A quadrature demodulator employing timing recovery interpolator

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is based upon and claims the benefit of priority from the prior Japanese Patent Application No. 2005-029493, filed on February 4, 2005, the entire contents of which are incorporated herein by reference.

BACKGROUND OF THE INVENTION

The present invention relates to an apparatus for communications, in particular, an apparatus for communications preferably using for a demodulation apparatus demodulating signals quadrature modulated.

The quadrature amplitude modulation/demodulation is known as modulation/demodulation technology, which is used for signal conveyance in the communications such as CATV (cable television), satellite communications, multi-channel radio communications. Fig. 9 is a block diagram showing a configuration of a demodulator for demodulating quadrature amplitude modulation wave (modulated signal) in the prior art. As shown in Fig. 9, the demodulator is composed of an RF unit 11, a MIX unit 12, A/D converters 13a, 13b, a rotator (AFC: Auto Frequency Control) 14, interpolators 91a, 91b, a rotator (CR: Carrier Recovery) 16, low-pass filters (LPFs) 17a, 17b, a clock (CLK) phase error operator 18, and a decoder 19.

A modulated signal input from a cable or an antenna is performed for frequency conversion in the RF unit 11, so that its frequency is converted into the frequency possibly to demodulate. Subsequently, at the MIX unit 12, the signal is quadrature demodulated into base band signals as I/Q axes (hereafter, I axis is called as "I channel", Q axis is called as "Q channel"), which are digitized by the A/D converters 13a, 13b.

That is, the modulated wave, which is frequency converted by the RF unit 11, is quadrature demodulated into the signals corresponding to the phase axes of I channel as in-phase and Q channel as quadrature-phase, and analog formatted I channel and Q channel signals are output. The analog formatted I channel signal is converted into a digital formatted I channel signal by the A/D converter 13a, and similarly, the analog formatted Q channel signal is converted into a digital formatted Q channel signal by the A/D converter 13b.

In addition, the A/D converters 13a, 13b may be arranged at different locations from those in Fig. 9 depending upon the processing method. For example, if they are arranged in front of the quadrature demodulation i.e. MIX unit 12, the quadrature demodulation within the MIX unit 12 will become a digital process.

The digital data of I/Q channels respectively output from the A/D converters 13a, 13b are synchronized with a carrier. This process is performed by shifting frequency in the rotators 14 and 16; roughly adjusting (coarse adjustment) in the rotator (AFC) 14 and perfectly synchronizing (fine adjustment) in the rotator (CR) 16. However, if the deviation of frequency is small, the process in the rotator (AFC) 14 may be omitted.

Since sampling in the A/D converters 13a and 13b is performed by using clock signal Fclk1 asynchronized to clock signal Fclk2 which is synchronized with the clock signal of the transmission side, in order to restore digital data from the demodulated base band signals of I/Q channels, it is necessary to provide synchronization for sampling timing, which is generally called clock synchronization. Fig. 9 illustrates a configuration which is not for synchronization between the clock signal Fclk1 of sampling timing and data, but for synchronization through adjusting the phase of data, and which creates a signal of a phase position obtaining data thereof, with an interpolation process through the interpolators 91a and 91b in order to synchronize the phase of the data. The created data become those synchronizing with the timing Fclk2, which is synchronized with the clock signal of the transmission side.

The base band signals of I/Q channels phase-adjusted by the interpolators 91a and 91b are finally synchronized with the carrier by the rotator (CR) 16; wave-shaping is carried out through low-pass filters (LPFs) 17a and 17b, which are called roll-off filters, and thereby signals are created. In general, because of processing such as error correcting and framing on the digital data, they will be decoded by the decoder (DEC) 19 as necessary.

The interpolating operation through the interpolators 91a and 91b shown in Fig. 9 will be described, referring to Fig. 10. In Fig. 10, DTI is input data, which is a sine wave in the figure for easy understanding of it. STA1 to STA4 shown by broken lines are sampling timings for the input data DTI based on the clock signal Fclk1, and STB1 to STB4 shown by chain-dot lines are original sampling timings based on the clock signal Fclk2.

In order to obtain signals at positions where original sampling timings are present from the signals sampled by the A/D converters 13a and 13b, original data sequence ID1 to ID4 at desirable sampling timings STB1 to STB4 are calculated and produced, from each point of the output data sequence SD1 to SD4 of the A/D converters 13a and 13b at sampling timings STA1 to STA4. Information relating to shifting phases Δt1 to Δt4 are output as phase error information PSI from the clock phase error operator 18 shown in Fig. 9.

