GB2364868A

GB2364868A - Sub-carrier demodulation frequency control using weighted partitioned correlation of guard period with information symbol

Info

Publication number: GB2364868A
Application number: GB0102820A
Authority: GB
Inventors: Takashi Fujiwara; Jun Ido
Original assignee: Mitsubishi Electric Corp
Current assignee: Mitsubishi Electric Corp
Priority date: 2000-02-28
Filing date: 2001-02-05
Publication date: 2002-02-06
Anticipated expiration: 2021-02-05
Also published as: CN1193524C; GB0102820D0; CN1311575A; GB2364868B; JP2001244910A; JP3789275B2

Abstract

A sub-carrier-frequency signal demodulating apparatus detects phase error by correlating a guard interval with a corresponding part of an effective symbol period. Each such error detection period is partitioned into sub-periods 5, a separate phase error value is obtained for each sub-period, and these phase error values are weighted 6, giving generally higher weights to sub-periods with higher correlation values. The weighted phase error values are then recombined 8 to obtain a final phase error signal ES, which controls the frequency of an oscillator 10 used in coherent demodulation of the sub-carrier-frequency signal. The partitioning and weighting enable the phase error to be detected with consistent accuracy, even if the signal is received on a multipath channel. The system is suitable for use with OFDM, DMT and similar multicarrier modulation systems.

Description

2364868 SUB-CARRIER-FREQUENCY SIGNAL DEMODULATING APPARATUS

BACKGROUND OF THE INVENTION

The present invention relates to a sub-carrierfrequency signal demodulating apparatus that demodulates a sub-carrier-frequency signal used in, for example, a receiver that demodulates a modulated signal transmitted by an orthogonal frequency division multiplexing system, more particularly a receiver that demodulates a sub-carrierfrequency signal by controlling a reference signal of the sub-carrier frequency and coherently detecting the subcarrier-frequency signal, using a phase error signal computed on the basis of a correlation value of a guard interval period in one symbol period of a digitally modulated signal transmitted by an orthogonal frequency division multiplexing system and a corresponding guard interval transfer period disposed at the end of the effective symbol period.

A receiver that demodulates a modulated signal transmitted by an orthogonal frequency division multiplexing system will be described, using a drawing.

FIG. 10 is a structural block diagram of an orthogonal frequency division multiplexing (OFDM) receiver that receives a signal modulated by an OFDM transmission system.

A receiving antenna 40 is attached to the OFDM receiver 50. The receiving antenna 40 is coupled to the main-carrierfrequency signal demodulator 30 in the OFDM receiver 50. A modulated signal of the OFDM transmission system, received at the antenna 40, is input to the main-carrierfrequency signal demodulator 30 and undergoes primary demodulation by a reference signal of the main carrier frequency; a subcarrier-frequency signal BS is output from the main-carrierfrequency signal demodulator 30 and undergoes secondary demodulation by a reference signal SS of the subcarrier 1 frequency in a sub-carrier-frequency signal demodulator 20; the demodulated signal DS is output from the sub-carrierfrequency signal demodulator 20.

Incidentally, the sub-carrier frequency may also be referred to as an intermediate frequency (IF), and the subcarrier-frequency signal BS as an intermediate-frequency signal or IF signal. The sub-carrier-frequency signal BS is the sum of a plurality of digitally modulated sub-carrier signals having regularly spaced frequencies centered around the frequency of the reference signal SS.

In FIG. 10, 2 denotes a multiplier, 4 denotes an effective-symbol-period delay circuit, 5 denotes a phase error detection circuit that detects phase error on the basis of correlation characteristics, 9 denotes a loop filter with adjustable gain, 10 denotes an oscillator circuit with a frequency controllable by numerical control, and SS denotes the output signal of the numerically controlled oscillator circuit 10, which oscillates according to the phase error signal input at sampling time N; SS is the reference signal of the sub-carrier frequency multiplied by the sub-carrier-frequency signal BS in the multiplier 2.

Next, the operations in the-conventional sub-carrierfrequency signal demodulator 20 shown in FIG. 10 will be described.

When the OFDM receiver 50 receives a modulated signal of the OFDM transmission system, the sub-carrier-frequency signal BS is input from the main-carrier-frequency signal demodulator 30 to the multiplier 2 in the sub-carrierfrequency signal demodulator 20, but in the initial state, a demodulated signal DS with uncorrected phase error is output from the multiplier 2. This demodulated signal DS and a signal obtained by delaying this demodulated signal DS by the effective symbol period in the effective-symbol-period delay circuit 4 are input to the phase error detection 2 circuit 5. The phase error detection circuit 5 detects the correlation characteristic of the two input signals, and outputs a phase error signal. The phase error signal is input to the loop filter 9, and a phase error signal from which high-frequency noise has been removed by the loop filter 9 is input to the numerically controlled oscillator circuit 10.

The reference signal SS output from the numerically controlled oscillator circuit 10 here comprises cosO(N) and sinO (N), corresponding to the output signal A 0 (N) of the loop f 11ter 9 at an arbitrary sampling time N (0 (N) = A 0 (N) + AO(N + 1)).

The reference signal SS output from this numerically controlled oscillator circuit 10 is controlled in oscillation so as to reduce the phase error signal ES, thereby reducing the phase error of the demodulated signal DS resulting from multiplication of the sub- carrierfrequency signal BS and reference signal SS in the multiplier 2.

The phase error signal ES is generated by detection, in the phase error detection circuit 5, of the correlation characteristic between the demodulated signal DS and a signal obtained by delay of the demodulated signal DS by an amount equal to the length of the effective symbol period in that signal, and by detection of the phase error of the demodulated signal on the basis of that correlation value. The demodulated signal DS input to the effective-symbolperiod delay circuit 4 in FIG. 10 is output after being delayed by an amount equivalent to the length of the effective symbol interval. In the phase error detection circuit 5, the correlation characteristic between the delayed demodulated signal DS and the non-delayed demodulated signal DS is detected, and the phase error signal ES is generated and output on the basis of the value 3 of the correlation characteristic. The phase error signal ES output from the phase error detection circuit 5 is output to the numerically controlled oscillator circuit 10 after harmonic components and other high- frequency noise therein have been removed in the loop filter 9.

Coherent detection is thereby performed in the subcarrier-frequency signal demodulator 20. The coherent detection operation can be outlined as follows. The phase error signal ES is generated on the basis of the demodulated signal DS, and the oscillation frequency of the reference signal SS of the sub-carrier frequency output from the numerically controlled oscillator circuit 10 is controlled by the phase error signal ES. The sub-carrier-frequency signal BS and the reference signal SS of the sub-carrier frequency are multiplied in the multiplier 2, from which the demodulated signal DS is output. If the phase error between the subcarrier-frequency signal BS and the reference signal SS of the subcarrier frequency is large, the value of the phase error signal ES becomes large.

