Method and Circuit for Detecting CN Ratio of QPSK Signal

1. Field of the Invention

The present invention relates to a method and a circuit for detecting a carrier-to-noise ratio (CN ratio), and more particularly to a CN ratio detecting circuit used for adjusting a satellite broadcast receiver at an installation of an antenna and also for checking receiving conditions of satellite broadcast signals.

2. Description of the Related Art

FIG. 1 shows a block diagram of a prior art CN detecting circuit for a satellite broadcast signal. In the figure, variable resistor 1 is directed to adjusting the input level of a frequency-demodulated signal FD of a received satellite broadcast signal. Wide-band amplifier 2 has a pass-band width wider than 10 MHz. The noise component included in the output of wide-band amplifier 2 passes through band-pass filter (BPF) 15 with a center frequency of approximately 9 MHz, and is amplified by noise amplifier 16. The gain of noise amplifier 16 is adjusted in advance with variable resistor 17 so that the value of the CN ratio to be determined has a predetermined relation to a detected noise level. The output level of noise amplifier 16 is next detected through smoothing or averaging by noise detector 18, and then converted to a digital noise signal by analog-to-digital (A-D) converter 19. A-D converter 19 is typically a 4 bit converter capable of representing the converted digital signal in 16 steps. Microcomputer 20 calculates the CN ratio in accordance with a predetermined program from the digital noise signal and delivers the calculated result to a given display device.

Wide-band amplifier 2 also provides an output OF, to be delivered to a transmission line. The output of wide-band amplifier 2 is converted so that the output impedance matches with the 75 .OMEGA. characteristic impedance of the transmission line by means of impedance converting circuit 21 which transmits the impedance-matched output to the transmission line as the frequency-demodulated output OF.

The problem encountered in the CN ratio detecting circuit described above is that, because of dispersions in the gain and of the signal-to-noise ratio of both the noise amplifier and the noise detector, and also because of errors caused by temperature variations and non-linear distortions in the circuit characteristics, it is difficult to adjust the gain of the noise amplifier so as to cause the detected noise level to have a predetermined correlation with the CN ratio of the frequency-demodulated signal. This involves a difficulty in obtaining a precise CN ratio through analog processing of the received signal, which results in a difficulty in adjusting and checking the satellite broadcast receiving system.

Another problem encountered is that using expensive parts such as an A-D converter, in the analog CN-ratio detecting circuit brings about an increase in the cost of the circuit.

It is an object of the present invention to provide a method of detecting the CN ratio of a QPSK signal, which is less influenced by dispersions in gains and signal-to-noise ratios of the circuit elements and by temperature variations and non-linear distortions in the circuit characteristics.

It is another object of the present invention to provide a circuit for detecting the CN ratio of a QPSK signal, which is less expensive and less influenced by dispersions in gains and signal-to-noise ratios of the circuit elements and by temperature variations and nonlinear distortions in the circuit characteristics.

In order to attain the first object above, the method according to the present invention is directed to determining the CN ratio of a QPSK signal which carries or includes a binary signal representing at least a message and an information for error-control to detect a double-error. The method comprises the steps of:

receiving the QPSK signal and reproducing the binary signal from the QPSK signal,

detecting the double-error included in the binary signal and generating a pulse whenever the double-error is detected,

counting the pulse and providing an occurrence number of the double-error during a predetermined time period which is long enough to compare the occurrence number with an average occurrence time during which successive double-errors occur, and

computing the CN ratio from the occurrence number in accordance with the theoretical relation between CN ratios and occurrence numbers of the double-errors.

The theoretical relation is expressed by the equation

where: .alpha. and M denote the CN ratio and the occurrence number of the double-error, respectively, k being a proportional constant, and the function .PHI. being defined so that ##EQU1## where: .zeta. denotes a variable integration having a dimension of .alpha..sup.-1/2.

