different data rates for the various bandwidths available.

Continue to Section 6.1

Return to DTTB Tutorial Table Of Contents

Return to Tutorial Index Page