Since the conventional sub-carrier-frequency signal demodulator apparatus 20 shown in FIG. 10 determines the phase error on the basis of a correlation characteristic of the demodulated signal DS and the signal obtained by delay of the demodulated signal DS for a predetermined time, however, when a demodulated signal delayed by reception from a multipath channel (a multipath delayed-wave demodulated signal) is received, the multipath delayed-wave demodulated signal is added to the demodulated signal received by the normal channel joining the transmitter and the receiver by the shortest path, and changes occur in the correlation characteristic determined in the phase error detection circuit 5, due to the effect of the multipath delayed-wave demodulated signal.

Specifically, as there is no correlation between the 4 guard interval transfer period in the demodulated signal and the part of the effective symbol period immediately preceding the guard interval transfer period in the multipath delayed-wave demodulated signal (the nonguardinterval-transfer period), the correlation from the beginning of the guard interval transfer period in the demodulated signal to the end of a period equivalent to the delay time of the multipath delayed wave is lowered. Moreover, since there is no correlation between the delayed guard interval period, which is the guard interval period delayed by an amount equivalent to the effective symbol period, and the guard interval transfer period in the effective symbol period in the multipath delayed- wave delayed demodulated signal wherein the multipdth delayedwave demodulated signal is further delayed by the effective symbol period, the correlation is lowered from the beginning of the delayed guard interval period, which is delayed by an amount equivalent to the effective symbol period in the demodulated signal delayed for correlation detection, to the end of a period equivalent to the delay time of the multipath delayed wave. A resulting problem that occurs is that the accuracy of the detected phase error is degraded.

The present invention addresses the problems described above, with the object of providing a sub-carrier-frequency signal demodulating apparatus that can mitigate degradation of the accuracy of a detected phase error signal and can detect phase error in a stable manner, even when there exists a delayed wave that propagates on a delayed multipath.

SUMMARY OF THE INVENTION

The invented sub-carrier-frequency signal demodulating apparatus receives a sub-carrier-frequency signal having effective symbol periods separated by guard intervals, generates a phase error signal and a demodulated signal therefrom, has an oscillator circuit generating a reference signal at a frequency controlled by the phase error signal, and has a first multiplier generating the demodulated signal by use of the reference signal. To attain the above object, the invented sub-carrier-frequency signal demodulating apparatus comprises: a delay circuit receiving one of said sub-carrier-frequency signal and said demodulated signal as a nondelayed signal, delaying said non-delayed signal by an amount equal to one effective symbol period in the subcarrier-frequency signal, and outputting a resulting delayed signal; a phase error detection circuit detecting a correlation characteristic between said delayed signal and said non-delayed signal during a detection period equal in length to one guard interval in said sub-carrier-frequency signal, dividing said detection period into a plurality of sub-periods, and generating a plurality of first partitioned phase error values, corresponding to said sub-periods, based on said correlation characteristic; a weighting function circuit receiving the first partitioned phase error values and calculating a corresponding plurality of weighting coefficients; a plurality of second multipliers generating a plurality of second partitioned phase error values by multiplying said first partitioned phase error values by the corresponding weighting coefficients; and a computation circuit performing a computation on said plurality of second partitioned phase error values, thereby obtaining said phase error signal.

In one aspect of the invention, the weighting function circuit calculates said weighting coefficients based on ratios of amplitudes of said first partitioned phase error values.

In another aspect of the invention, the weighting function circuit calculates said weighting coefficients based on ratios of powers of amplitudes of said first 6 partitioned phase error values.

In another aspect of the invention, the computation circuit obtains said phase error signal by computing an average of said second partitioned phase error values.

In another aspect of the invention, the computation circuit obtains said phase error signal by computing an average of powers of said second partitioned phase error values.

In another aspect of the invention, the delay circuit and said phase error detection circuit receive said demodulated signal as said non- delayed signal, and said first multiplier multiplies said sub-carrier- frequency signal by said reference signal. The oscillator frequency is thereby controlled by a feedback control system.

In another aspect of the invention, the delay circuit and said phase error detection circuit receive said subcarrier-frequency signal as said non-delayed signal, and said first multiplier multiplies said sub-carrierfrequency signal by said reference signal. The oscillator frequency is thereby controlled by a feed-forward control system.

In another aspect of the invention, the delay circuit and said phase error detection circuit receive said subcarrier-frequency signal as said non-delayed signal, and said first multiplier multiplies said delayed signal by said reference signal. The oscillator frequency is thereby controlled by a feed-forward control system.

The sub-carrier-frequency signal is, for example, an orthogonal frequencyd1vision multiplexed signal.

The invented sub-carrier-frequency signal demodulating apparatus may also have a loop filter coupled between said computation circuit and said oscillator circuit, removing high-frequency noise from said phase error signal.

BRIEF DESCRIPTION OF THE DRAWINGS

7 In the attached drawings:

FIG. 1 is a block diagram showing the configuration of a sub-carrierfrequency signal demodulating apparatus which is Embodiment One of the present invention; FIG. 2 (a) to (g) are timing diagrams showing the demodulated signal DS input to the phase error detection circuit 5 in FIG. 1, etc.; FIG. 3 is a timing diagram that shows the signals of FIG. 2 (c), (e), (f), and (g) on an expanded time axis, and shows the sub- periods into which the correlation characteristic detection period is partitioned in (h); FIG. 4 Is a block diagram showing the configuration of a sub-carrier- frequency signal demodulating apparatus which is Embodiment Two of the present invention; FIG. 5 is a block diagram showing the configuration of a sub-carrier- frequency signal demodulating apparatus which is Embodiment Three of the present invention; FIG. 6 is a block diagram showing the configuration of a sub-carrier- frequency signal demodulating apparatus which is Embodiment Four of the present invention; FIG. 7 is a block diagram showing the configuration of a sub-carrierfrequency signal demodulating apparatus which is Embodiment Five of the present invention; FIG. 8 is a block diagram showing the configuration of a sub-carrier- frequency signal demodulating apparatus which is Embodiment Six of the present invention; FIG. 9 is a block diagram showing the configuration of a sub-carrier- frequency signal demodulating apparatus which is Embodiment Seven of the present invention; and FIG. 10 is a block diagram of an orthogonal frequency division multiplexing (OFDM) receiver that receives a signal modulated by an OFDM transmission system.

DETAILED DESCRIPTION OF THE INVENTION

8 The present invention will now be described on the basis of the illustrated embodiments. In Figs. 1 to 9, parts with the same functions as in the conventional sub-carrierfrequency signal demodulating apparatus shown in FIG. 10 have the same reference characters. Embodiment One.

FIG. 1 Is a block diagram showing the configuration of a sub-carrierfrequency signal demodulating apparatus which is Embodiment One of the present invention.