In order to attain the second object above, the circuit according to the present invention is directed to determining the CN ratio of a QPSK signal which includes a binary signal representing at least a message and an information for error-control to detect a double-error. The circuit comprises:

converter means which receives said QPSK signal and reproduces said binary signal from said QPSK signal,

error-control means which at least detects said double-error included in said binary signal, and delivers a pulse whenever said double-error is detected,

counter means which counts said pulse and provides an occurrence number of said double-error for a predetermined time period which is long enough to compare with an average occurrence time during which successive doubleerrors occur, and

computer means which computes the CN ratio from the occurrence number in accordance with the theoretical relation between CN ratios and occurrence numbers of the double-errors, as described above.

The above and other objects, features, and advantages of the present invention will become apparent from the following description referring to the accompanying drawings which illustrate an example of a preferred embodiment of the present invention.

FIG. 1 shows a block diagram of a prior art CN-ratio detecting circuit for a satellite broadcast signal.

FIG. 2 shows a block diagram illustrating an embodiment of the CN-ratio detecting circuit according to the present invention.

FIG. 3 is a diagram for explaining a mathematical formulation of the double-error rate.

FIG. 4 shows a curve of double-error rates plotted against CN ratios.

Referring to the drawings, the embodiment of the present invention will be presented below.

FIG. 2 is a block diagram illustrating an embodiment of the CN-ratio detecting circuit according to the present invention. In the figure, variable resistor 1, wide-band amplifier 2 and impedance converter 21 are identical with those illustrated with reference to FIG. 1. Band pass filter (BPF) 3 with a pass-band width of 5.7272 MHz is adapted for passing a QPSK voice sub-carrier. QPSK demodulator 4 synchronous-demodulates the QPSK voice subcarrier to provide coherent component signal I and othogonal component signal Q. Difference converter 5 regenerates a PCM signal, including a voice PCM signal, from two component signals I, Q. Buffer amplifier 6, fed with the PCM signal from difference converter 5, delivers the PCM signal as a bit stream BST. Frame-synchronization detector (FSD) 7 is supplied with the PCM signal from difference converter 5 and detects the frame-synchronization signal from the supplied PCM signal, delivering synchronization-detection pulse Syn whenever the frame-synchronization signal is detected. Reference timing generator 8 generates reference timing signals (CKa, CKb, CKc, CKd) in synchronization with synchronization-detection pulse Syn.

Descrambler 9 is supplied with both the PCM signal from difference converter 5 and reference timing signal CKa from reference timing generator 8, and descrambles the PCM signal in synchronization with reference timing signal CKa. Deinterleaver 10 effects a deinterleaving of the output signal of descrambler 9 in synchronization with reference timing signal CKb, thereby reproducing the voice PCM signal.

Error detecting circuit 11 detects a double-error in the output signal of deinterleaver 10 in synchronization with reference timing signal CKc and provides an occurrence pulse E whenever a double-error is detected. Counter 12 counts occurrence pulses E during the period designated by reference timing signal CKd to provide an occurrence number M. The period is long enough to compare the occurrence number with the average occurrence time that two successive double-errors occur. Microcomputer 13 calculates the CN ratio from the occurrence number in accordance with the theoretical relation between the CN ratio and the occurrence number, which will be described in detail below.

In operation, broadcast radio waves is transmitted from a broadcasting satellite at frequencies ranging from 11.7 GHz to 12 GHz in the SHF band. The frequencies of radio waves received by a receiving antenna are down-converted into a frequency range from 1.0 GHz to 1.3 GHz by a first converter. The down-converted signal is then supplied to a satellite broadcast receiver in which its frequency is converted into an intermediate frequency by a second converter. The intermediate-frequency signal is then demodulated by a frequency demodulator to produce a frequency-demodulated signal FD.

Frequency-demodulated signal FD is applied through variable resistor 1 to wide-band amplifier 2. Wide-band amplifier 2 supplies an output signal FO to a transmission line through impedance converter 21, as frequency-domodulated signal OF. Wide-band amplifier 2 also supplies its output to BPF 3 which passes an QPSK voice subcarrier. The QPSK signal is then demodulated by QPSK demodulator 4 into orthogonal component signal Q and coherent component signal I. Orthogonal component signal Q and coherent component signal I are supplied to difference converter 5 to reproduce a PCM signal of 2.048 MBPS.