Fig. 11 is a block diagram showing a configuration of the clock phase error operator 18. The clock phase error operator 18 comprises a clock phase error extractor 111, a phase comparator (PD) 112, a loop filter (LPF) 113, and an oscillator (NCO) 114.

A phase error is extracted from input I/Q channel data through the clock phase error extractor 111. The extracted phase error is compared with output timing of the oscillator 114 by the phase comparator 112. With processing of the clock phase error extractor 111 and the phase comparator 112, the phase error Δt is calculated.

Then, the comparison result at the phase comparator 12 is integrated (smoothed out) by the loop filter 113, whereby the oscillator 114 is controlled on the basis of the integration result. That is, a PLL (Phase Locked Loop) is composed of the phase comparator 112, the loop filter 113, and the oscillator 114. Though the oscillator 114 outputs the timing of clock signal Fclk2, actually it will inform the interpolators 91a and 91b about the phase difference from the clock signal Fclk1 as phase error information.

In US Patent No. 5872818, detecting a symbol timing error from an output signal of a matching filter through a timing error detector and changing tap coefficients of the matching filter based on the detected timing error information are described.

SUMMARY OF THE INVENTION

An object of the present invention is to provide an apparatus for communications that can suitably control a band characteristic of interpolators on the basis of a receiving signal.

An apparatus for communications according to the present invention includes a front processor for outputting a digital formatted base band signal by quadrature demodulating a modulated signal, an interpolation processor for interpolating to adjust a phase error in accordance with sampling timing on the basis of the base band signal output from the front processor and for creating and outputting a base band signal synchronized with the sampling timing, and a back processor for wave-shaping the base band signal output from the interpolation processor and for outputting. The interpolation processor includes an FIR filter which controls tap coefficients supplying to the FIR filter on the basis of the modulated signal input therein. The invention is also concerned with corresponding methods.

According to the configuration as described above, it is possible to control a band characteristic of the FIR filter by supplying tap coefficients on the basis of the receiving modulated signal to the FIR filter.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of the invention, embodiments of it will now be described, by way of example, with reference to the accompanying drawings, in which: <ul id="ul0001" list-style="none"><li>Fig. 1 is a block diagram showing an example of a configuration of a demodulator according to a first embodiment;</li><li>Figs. 2A and 2B are diagrams showing an example of a configuration of an interpolator;</li><li>Figs. 3A and 3B are diagrams showing examples of changing bandwidth of the interpolator;</li><li>Fig. 4 is a block diagram showing an example of a configuration of a demodulator according to a second embodiment;</li><li>Fig. 5 is a block diagram showing an example of a configuration of a demodulator according to a third embodiment;</li><li>Fig. 6 is a block diagram showing a configuration of a controller according to the third embodiment;</li><li>Fig. 7 is a block diagram showing an example of a configuration of a demodulator according to a fourth embodiment;</li><li>Fig. 8 is a block diagram showing a configuration of a controller according to the fourth embodiment;</li><li>Fig. 9 is a block diagram showing a configuration of a demodulator in the prior art;</li><li>Fig. 10 is a diagram to explain an interpolating action through an interpolator;</li><li>Fig. 11 is a block diagram showing a configuration of a clock phase error operator;</li><li>Figs. 12A and 12B are diagrams showing a configuration of an interpolator within a demodulator in the prior art;</li><li>Fig. 13 is a diagram to explain a filter band of an interpolator;</li><li>Fig. 14 is a conceptual view of an interpolator within a demodulator in the prior art; and</li><li>Fig. 15 is a diagram to explain influence of an adjacent wave.</li></ul>

Each of the interpolators shown in Fig. 9 is composed of a FIR (finite impulse response) filter 121 shown in Fig. 12A and a coefficient operator (coefficient memory) 127 shown in Fig. 12B which outputs coefficients (tap coefficients) Ci (i is a subscript equal to an integer between minus 4 and plus 4) to a FIR filter 121.

As shown in Fig. 12A, the FIR filter 121 is composed of a group of registers 123 which is composed of a plurality of registers 122 in cascade connection and to which input signal data IDT are input; multipliers 124 each of which multiplies the output of each register 122 (or input signal data IDT) by each tap coefficient Ci; and an adder 125 which adds multiplying results of each multiplier 124 and outputs it as output signal data ODT. In Fig. 12A, a FIR filter 121 having nine taps is illustrated as an example.