In FIG. 1, 21 denotes a sub-carrier-frequency signal demodulator, which is the sub-carrier-frequency signal demodulating apparatus of the present embodiment; BS denotes a sub-carrier-frequency signal; 2 denotes a first multiplier; DS denotes a demodulated signal; 4 denotes an effectivesymbol-period delay circuit; 5 denotes a phase error detection circuit that detects phase errors on the basis of the correlation characteristics in periods into which the correlation detection period is partitioned; 6 denotes a weighting function circuit that calculates a weighting function by which the phase error value is multiplied in each of the periods into which the correlation detection period is partitioned; 7 denotes a second multiplier; 8 denotes a computation circuit that operates on the weighted signals in each sub-period; 9 denotes a loop filter with adjustable gain; 10 denotes an oscillator circuit with a frequency controllable by numerical control; SS denotes the reference signal output from the numerically controlled oscillator circuit 10; and ES denotes a phase error signal output from the computation circuit 8.

The first multiplier 2 corrects the phase error of the sub-carrierfrequency signal BS by multiplying the subcarrier-frequency signal BS by the reference signal SS output from the numerically controlled oscillator circuit 10 in correspondence to the phase error AO(N), and outputs the 9 demodulated signal DS. The phase error detection circuit 5 detects the correlation characteristic with the output signal of the effective-symbolperiod delay circuit 4, detects a first phase error of the demodulated signal on the basis of the value of that correlation characteristic, and outputs, as first partitioned phase error values, the value of the first phase error partitioned into a predetermined number N of sub-periods (where N is an integer equal to or greater than two and smaller than the total number of pulses of the clock signal used in the guard interval period of one symbol period). The weighting function circuit 6 calculates a plurality of weighting coefficients on the basis of the first partitioned phase error values, and outputs them to have the second multipliers 7 weight the first partitioned phase error values according to the effects of the respective multipath delayed waves. The second multipliers 7 weight the first partitioned phase error values by multiplying the first partitioned phase error values by the weighting coefficients, and output second partitioned phase error values. The computation circuit 8 outputs the phase error signal ES, which is a second phase error value, by performing an operation such as addition, for example, on the second partitioned phase error values. The loop filter 9 has, for example, at least one filter with a gain that can be varied with differing frequencies, and outputs the sum of the outputs of the filters.

Next, the operations in the sub-carrier-frequency signal demodulator 21 of the first embodiment, shown in FIG. 1, will be described, with reference to FIG. 10.

When the OFDM receiver 50 receives a signal modulated by the OFDM transmission system, the sub-carrier-frequency signal BS is input from the main-carrier-frequency signal demodulator 30 to the first multiplier 2 in the sub-carrierfrequency signal demodulator 21.

Here, coherent detection is carried out in the subcarrier-frequency signal demodulator 21; the overall operation of coherent detection is similar to the operation in the conventional sub-carrier-frequency signal demodulator 20 shown in FIG. 10.

In the initial state of the sub-carrier-frequency signal demodulator 21, a demodulated signal DS with uncorrected phase error is output from the first multiplier 2. This demodulated signal DS and a signal obtained by delaying the demodulated signal DS by the effective symbol period in the effective-symbol-period delay circuit 4 are input to the phase error detection circuit 5. The phase error detection circuit 5 detects the correlation characteristic between the two input signals, and outputs a phase error signal. This phase error signal ES is output so as to synchronize the guard interval period with the last part of the corresponding effective symbol period.

The phase error signal output from the phase error detection circuit 5 is output piecewise, -in periods into which the correlation-characteristic detection period is partitioned. That is, the phase error detection circuit 5 partitions the correlation-characteristic detection period into sub-periods 1 to N, and detects and outputs the phase error on the basis of the correlation value detected in each sub-period.

The weighting function circuit 6 calculates a weighting coefficient for each sub-period on the basis of the value output by the phase error detection circuit 5 for each subperiod. For example, It makes the weighting coefficient large for sub-periods with strong correlation, and small for sub-periods with weak correlation, that is, for periods in which the phase-error detection accuracy is low due to effects of signals (multipath delayed-wave demodulated signals) demodulated from delayed waves (multipath delayed 11 - 13 waves) that take paths other than the shortest path in a multipath, etc.; whereby the effect of multipath delayed waves on the demodulated signal is mitigated.

The weighting coefficients output from the weighting function circuit 6 in correspondence to the outputs of the phase error detection circuit 5 are multiplied together with the outputs of the phase error detection circuit 5 by the second multipliers 7, weighting the phase error signal in each sub-period. The computation circuit 8 inputs all of the signals output from the second multipliers 7, and outputs a final phase error signal ES by performing operations based on the input signals.

The output of the computation circuit 8, namely, the phase error signal ES, is input to the loop filter 9, which has adjustable gain. A phase error signal from which highfrequency noise has been removed by the loop filter 9 is input to the numerically controlled oscillator circuit 10.

The numerically controlled oscillator circuit 10 outputs cosO(N) and sinO(N), corresponding to the output signal AO(N) of the loop filter 9 at an arbitrary sampling time N (O(N) = AO(N) + AO(N + 1)); that is, it outputs the reference signal SS, which is a signal of the recovered subcarrier frequency. A demodulated signal DS with corrected phase error is obtained by multiplication of the subcarrier-frequency signal BS and the reference signal SS output from this numerically controlled oscillator circuit 10 by the first multiplier 2.

The method of detecting the correlation characteristic between the signal delayed from above-described demodulated signal DS and the non-delayed demodulated signal DS will now be described.

FIG. 2 is a timing diagram showing the demodulated signal DS input to the phase error detection circuit 5 in FIG. 1, etc.

12 FIG. 2 (a) shows the non-delayed demodulated signal DS; STO, ST1, and ST2 indicate one-symbol periods, which are the signal units of the transmitted signal. In symbol period STO (denoted STO below), there are a guard interval GIO (denoted GIO below) and an effective symbol period ESO (donated ESO below). The period from the last part of ESO forward for a period equivalent to GIO is the guard interval transfer period RGO (denoted RGO below). Similarly, in symbol periods ST1 and ST2 there are guard intervals GI1 and G12 (denoted GIl and G12 below) and effective symbol periods ES1 and ES2 (denoted ES1 and ES2 below), and guard interval transfer period RG1 (denoted RG1 below) occurs from the last part of effective symbol period ES1, preceding forward.

FIG. 2 (b) shows the demodulated signal DS delayed as a signal for detection of the correlation characteristic; DLO, DL1, and DL2 indicate delayed periods equivalent to ESO, ES1, and ES2 in FIG. 2 (a), respectively. DGO and DG1 are delayed guard interval periods, being GIO and GIl in FIG. 2 (a) delayed by the delay periods DLO and DL1.

As stated above, STO to ST2, which are signal units of the modulated signal transmitted by an orthogonal frequency division multiplexing system, comprise GIO to G12 and ESO to ES2. The guard interval periods GIO to G12 are provided at the beginning of each symbol period STO to ST2 for the purpose of obtaining signal synchronization between the transmitter and the receiver, and to prevent the received signal from being interfered with by multipath delayed signals. The effective symbol periods ESO to ES2 are the periods that include the transmitted data actually demodulated; the final portions RGO to RG1 of those periods are copied into the corresponding guard interval periods GIO to G12 in the symbol periods STO to ST2.