The reproduced PCM signal is transmitted through three routes. A first transmission route includes buffer amplifier 6 which delivers the PCM signal as digital data or bit stream BST. The reproduced PCM signal is also transmitted through a second transmission route including frame synchronization detector 7 and reference timing generator 8, wherein the start position of a frame made up of 2048 bits is detected by a 16-bit frame-synchronization-pattern extractor (not shown) provided in frame synchronization detector 7. The output of the extractor is supplied to reference timing generator 8 to cause reference timing signals CKa through CKd to be generated. Reference timing signals CKa, CKb, CKc, CKd provide reference timings for signal processing in a third transmission route.

Usually, the PCM signal is transmitted in the NRZ format, and thus possibly continues in the state logic 1 or logic 0 for a long time. Such a continued state, however, is likely to cause it to be impossible to regenerate time information from the PCM signal. In order to avoid such an undesirable situation, the original PCM signal is subject to a scrambling process in which, to the PCM signal, data in the PN pattern made up of pseudo random numbers belonging to the 10th-order M-sequence are added. The original PCM signal is further subject to a bit-interleaving process in order to prevent a burst error.

In the third transmission route, the PCM signal delivered from difference converter 5, which has been scrambled, is descrambled by descrambler 9 in synchronization with reference timing signal CKa, as described above. The descrambled voice PCM signal is deinterleaved by deinterleaver 10 in synchronization with reference timing signal CKb in order to remove the bit-interleave. Voice PCM signal PPC, after thus being processed, undergoes both single-error correction and error control through a double-error detecting code. Error detector 11, supplied with voice PCM signal PPC, detects each occurrence of a double-error in synchronization with reference timing signal Kc. Counter 12 counts occurrence pulses E during the counting period designate by reference timing signal Kd in order to provide occurrence number M. Occurrence number M is supplied to microcomputer 13 which computes the CN ratio from both occurrence number M and the theoretical relation between the CN ratio and occurrence number M. The CN ratio is delivered to a display.

The theoretical relation is derived as described below.

QPSK signal S of a unit amplitude is expressed by

where f.sub.c stands for a carrier frequency and k.sub.i =0, 1, 2 or 3 corresponding to dibits (00), (01), (11) or (10), respectively, of a modulating PCM signal. Transforming equation (1) into a linear combination of coherent and orthogonal components gives

Signal S(t) above can be represented intuitively as a vector (cos(k.sub.i .pi./2), sin(k.sub.i .pi./2)) in a two-dimensional vector space defined by two orthogonal basis vectors corresponding to cos 2.pi.f.sub.c t and sin 2.pi.f.sub.c t. This vector space is commonly referred to as a signal space. QPSK signal S(t)s modulated (or phase-shift keyed) by PCM dibits (00), (01), (11), (10) and the corresponding coordinates in the signal space are listed in Table 1.

The noise of carrier frequency f.sub.c is also represented by a linear combination of coherent and orthogonal components,

which is received together with QPSK signal S(t) by the CN ratio detecting circuit and causes input signal R(t) of the CN ratio detecting circuit to be randomly fluctuated. The input signal is represented by

Signals R(t) corresponding to the PCM dibits and their coordinates in the signal space are also listed in Table 2.

It is assumed that the values of signal R(t) fluctuating with respect to the value of signal S(t) distribute in accordance with the normal distribution. This assumption means in the scheme of the signal space concept that coordinates of signal R(t) distribute around the coordinate of QPSK signal (cos(k.sub.i .pi./2), sin(k.sub.i .pi./2)) at a density defined by the probability density function,

wherein .sigma. stands for the mean square error, and a set of constants (x.sub.0, y.sub.0) and a set of variables (x, y) represent the coordinate of signal S(t) as shown in Table 1 and the coordinate of signal R(t) as shown in Table 2, respectively. More specifically, the probability density function for each PCM dibit becomes as follows: For dibit (0 0), ##EQU2## and for dibit (1 1), ##EQU3## for dibit (0 1), ##EQU4## and for dibit (1 0), ##EQU5##