The FIR filter 121, which holds input signal data IDT during a plurality of time, treats each item of data as an impulse, and changes the phase by controlling the tap coefficients Ci multiplied by each impulse. In other words, with appropriate control of the tap coefficients Ci, the FIR filter 121 can output a signal which suitably shifts the phase for the input signal data IDT as the output signal data ODT. These coefficients Ci can be calculated from the impulse response waveform.

That is, the FIR filter 121 has an impulse response as a filter limiting a band (for example, low-pass filter: LPF), and outputs a value derived from a timing position of clock signal Fclk1 to a timing position of clock signal Fclk2 of the impulse response, as coefficients Ci through the coefficient operator 127. This shows that the coefficients Ci of the FIR filter 121 change from moment to moment corresponding to the change of the phase error information PSI input to the interpolators.

The coefficient operator 127 outputs the tap coefficients Ci according to the phase error information PSI which is input. The coefficient operator 127 has previously stored a table including the tap coefficients Ci corresponding to the amount of the phase errors Δt, on the basis of the phase errors indicated by the phase error information PSI, and reads and outputs tap coefficients Ci corresponding to the phase errors. In addition, the coefficient operator 127 may be configured to calculate the tap coefficients Ci with calculation on an as-needed basis of the input phase error information PSI.

There follows a description of the operation of the interpolators (FIR filter 121) in a frequency domain, in other words, the operation as a filter. Fig. 13 is a diagram to explain a filter band of the interpolator. In Fig. 13, the interpolator operates as a filter limiting band, and herein operates as a low-pass filter (LPF).

In Fig. 13, F1 is a band of a base band signal of I channel or Q channel inputting to the interpolator, and BN1 is a band of the interpolator (pass band as filter). Further, Fclk1 and Fclk2 respectively illustrate the frequency of the sampling clock in the A/D converter and the clock frequency of the transmission side (actually desirable sampling timing).

The filter characteristic of the interpolator is preferably a conveyance characteristic without affecting the base band signal of the input I channel or Q channel, as shown in Fig. 13. The band BN1 of the interpolator generally as a wider band characteristic than that of the band F1 of the base band signal. With this operation, it is possible to process the interpolating operation.

Though the interpolator operates as a low-pass filter, it also operates as a modulator (modulating operation), because the tap coefficients Ci change from moment to moment as described above. Fig. 14 is a conceptual view to explain the modulating operation through the interpolator (FIR filter 121). In Fig. 14, blocks having functions being identical to those of the blocks shown in Fig. 9 are indicated the same numbers/characters, so that the description will not be repeated.

As shown in multipliers 141a, 141b in Fig. 14, the interpolators perform not only the operation as a low-pass filter but also the operation as the multiplying operation at the frequency (Fclk1 - Fclk2) for the base band signal of input I/Q channels. This operation is, as shown in Fig. 15, the same operation of modulating the input signal with the frequency (Fclk1 - Fclk2) in the frequency axis.

As shown in Fig. 15, it is desirable for the signal input to the interpolator to be only the base band signal (band F1) of the receiving I/Q channels. However, in many cases, a signal (band F2) of an adjacent channel remains. The signal of the adjacent channel means a signal being adjacent to the waveform of the desirable channel when a plurality of waveforms are arranged in the received highfrequency signal, because, in many cases, multi-channel signals are transmitted, such as in CATV, satellite communications, and multi-radio communications; such signals from the adjacent channels can become an interference source in the waveform of the desirable channel.

This signal of the adjacent channel is also modulated by the interpolator. Consequently, as shown in Fig. 15, signals modulated by the interpolator appear in bands F2A and F2B which are lower and upper frequencies of the signal band F2 of the adjacent channel. However, since a band limitation is also simultaneously effected by the operation as the low-pass filter actually, while the components of higher frequency (the components of the band F2 and F2B) are cut off from the right signal part of the adjacent channel being out of the pass band BN1, the components of the modulated signal (the components of the band F2A) relating to the adjacent channel within the pass band BN1 are not cut off, and are transferred to the latter stage without being processed, and are likely to become a noise source. The components transferred to the latter stage among those of the adjacent channel illustrate as the hatching area NS in Fig. 15.

It is considered that the interference owing to the signal components of this hatching area NS affects the base band signal of I/Q channels of the desirable receiving channel. In general, since the signal components of the hatched area NS are removed by roll-off filters (LPFs) 17a and 17b positioned at the latter stage shown in Fig. 9, it is considered that the signal components of the hatched area NS do not affect the desirable receiving channel finally. However, the interference components affect the dynamic range until the input terminals of the roll-off filters (LPFs) 17a and 17b, which causes a reduction in quality.