Since the demodulated signals in GIO and GI1 in FIG. 2 (a) are transferred copies of the demodulated signals in RGO 13 and RG1 in FIG. 2 (a), the demodulated signal in GIO has the same content as the demodulated signal in RGO, and the demodulated signal in GIl has the same content as the demodulated signal in RG1. Accordingly, the demodulated signal in RGO has the same content as the demodulated signal in DGO, which is GIO delayed by delay period DLO, and the demodulated signal in RG1 has the same content as the demodulated signal in DG1, which is GI1 delayed by delay period DL1. It is not necessarily true, however, that DGO is synchronized with RGO, or DG1 with RG1.

Correlation-characteristic detection periods DTO and DT1 (denoted DTO and DT1 below) are then provided so as to occupy the same periods as RGO and RG1, as shown in Fig. 2 (c); the correlation characteristic in DTO between the content of the demodulated signal in RGO and the content of the demodulated signal in DGO is detected; and the correlation characteristic in DT1 between the content of the demodulated signal in RG1 and the content of the demodulated signal in DG1 is detected. From the detected value of the correlation characteristic, it is possible to detect the amount of phase offset or phase error between the demodulated signal in RGO and the demodulated signal in DGO, so the phase error signal is output from the phase error detection circuit 5.

That phase error can be detected in this way has even been proven mathematically. For example, the principle by which a modulated signal in a conventional orthogonal frequency division multiplexing system is demodulated by the sub-carrier-frequency signal demodulating apparatus has been shown mathematically by Paul H. Moose in "A Technique for Orthogonal Frequency Division Multiplexing Frequency Offset Correction," IEEE Transactions on Communications, Vol. 42, No. 10, October 1994.

The phase error is obtained here by use of the 14 frequency offset. First, let sn by the n-th sample in the Ith symbol of an OFDM signal including a frequency offset 8f. In the case in which n is in the guard interval period, the autocorrelation function RN is defined as shown in equation RN E [Sn S.,N Where E[xl is the expectation of x, and x is the complex conjugate of x. N is the size of the FFT (fast Fourier transform).

When a frequency offset is present, if noise is ignored, sn+N can be expressed as In the next equation (2) using Sn and the frequency offset 8f.

Sn+N = s,, exp[j27rbfN] (2) If equation (2) is substituted into equation (1), the next equation (3) is obtained.

RN = -!E S"12 XP[j 2jr6ff] (3) If it is assumed here that the calculation of the expectation E[x] can be approximated by the time average in the guard interval GI, equation (1) is indicated as in the next equation (4).

1 -1.

RN = S nSn+N 2N g n.- Accordingly, the frequency offset 6f can be estimated as in the following equation (5) by extracting the phase component from the complex data obtained in equation (4).

8f tan -1 Im[R,, 2nN ReFR] From this equation (5) for the frequency offset 8f, the phase error 80 can be obtained as shown in the next equation (6).

= tan- (6) [ _R_eTR_N 1.

From the above, it can be understood mathematically that the phase error of the demodulated signal can be detected by finding the correlation characteristic of the guard interval period and the final part of the corresponding effective symbol period, as shown in equations (1) to (4), and extracting the phase component therefrom, as shown in equations (5) and (6).

FIG. 2 (d) to (g) are timing diagrams showing relationships between the above-mentioned multipath delayedwave demodulated signal and the normal demodulated signal.

FIG. 2 (d) shows a multipath delayed-wave demodulated signal; (e) shows the demodulated signal of (a) and the multipath delayed-wave demodulated signal placed one above another; (f) shows the period in which the correlation characteristic varied and the period in which the correlation characteristic does not vary in the correlation characteristic detection period; and (g) shows the demodulated signal delayed for correlation detection and the delayed multipath delayed-wave demodulated signal placed one 16 above another.

The symbol periods MSO to MS2 in the multipath delayedwave demodulated signal shown in FIG. 2 (d) propagate on a delayed path in the multipath channel (denoted the delayed multipath below), so they are each delayed by an amount dt with respect to the symbol periods STO to ST2 in the normaldemodulated signal shown in (a). Therefore, the MEOb shown at the bottom of FIG. 2 (e) in the demodulated signal that propagates on the delayed multipath (the period in the multipath-delayed-wave demodulated signal corresponding to the above RGO) is also delayed by the period dt with respect to RGO in the demodulated signal that propagates on the normal path, shown at the top of FIG. 2 (e). Accordingly, an interval lasting for a period dt from the beginning of RGO in the demodulated signal, that is, the correlation characteristic variation period DTOa shown in FIG. 2 (f), is affected by MEOa in the demodulated signal that propagates on the delayed multipath (the period in the multipath delayed- wave demodulated signal corresponding to the non-RGO part of the effective symbol period ESO). Similarly, the correlation characteristic variation period DTOb shown in FIG. 2 (f) is affected by MEOb in the demodulated signal that propagates on the delayed multipath. Here, RGO and MEOb are correlated, because they are the same signal in a delayed relationship, but RGO and MEOa are not correlated.

Meanwhile, the MIO shown at the bottom of FIG. 2 (g) in the modulated signal that propagates on the delayed multipath and is delayed for correlation detection (the period in the multipath delayed-wave demodulated signal delayed for correlation detection corresponding to the above DGO) is also delayed by the period dt with respect to DGO In the modulated signal that propagates on the normal path and is delayed for correlation detection, shown at the top of FIG. 2 (g). Accordingly, an interval lasting for a period dt 17 from the beginning of DGO in the demodulated signal that was delayed for correlation detection, that is, the correlation characteristic variation period DTOa shown in FIG. 2 (f), is affected by MDOb in the modulated signal that propagates on the delayed multipath and is delayed for correlation detection (the period in the multipath delayed-wave demodulated signal corresponding to the non-RGO part of the effective symbol period ESO). Similarly, the correlation characteristic variation period DTOb shown in FIG. 2 (f) is affected by MIO in the demodulated signal that propagates on the delayed multipath and is delayed for correlation detection. Here, DGO and MIO are correlated, because they are the same signal in a delayed relationship, but DGO and MDOb are not correlated.

Accordingly, since there is no correlation between RGO and MEOa, the correlation from the beginning of RGO for a period dt in the demodulated signal is lowered, and since there is no correlation between DGO and MDOb, the correlation from the beginning of DGO for a period dt in the demodulated signal delayed for correlation detection is lowered, as a result of which, for example, the accuracy of the detected pha.s,e error is degraded in the conventional sub-carrier-frequency signal demodulator 20 shown in FIG. 10.