It is to be noted that probability density functions P.sub.0, P.sub.1 for dibits (0 0) and (1 1) are in mirror image relation with respect to x being equal to 0, i.e., the two probability density functions are interchanged through the mirror image x.fwdarw.-x as shown in FIG. 3, and that probability density functions for dibits (0 1) and (1 0) (not shown) are also in mirror image relation with respect to y being equal to 0. This means that area S.sub.0 enclosed by probability density function curve P.sub.0 in region x<0 equals area S.sub.1 for P.sub.1 in region x>0. Accordingly if criterion point x.sub.c for dibits (0 0), (1 1) is set at coordinate x=0, area S.sub.0 is proportional to probability Pe.sub.0 in that dibit (0 0) is erroneously detected as (1 1), and area S.sub.1 is proportional to probability Pe.sub.1 in that dibit (11) is erroneously detected as (0 0), wherein criterion point x.sub.c refers to the x-coordinate above which signal R(t) is judged to carry dibit (0 0) and below which signal R(t) is judged to carry dibit (1 1), regardless of whether QPSK signal S(t) included in signal R(t) actually carries PCM dibit (0 0) or (1 1). Since area S.sub.0 is equal to area S.sub.1 as described above, Pe.sub.0 =Pe.sub.1. Criterion limit x.sub.c which makes Pe.sub.0 and Pe.sub.1 equal is commonly called the optimum limit. (In the case that criterion limit x.sub.c is set at the optimum limit, the sum of the probability of judging (0 0) as (1 1) and the probability of judging (1 1) as (0 0) is minimal.) Probabilities Pe.sub.0 and Pe.sub.1 are commonly called the error rate for (0 0) and for (1 1), respectively. The error rate for (0 0) becomes ##EQU6## On the other hand, the carrier power of the QPSK signal per period is given by ##EQU7## Similarly the noise power per period is given by ##EQU8## It is assumed that the noise is superposed isotropically on QPSK signal S(t); in other words, vector n(n.sub.c, n.sub.s) representing noise N distributes in the signal space isotropically around vector (cosk.sub.i .pi./2, sink.sub.i .pi./2) for QPSK signal S. This involves the fact that the probability density function is represented in terms of same variance .sigma..sup.2 for any of n.sub.c, n.sub.s and n. Thus, the average value of n.sup.2 is given by

and

wherein ##EQU9## and N.sub.p denotes the averaged noise power derived by assuming that one-period noise power P.sub.N distributes in accordance with the normal distribution. The theoretical CN ratio, accordingly, is given, from equations (12) and (15), by

It follows from equation (10) that

Theoretically, error rate Pe.sub.0 is proportional to the means of the occurrence number counted for an infinitely long time. Experimentally, however, it is measured by counting the occurrence number M throughout a time period which is long enough to compare with the average occurrence time as described above. Thus, it results that

wherein k and .alpha. denote the proportional constant and the CN ratio, respectively. Proportional constant k can be experimentally determined.

The same argument holds in double-error rates Pe.sub.1, Pe.sub.2, Pe.sub.3 for dibits (1 1), (0 1), (1 0), respectively. Accordingly, the same relation holds between double-error rate Pe and CN ratio .alpha. for any of the PCM dibits. The curve of double-error rate Pe plotted against CN ratio .alpha. is represented in the logarithmic scale in FIG. 4. Thus, by storing a conversion table corresponding to equation (19) or FIG. 4 in a memory provided in microcomputer 13, CN ratio o can be obtained from double-error rate Pe determined by occurrence number M for a given counting period.

The use of a double-error detecting counter and a microcomputer for computing the CN ratio from the theoretical relation between CN ratios and double-error rates makes it unnecessary to use the noise amplifier and noise detector as in the CN ratio detecting circuit of the prior art. By virtue of the absence of errors caused by dispersions in the gain and the SN ratio and also caused by the temperature variation and the non-linear deformation, the present invention has the advantage that it is capable of obtaining a high-precesion CN ratio without any special adjustment. The present invention has another advantage in that, since it is not necessary to use expensive parts such as an A-D converter or the like, the cost can be reduced.

It is to be understood, however, that although the characteristics and advantages of the present invention have been set forth in the foregoing description, the disclosure is illustrative only, and changes may be made in the arrangement of the parts within the scope of the appended claims.