GB2232854A

GB2232854A - Determining channel quality

Info

Publication number: GB2232854A
Application number: GB8912942A
Authority: GB
Inventors: Andrew Colin Baslington
Original assignee: GEC Marconi Ltd; Marconi Co Ltd
Current assignee: BAE Systems Electronics Ltd
Priority date: 1989-06-06
Filing date: 1989-06-06
Publication date: 1990-12-19
Also published as: GB8912942D0

Abstract

A method of assessing the quality of a data channel based on the processing of soft decision demodulation information. Using the method it is now possible to determine the Bit Error Rate (BER) of a data channel in a much shorter time than previously possible. Soft decisions within a known band of levels are measured and used instead of hard decisions. The sampling may therefore be performed over a shorter time period. The method may also be used to indicate signal to noise ratios or any other selected quality scale. <IMAGE>

Description

DETERMINING CHANNEL OUALITY This invention relates to a method of determining the quality of a data channel or waveguide. In particular the invention relates to a method of estimating the Bit Error Rate (BER) occurring in a data channel.

Conventional techniques of determining BER in digital information involve counting the number of errors in a number of bits. When the bit rate or the bit error rate are low a sampling period of several seconds may be necessary to obtain enough accurate information to indicate the quality of the data channel.

In certain circumstances this time period is unacceptably long, a typical example being in satellite communication when large blocks of data must be transmitted with a high bit integrity. Another example is in fibre optic communication links. In these and other data transmission systems, which require high bit integrity, or operate at low bit rates, it is important that the bit error rate (BER) is assessed rapidly so that any degradation is detected within as a short a time period as possible. This enables a control system to implement any required correcting action as quickly as possible.

According to the present invention, apparatus for determining the quality of a data channel comprises a sampler arranged to sample a demodulated signal and assign a discrete value, from a predetermined range, to one or more sampled values; means for determining the proportion of assigned values lying within a chosen band of values, within the predetermined range, and means for assessing the quality according to the proportion.

Preferably in one embodiment there are eight equispaced levels, four being representative of a digital '1' and the remaining four being representative of a digital '0', and the chosen band occupies the central four revels of the eight levels. The proportion of values in the chosen band may be determined by counters. The means for assessing channel quality may be a memory element such as a look-up table.

The invention may be used in a communication system, particularly a satellite communication system.

The present invention will now be described by way of example only and with reference to the following figures, in which: Figure 1 illustrates schematically a satellite communication system; Figure 2 illustrates a channel "Eye Diagram" of a digital signal with superimposed noise; Figure 3 is an illustrative table relating Signalto-Noise ratio with Bit Error Rate.

Referring firstly to Figure 1, this illustrates a block diagram of a communication demodulation system which has an antenna 1, an amplifier 2 and a demodulator 3. A digital signal received at 1 passes via the amplifier 2 to the demodulator 3. The demodulated signal is then sampled at a sampler 4. The information then passes to a processor 5.

Referring now to figure 2 the form of the information is visualised as an Eye diagram 6.

Successive possible values of either 1 or 0 are shown as a series of waves 8 (in dotted lines). The received information 9 in this example is 10110, which has arisen as a result of sampling at times T1, T2, T3, T4 and T5 and is shown in solid lines. Superimposed noise 10 may cause the effective signal to drift and to take values which deviate from the true value of a bit. On sampling, because of this noise effect, so-called Hard decisions (i.e. decisions which assign one of only two levels, 1 or 0 to each sample), may not always assign a true value to the data if the noise is so pronounced as to alter the value of a bit from a 1 to a 0 or vice versa.

An improved method of data sampling is by using 'Soft Decision Making'. This involves data sampling but assigns one of a number of levels to the data, instead of either a 1 or a 0. In this example there are eight levels; +7, +5, +3, +1, -1, -3, -5 and -7 and which equi-spaced on the vertical axis of the Eye diagram.

Values +1 to +7 correspond to a digital '1' and -1 to -7 to a digital '0'. The soft decisions which arose from sampling at times T1, T2, T3, T4 and T5 in this example are +7, -5, +7, +3 and -7 respectively.

By counting the number of decisions occurring at a particular soft decision level or within a predetermined band of soft decision levels it is possible to derive information as to the channel quality. if a large number of +1 or -1 are counted, this would indicate a poor signal quality since a large amount of noise must be present. On the other hand, if all soft decision samples are of +7 or -7 this would indicate good channel quality. It has been found that if the value of 43 is taken this is particularly representative of channel quality. It is possible, with knowledge of the type of noise within the system, to correlate the ratio of the number of samples occurring at level 3 to the total number of samples, with the channel quality.

Figure 3 indicates how the proportion or percentage of decisions made at level 3 may be related to the Bit Error Rate and hence to the signal to noise ratio. Thus if 10% of decisions fall within the level 3 band, the Signal-to-Noise ratio is considered to be 8.5 dB and the Bit Error Rate is 10 4 or 1 error in 10,000 bits. The data in the Figure relates to transmission noise only, for other types of noise, different values may apply.

The decisions at each level may conveniently be measured by one or more counters. Data shown in Figure 3 may conveniently be held in a look-up table or other memory element. By implementing the information derived from soft decisions it is possible to obtain an indication of general quality and to obtain such a measurement in less time than has previously been possible. Measures, such as an increase in transmission power, for example in the satellite communication system may then be implemented to maintain signal quality.

Claims

1. Apparatus for determining the quality of a data channel, comprising a sampler arranged to sample a demodulated signal and assign a discrete value, from a predetermined range, to one or more sampled values; means for determining the proportion of assigned values lying within a chosen band of values, within the predetermined range, and means for assessing the quality according to the proportion.

2. Apparatus as claimed in Claim 1 wherein there are eight equispaced levels, four being representative of a digital '1' and the remaining four being representative of a digital '0', and the chosen band occupies the central four levels of the eight levels.

3. Apparatus as claimed in Claim 1 or Claim 2 including a memory for relating the proportion of values to a channel quality parameter.

4. Apparatus as claimed in Claim 3 wherein the memory comprises a look-up table.

5. Apparatus as claimed in any one of the preceding claims, adapted for use with a signal having superimposed Gaussian Noise.

6. A communication system including apparatus for determining channel quality as claimed in any preceding claim.

7. A communication system as claimed in Claim 6 which is a satellite communication system.

8. Apparatus for determining channel quality substantially as hereinbefore described with reference to and as illustrated by the accompanying figures.