The present embodiment detects the interval lasting for the period dt from the beginning of RGO and the interval lasting for the period dt from the beginning of DGO, and reduces the weights of these periods, so that the accuracy of the final phase error is not degraded.

FIG. 3 is a timing diagram that shows the signals of FIG. 2 (c), (e), (f), and (g) on an expanded time axis, and shows the sub-periods into which the correlation characteristic detection period is partitioned in (h).

FIG. 3 (c) and (e) to (g) are diagrams showing the corresponding signals in FIG. 2 on an expanded time axis; 18 (h) shows the sub-periods into which the correlation characteristic detection period is partitioned by the phase error detection circuit 5 of the present embodiment. Finely divided sub-periods are desirable, but the ability to detect correlation cannot be finer than the wavelength of the clock signal, so in the present embodiment, the number of subperiods into which the period is partitioned is a predetermined number N (where N is an integer equal to or greater than two, and smaller than the total number of pulses of the clock signal used in the guard interval period of one symbol period).

By thus partitioning the correlation characteristic detection period, it is possible to identify the period dt by which the sub-carrier-frequency signal that propagated on the delayed multipath is delayed; accordingly, it is possible to identify the correlation characteristic variation period DTOa, this being the period in which the correlation is reduced when the sub-carrier-frequency signal that propagated on the delayed multipath is added to the sub-carrier signal that propagated on the normal path.

In the present embodiment, the weighting coefficients corresponding to the correlation characteristic variation period DTOa are set low by the weighting function circuit 6, and the weighting coefficients corresponding to the correlation characteristic variation period DTOb are set high. Thus in the final phase error signal computed by the computation circuit 8, the effect of the sub-carrierfrequency signal that propagated on the delayed multipath is reduced.

As described above, the sub-carrier-frequency signal demodulating apparatus 21 of the present embodiment comprises a phase error detection circuit 5 that partitions the correlation characteristic detection period into several sub-periods and outputs a first partitioned phase error 19 value for each sub-period, a weighting function circuit 6 that calculates a weighting function for weighting the first partitioned phase error values, second multipliers 7 that multiply the first partitioned phase error values output from the phase error detection circuit 5 by the corresponding weighting coefficients and output second partitioned phase error values, and a computation circuit 8 that operates on the second partitioned phase error values; partitioning the correlation characteristic detection period into several sub-periods, it weights the phase error value in each sub-period, and outputs, as the final phase error signal, a value resulting from operations on the weighted phase error value in each sub-period; on the basis of the phase error signal, it recovers a signal of the sub-carrier frequency and generates a reference signal, and is adapted to detect the sub-carrier-frequency signal coherently, so even when a delayed signal that propagates on a delayed multipath is present, the effect of the period in which the correlation is weakened by that delayed signal can be mitigated, degradation of the accuracy of the detected phase error can be mitigated, and the phase error can be detected in a stable manner. Embodiment Two.

FIG. 4 is a block diagram showing a sub-carrierfrequency signal demodulating apparatus which is Embodiment Two of the present invention. Incidentally, in FIGs. 4 to 9 below, parts with the same functions as in the sub-carrierfrequency signal demodulating apparatus of the first embodiment, shown in FIG. 1, have the same reference characters, and repeated descriptions will be omitted.

In the sub-carrier-frequency signal demodulator 22 of the present embodiment, in regard to the setting of the weighting function in the weighting function circuit 6 shown in FIG. 1, the weights are calculated by comparing the maximum values of the amplitudes of the first partitioned phase error values, and normalizing the comparison results.

In FIG. 4, 12 denotes a maximum value detection circuit that detects a maximum value from the first partitioned phase error values output from the phase error detection circuit 5 in each sub-period, and 13 denotes normalizing circuits that calculate the weighting function by normalizing the output of the phase error detection circuit 5 in each sub-period according to the maximum value detected by the maximum value detection circuit 12.

The operation of the sub-carrier-frequency signal demodulator 22 of the present embodiment will be described below.

The operation of the sub-carrier-frequency signal demodulator 22 of the present embodiment in the initial state from the output from multiplier 2 of the demodulated signal DS with uncorrected phase error to the output of the first partitioned phase error values partitioned by the phase error detection circuit 5 is similar to Embodiment One.

The weighting function circuit 6 calculates the weighting coefficient for each sub-period on the basis of the first partitioned phase error values of the phase error detection circuit 5. In the present embodiment, the first partitioned phase error values, which are the output of the phase error detection circuit 5, are input to the maximum value detection circuit 12, and the largest phase error value among the first partitioned phase error values is detected as the maximum value. The first partitioned phase error values are normalized by the normalizing circuits 13, making the detected maximum value equal to one, for example, and these normalized results constitute the weighting function of the sub- periods. By normalization of the f irst partitioned phase error values, the weighting function becomes large in the sub-periods with strong correlation, 21 I and small in sub-periods with weak correlation; that is, in periods in which the phase error detection accuracy is lowered by a delayed signal that propagated on a delayed multipath.

The weighting of the first partitioned phase error values is performed by multiplying the first partitioned phase error values of the phase error detection circuit 5 together with the weighting coefficients output from the weighting function circuit 6 corresponding to the first partitioned phase error values of the phase error detection circuit 5, by means of the second multipliers 7, creating the second partitioned phase error values. The second partitioned phase error values output from all of the second multipliers 7 are Input to the computation circuit 8, which outputs a phase error signal ES, which is the second phase error value, by performing operations such as addition, for example, on the input signals. Subsequent operations are similar to Embodiment One.

The amplitude of the phase error detection value in periods in which the phase error value is smaller than in other periods, that is, in periods of weak correlation due to the effects of multipath delayed waves, is curtailed by setting the weighting coefficients in this way. Accordingly, the effects of sub-periods in which the correlation is weakened by a delayed signal that propagated on a delayed multipath can be mitigated.

As described above, the sub-carrier-frequency signal demodulator 22 according to the present embodiment is adapted to calculate the weighting function calculated in the weighting function circuit 6 on the basis of a ratio of amplitudes of the first partitioned phase error values, carried out by the maximum value detection circuit 12 and normalizing circuits 13, so with a comparatively simple circuit configuration, even when a delayed wave that 22 propagates on a delayed multipath is present, the ef f ect of periods in which correlation is weakened by the multipath delayed signal can by mitigated even more than in the f irst embodiment, the degradation of the accuracy of the detected phase error can be mitigated, and the phase error can be detected in a stable manner. Embodiment Three.

FIG. 5 is a block diagram showing a sub-carrierfrequency signal demodulating apparatus which is Embodiment Three of the present invention.

In the sub-carrier-frequency signal demodulator 23 of the present embodiment, in regard to the setting of the weighting function in the weighting function circuit 6 shown in FIG. 1, the weights are calculated by comparing powers of the amplitudes of the first partitioned phase error values input In the sub-periods, and normalizing the comparison results.

In FIG. 5, 14 denotes power circuits that raise the first partitioned phase error signals output from the phase error detection circuit 5 in each sub-period to a power, 12 denotes a maximum value detection circuit that detects a maximum value from the outputs of the power circuits 14, and 13 denotes normalizing circuits that calculate the weighting function by normalizing the outputs of the power circuits 14 in each sub-period according to the maximum value detected by the maximum value detection circuit 12.

The operation of the sub-carrier-frequency signal demodulator 23 of the present embodiment will be described below.

The operation of the sub-carrier-frequency signal demodulator 23 of the present embodiment in the initial state from the output from multiplier 2 of the demodulated signal DS with uncorrected phase error to the output of the first partitioned phase error values partitioned by the 23 k phase error detection circuit 5 is similar to Embodiment One.

The weighting function circuit 6 calculates the weighting function for the sub-periods on the basis of the first partitioned phase error values of the phase error detection circuit 5. In the present embodiment, the first partitioned phase error values of the phase error detection circuit 5 are input to the power circuits 14, and a power value of each first partitioned phase error value is calculated in the power circuits 14. The outputs of the power circuits 14 are input to the maximum value detection circuit 12, and the largest phase error value is detected as the maximum value. The powers of the phase error values of the sub-periods are normalized by the normalizing circuits 13, making the detected maximum value equal to one, for example, and these normalized results constitute the weighting function of the sub-periods. Normalization of the powers of the phase error values in the sub-periods makes the weighting function large in sub-periods with strong correlation, and small in sub-periods with weak correlation; that is, in periods in which the phase error detection accuracy is lowered by a delayed signal that propagated on a delayed multipath.

The first partitioned phase error values, which are the outputs of the phase error detection circuit 5, are multiplied together with the weighting coefficients output from the weighting function circuit 6 corresponding to the outputs of the phase error detection circuit 5, by means of the second multipliers 7, performing the weighting of the first partitioned phase error signals in each sub-period, and creating the second partitioned phase error values. The second partitioned phase error values output from all of the second multipliers 7 are input to the computation circuit 8, which outputs a phase error signal ES, which is the second phase error value, by performing operations on the input 24 signals. Subsequent operations are similar to Embodiment One.

The amplitude of the phase error detection value in periods in which the first partitioned phase error value is smaller than in other periods, that is, in periods of weak correlation due to the effects of multipath delayed waves, is curtailed by setting the weighting coefficients in this way. Accordingly, the effects of sub-periods in which the correlation is weakened by a delayed signal that propagated on a delayed multipath can be mitigated.

As described above, the sub-carrier-frequency signal demodulator 23 according to the present embodiment is adapted to calculate the weighting function calculated in the weighting function circuit 6 on the basis of a ratio of powers of amplitudes of the first partitioned phase error values, carried out by the power circuits 14, maximum value detection circuit 12, and normalizing circuits 13, so the degree of weighting can be enhanced even more than in Embodiment Two, and with a comparatively simple circuit configuration, even when a delayed wave that propagates on a delayed multipath is present, the effect of periods in which correlation is weakened by the multipath delayed signal can by mitigated still further, the degradation of the accuracy of the detected phase error can be mitigated, and the phase error can be detected in a stable manner. Embodiment Four.

FIG. 6 is a block diagram showing a sub-carrierfrequency signal demodulating apparatus which is Embodiment Four of the present invention.

In the sub-carrier-frequency signal demodulator 24 of the present embodiment, in the computation circuit 8 shown in FIG. 1, an operation is carried out that averages the second partitioned phase error values, which have been weighted by the weighting function circuit 6 and second multipliers 7.

In FIG. 6, 15 denotes an averaging circuit, provided in the computation circuit 8, that calculates the average of the second partitioned phase error values, which are the outputs of the second multipliers 7.

The operation of the sub-carrier-frequency signal demodulator 24 of the present embodiment will be described below.

The operation of the sub-carrier-frequency signal demodulator 24 of the present embodiment in the initial state from the output from multiplier 2 of the demodulated signal DS with uncorrected phase error to the weighting of the first partitioned phase error values in each sub-period by the weighting function circuit 6 and the second multipliers 7 is similar to Embodiment One.

The second partitioned phase error values output from all of the second multipliers 7 are input to the computation circuit 8, which outputs a phase error signal ES, which is the second phase error value, by performing operations on the input signals.

Here, since the computation circuit 8 of the present embodiment is an averaging circuit 15, when the second partitioned phase error values output from all of the second multipliers 7 are input to the averaging circuit 15, the averaging circuit 15 calculates their average value, and that average value is output from the averaging circuit 15 as the phase error signal ES, which is the second phase error value. Subsequent operations are similar to Embodiment One.

The effect of sub-periods in which the correlation is weakened by a delayed signal that propagated on a delayed multipath can be mitigated by calculating the average value of the phase error signals in this way.

As described above, the sub-carrier-frequency signal demodulator 24 according to the present embodiment 26 configures the computation circuit 8 with an averaging circuit 15, and is adapted to obtain the second phase error value by calculating the average of the weighted second partitioned phase error values in the computation circuit 8, so with a comparatively simple circuit configuration, even when a delayed wave that propagates on a delayed multipath is present, the effect of periods in which correlation is weakened by the multipath delayed signal can by mitigated still further, the degradation of the accuracy of the detected phase error can be mitigated, and the phase error can be detected in a stable manner. Embodiment Five.

FIG. 7 is a block diagram showing a sub-carrierfrequency signal demodulating apparatus which is Embodiment Five of the present invention.

In the sub-carrier-frequency signal demodulator 25 of the present embodiment, in the computation circuit 8 shown in FIG. 1, an operation is carried out that calculates an average of powers of the second partitioned phase error values weighted in the weighting function circuit 6 and the second multipliers 7.

In FIG. 7, 16 denotes a power averaging circuit, provided in the computation circuit 8, that calculates the average of powers of the second partitioned phase error values, which are the outputs of the second multipliers 7.

The operation of the sub-carrier-frequency signal demodulator 25 of the present embodiment will be described below.

The operation of the sub-carrier-frequency signal demodulator 25 of the present embodiment in the initial state from the output from multiplier 2 of the demodulated signal DS with uncorrected phase error to the weighting of the first partitioned phase error values in each sub-period by the weighting function circuit 6 and the second 27 multipliers 7 is similar to Embodiment One.

Here, since the computation circuit 8 of the present embodiment is a power averaging circuit 16, when the second partitioned phase error values output from all of the second multipliers 7 are input to the power averaging circuit 16, the power averaging circuit 16 calculates an average value of powers thereof, and that average value of powers is output from the power averaging circuit 16 as the final phase error signal. Subsequent operations are similar to Embodiment One.

The effect of sub-periods in which the correlation is weakened by a delayed signal that propagated on a delayed multipath can be mitigated by calculating the average value of powers of the second partitioned phase error values in this way.

As described above, the sub-carrier-frequency signal demodulator 25 according to the present embodiment configures the computation circuit 8 with a power averaging circuit 16, and is adapted to obtain the second phase error value by calculating an average of powers of the weighted second partitioned phase error values in the computation circuit 8, so the degree of weighting can be enhanced even more than in Embodiment Four shown in FIG. 6, and with a comparatively simple circuit configuration, even when a delayed wave that propagates on a delayed multipath is present, the effect of periods in which correlation is weakened by the multipath delayed signal can by mitigated still further, the degradation of the accuracy of the detected phase error can be mitigated, and the phase error 28 can be detected in a stable manner. Embodiment Six.

FIG. 8 is a block diagram showing a sub-carrierfrequency signal demodulating apparatus which is Embodiment Six of the present invention.

The sub-carrier-frequency signal demodulator 26 of the present embodiment changes the position of the first multiplier 2, which corrects the phase error of the subcarrier-frequency signal BS and outputs the demodulated signal DS, to a position different from that in the abovedescribed Embodiment One. More specifically, in Embodiment One, the first multiplier 2 was disposed so as to feed back the reference signal SS, which is a signal of the recovered sub-carrier frequency output from the numerically controlled oscillator circuit 10, but in the present embodiment, the first multiplier 2 is disposed so as to feed the reference signal SS forward. That is, the first multiplier 2 is disposed so that its output is output as the demodulated signal DS, but is not output toward the effective-symbolperiod delay circuit 4 and the phase error detection circuit 5.

The operation of the sub-carrier-frequency signal demodulator 26 of the present embodiment will be described below.

The sub-carrier-frequency signal BS of the sub-carrierfrequency signal demodulator 26 of the present embodiment is input directly to the phase error detection circuit 5, and is also input directly to the effectivesymbol-period delay circuit 4. The signal delayed by the effective symbol period by the effective-symbol-period delay circuit 4 is also input to the phase error detection circuit 5. The operations performed in the phase error detection circuit 5 and the subsequent operations, until the demodulated signal DS with corrected phase error is obtained by multiplication of the 29 reference signal SS output from the numerically controlled oscillator circuit 10 together with the sub-carrierfrequency signal BS by the first multiplier 2, are similar to Embodiment One.

The convergence time can be shortened by thus making the coherent detection circuit in the sub-carrier-frequency signal demodulator 26 a feed-forward circuit instead of a feedback circuit.

Since the sub-carrier-frequency signal demodulator 26 of the present embodiment makes the coherent detection circuit a feed-forward circuit as described above, with a comparatively simple circuit configuration, even when a delayed wave that propagates on a delayed multipath is present, the effect of periods in which correlation is weakened by the multipath delayed signal can by mitigated, the degradation of the accuracy of the detected phase error can be mitigated, the phase error can be detected in a stable manner, and furthermore, the convergence time can be shortened. Embodiment Seven.

FIG. 9 is a block diagram showing a sub-carrierfrequency signal'demodulating apparatus which is Embodiment Seven of the present invention.

The sub-carrier-frequency signal demodulator 27 of the present embodiment changes the position of the first multiplier 2, which corrects the phase error of the subcarrier-frequency signal BS and outputs the demodulated signal DS, to a position different from that in the abovedescribed Embodiment One. More specifically, in Embodiment One, the first multiplier 2 was disposed so as to feed back the reference signal SS, which is a signal of the recovered sub-carrier frequency output from the numerically controlled.oscillator circuit 10, but in the present embodiment, the first multiplier 2 is disposed so as to feed the reference signal SS forward to a point following the effective-symbolperiod delay circuit 4. That is, the first multiplier 2 is disposed so as to input the output of the effective-symbolperiod delay circuit 4, and its output is output as the demodulated signal DS, but is not output toward the effective-symbol-period delay circuit 4 and the phase error detection circuit 5.

The operation of the sub-carrier-frequency signal demodulator 27 of the present embodiment will be described below.

The sub-carrier-frequency signal BS of the sub-carrierfrequency signal demodulator 27 of the present embodiment is input directly to the phase error detection circuit 5, and is also input directly to the effectivesymbol-period delay circuit 4. The signal delayed by the effective symbol period by the effective-symbol-period delay circuit 4 is input to the phase error detection circuit 5, and is also input to the first multiplier 2. The operations performed in the phase error detection circuit 5 and the subsequent operations, until the demodulated signal DS with corrected phase error is obtained by multiplication of the reference signal SS output from the numerically controlled oscillator circuit 10 together with the sub- carrier- frequency signal BS by the first multiplier 2, are similar to Embodiment One.

The convergence time can be shortened by thus making the coherent detection circuit in the sub-carrier-frequency signal demodulator 27 a feed-forward circuit instead of a feedback circuit.

Since the sub-carrier-frequency signal demodulator 27 of the present embodiment makes the coherent detection circuit a feed-forward circuit in which the signal delayed by the effective symbol period by the effectivesymbolperiod delay circuit 4 is input to the first multiplier 2 as described above, with a comparatively simple circuit 31 configuration, even when a delayed wave that propagates on a delayed multipath is present, the effect of periods in which correlation is weakened by the multipath delayed signal can by mitigated, the degradation of the accuracy of the detected phase error can be mitigated, the phase error can be detected in a stable manner, and furthermore, the convergence time can be shortened, phase error correction can be carried out on the symbol period in which the phase error was actually detected, and a sub-carrier-frequency signal demodulating apparatus that operates in a stable manner with better accuracy can be obtained.

The sub-carrier-frequency signal demodulator in each of the above embodiments was described as applied in a receiver for an orthogonal frequency division multiplexing (OFDM) system, but it can also be incorporated as a component circuit into the demodulator of, for example, a digital television receiver or the like.

Furthermore, the sub-carrier-frequency signal demodulating apparatus of the present invention is not limited to circuits of the various types constituting the sub-carrier-frequency signal demodulators described inthe embodiments above, to their interconnection states, to the types of principal signals connected to the sub-carrierfrequency signal demodulators, or to their control methods or the like as described in the embodiments above.

In all aspects of the present invention, a sub-carrierfrequency signal demodulating apparatus that demodulates a sub-carrier-frequency signal provides means of partitioning the correlation characteristic detection period into several sub-periods, when the correlation characteristic of the guard interval period and the corresponding last part of the effective symbol period is determined in order to detect the phase error; means for weighting a first partitioned phase error value in each sub- period; and means for determining a 32 computed value thereof as a second phase error value, with the effect that:

the influence of periods in the correlation detection period in which the correlation is weakened by a subcarrier-frequency signal delayed by propagation through a multipath channel can be mitigated, degradation of phase error detection accuracy can be prevented, and operation can be stabilized.

According to one aspect of the invention, when the correlation characteristic detection period is partitioned Into several sub-periods and the first partitioned phase error values are weighted in each subperiod, means are provided for calculating the weighting function on the basis of ratios of amplitudes of the first phase error values, with the effect that:

with a comparatively simple circuit configuration, the influence of periods in the correlation detection period in which correlation is weakened by a sub-carrier-frequency signal delayed by propagation through a multipath channel can be mitigated, degradation of the phase error detection accuracy can be prevented, and operation can be stabilized.

According to another aspect of the invention, when the correlation characteristic detection period is partitioned into several sub-periods and the first partitioned phase error values are weighted in each subperiod, means are provided for calculating the weighting function on the basis of ratios of powers of amplitudes of the first partitioned phase error values, with the effect that:

the degree of weighting can be further enhanced, the influence of periods in the correlation detection period in which correlation is weakened by a sub-carrier-frequency signal delayed by propagation through a multipath channel can be mitigated, degradation of the phase error detection accuracy can be prevented, and operation can be stabilized.

33 According to another aspect of the invention, means are provided for partitioning the correlation characteristic detection period into several sub-periods, weighting the first partitioned phase error value in each sub-period, and determining their average value as the second phase error value, with the effect that:

According to another aspect of the invention, means are provided for partitioning the correlation characteristic detection period into several sub-periods, weighting the first partitioned phase error value in each sub-period, and determining an average value of powers thereof as the second phase error value, with the effect that:

the degree of weighting can be further enhanced, the influence of periods in the correlation detection period in which correlation is weakened by a sub-carrier-frequency signal delayed by propagation through a multipath channel is mitigated, degradation of the phase error detection accuracy is prevented, and operation is stabilized.

According to another aspect of the invention, a feedback system is configured in which the phase error is determined from the demodulated signal, with the effect that:

phase error is reliably corrected in the demodulated signal.

According to another aspect of the invention, a feedforward system is configured in which the phase error is determined from the sub-carrierfrequency signal, with the effect that:

34 with a comparatively simple circuit configuration, even when a delayed wave that propagates on a delayed multipath is present, the influence of periods in the correlation detection period in which correlation is weakened by the multipath delayed signal can be mitigated, degradation of the phase error detection accuracy can be prevented, operation can be stabilized, and furthermore, the convergence time can be shortened.

According to another aspect of the invention, a feedforward system is configured in which the phase error is determined from the sub-carrierfrequency signal, and the demodulated signal is obtained from the delayed sub-carrierfrequency signal, with the effect that:

with a comparatively simple circuit configuration, even when a delayed wave that propagates on a delayed multipath is present, the influence of periods in the correlation detection period in which correlation is weakened by the multipath delayed signal can be mitigated, degradation of the phase error detection accuracy can be prevented, operation can be stabilized, and furthermore, the convergence time can be shortened, phase error correction can be carried out on the symbol in which the phase error was actually detected, and operation can be stabilized more accurately.

According to another aspect of the invention, the subcarrier-freguency signal is an orthogonal frequency-division multiplexed signal, with the effect that:

the stringent frequency-control requirements in demodulation of an orthogonal frequency-division multiplexed signal can be satisfied.

According to another aspect of the invention, the subcarrier-frequency signal demodulating apparatus also provides a loop filter, with the effect that:

high-freguency noise can be removed from the phase error signal.

36

Claims

CLAIMS:

1. A sub-carrier-frequency signal demodulating apparatus receiving a sub-carrier-frequency signal having effective symbol periods separated by guard intervals, and generating a phase error signal and a demodulated signal therefrom, having an oscillator circuit generating a reference signal at a frequency controlled by the phase error signal and a first multiplier generating the demodulated signal by use of the reference signal, comprising:

a delay circuit receiving one of said sub-carrier frequency signal and said demodulated signal as a non delayed signal, delaying said non-delayed signal by an amount equal to one effective symbol period in the sub carrier-frequency signal, and outputting a resulting delayed signal; a phase error detection circuit detecting d correlation characteristic between said delayed signal and said non delayed signal during a detection period equal in length to one guard interval in said sub-carrier-frequency signal, dividing said detection period into a plurality of sub periods, and generating a plurality of first partitioned phase error values, corresponding to said sub-periods, based on said correlation characteristic; a weighting function circuit receiving the first partitioned phase error values and calculating a corresponding plurality of weighting coefficients; a plurality of second multipliers generating a plurality of second partitioned phase error values by multiplying said first partitioned phase error values by the corresponding weighting coefficients; and a computation circuit performing a computation on said plurality of second partitioned phase error values, thereby obtaining said phase error signal.

37

2. The sub-carrier-frequency signal demodulating apparatus of claim 1, wherein said weighting function circuit calculates said weighting coefficients based on ratios of amplitudes of said phase error values.

3. The sub-carrier-frequency signal demodulating apparatus of claim 1, wherein said weighting function circuit calculates said weighting coefficients based on ratios of powers of amplitudes of said phase error values.

4. The sub-carrier-frequency signal demodulating apparatus of claim 1, wherein said computation circuit obtains said phase error signal by computing an average of said second partitioned phase error values.

5. The sub-carrier-frequency signal demodulating apparatus of claim 1, wherein said computation circuit obtains said phase error signal by computing an average of powers of said second partitioned phase error values.

6. The sub-carrier-frequency signal demodulating apparatus of any of claims 1, wherein said sub-carrier-frequency signal is an orthogonal frequency-division multiplexed signal.

7. The sub-carrier-frequency signal demodulating apparatus of any of claims 1, further comprising:

a loop filter coupled between said computation circuit and said oscillator circuit, removing high-frequency noise from said phase error signal.

8. The sub-carrier-frequency signal demodulating apparatus of any of claims 1-7, wherein said delay circuit and said 38 phase error detection circuit receive said demodulated signal as said non- delayed signal, and said first multiplier multiplies said sub-carrier- frequency signal by said reference signal.

9. The sub-carrier-frequency signal demodulating apparatus of any of claims 1-7, wherein said delay circuit and said phase error detection circuit receive said sub-carrier frequency signal as said non-delayed signal, and said first multiplier multiplies said sub-carrier-frequency signal by said reference signal.

10. The sub-carrier-frequency signal demodulating apparatus of any of claims 1-7, wherein said delay circuit and said phase error detection circuit receive said sub-carrier frequency signal as said non-delayed signal, and said first multiplier multiplies said delayed signal by said reference signal.

11. A method of controlling the frequency of a local oscillator used to demodulate a signal comprising effective symbol periods separated by guard intervals, the method comprising providing correlation signals representing the degree of correlation between samples of a received signal which are separated by a period corresponding to the effective symbol period, the correlation signals representing the correlation in respective sub-periods which are shorter than the guard interval, and providing a local oscillator control signal on the basis of a weighted combination of the correlation signals.

12. A sub-carrier-frequency signal demodulation apparatus substantially as herein described with reference to any of Figures 1 and 4 to 9 of the accompanying drawings.

39

13. A method of controlling the frequency of a local oscillator, the method being substantially as herein described with reference to any of Figures 1 and 4 to 9 of the accompanying drawings.