In particular, if the level of the signal relating to the adjacent channel is very large, the signal level will exceed the possible range to process owing to the noise of the hatching area NS. This will become a serious problem such that performing the process in the latter stage is impossible.

In order to solve the problem described above, a method will be considered which makes the band width BN1 of the interpolator narrow, of the same order of that of the roll-off filters 17a and 17b in the latter stage. However, when the band width BN1 is simply made narrow without study, the following problem occurs. At the time of performing a process in the interpolator, frequency shift components exist in the input signal owing to imperfect synchronization of the carrier with only coarse adjustment through the rotator (AFC) 14. If the frequency shift components are cut off through the interpolator by making the bandwidth BN1 narrow, a deterioration of the characteristic will occur.

Accordingly, the band width of the interpolator should be determined in an optimal manner, with consideration of the effect of the adjacent channel and the attenuation of the self-wave owing to carrier shift. Therefore, a demodulator applied an apparatus for communications according to embodiments of the present invention is a modulator which can suitably control the band width of the interpolator corresponding to an input modulated signal.

Initially, a first embodiment according to the present invention will be described.

Fig. 1 is a block diagram showing an example of a configuration of a demodulator 10 provided in the apparatus for communications representing the first embodiment of the present invention.

The demodulator 10 includes an RF (frequency converter) unit 11, a MIX (mixer) unit 12, A/D (analog-to-digital) converters 13a and 13b, a rotator (AFC: Auto Frequency Control) 14, interpolators 15a and 15b, a rotator (CR: Carrier Recovery) 16, low pass filters (LPFs) 17a and 17b, a clock phase error operator 18, and a decoder 19.

The RF unit 11 performs frequency conversion on a modulated signal input therein, and converts its frequency into a demodulatable frequency. For example, the RF unit 11 converts a demodulated frequency that is input with 1 to 2 GHz frequency into a frequency with around 500 MHz. The MIX unit 12 quadrature demodulates (separates) the modulated wave performed frequency conversion through the RF unit 11 into base band signals corresponding to the phase axes of I channel and Q channel.

The A/D converters 13a and 13b digitally convert analog formatted base band signals of I/Q channels output from the MIX unit 12 with sampling on the basis of clock signal Fclk1. In addition, the locations of the A/D converters 13a and 13b are not limited to this case, which may be arranged in different locations depending upon the processing method. For example, if they are arranged in front of the MIX unit 12, the quadrature demodulation within the MIX unit 12 will become a digital process. Further, the clock signal Fclk1 and the clock signal Fclk2 are asynchronous, the clock signal Fclk2 is synchronous with a clock signal of a transmission side.

The rotator (AFC) 14 and the rotator (CR) 16 make the carriers synchronous with digital data of the I/Q channels. The rotator 14 (AFC) roughly adjusts the frequency by coarse adjustment and the rotator (CR) 16 performs fine adjustment for perfect synchronization. If the deviation of the frequency is small, the performance through the coarse rotator (AFC) 14 may be omitted.

To the interpolators 15a and 15b, phase error information PSI is input from the clock phase error operator 18, and also reception rate information SRI is input from a receiving controller (not shown). The interpolators 15a and 15b are controlled on the basis of the phase error information PSI, and the reception rate information SRI, etc., and, in order to restore digital data from the demodulated base band signals of the I/Q channels, the interpolators 15a and 15b perform synchronism, which is generally called as clock synchronism, for the sampling timing. Concretely, the interpolators 15a and 15b, as described by using Fig. 10, adjust the phase of data on the basis of the data of the I/Q channels performed sampling by using the clock signal Fclkl, and create data through calculation at the sampling timing by using the clock signal Fclk2 synchronized with the clock signal of the transmission side.

The low-pass filters (LPFs) 17a and 17b, which are so called roll-off filters, perform wave-shaping for the base band signals of the I/Q channels where the carriers are synchronized by the rotator (CR) 16. The decoder 19 performs processing such as error correction and framing performed on digital data as necessary, and also decodes and outputs it as the data of the I/Q channels.

The clock phase error operator 18 acquires the phase difference between the clock signal Fclk1 and the clock signal Fclk2, in other words the difference between the actual sampling timing and the desirable sampling timing with calculation. The clock phase error operator 18, on the basis of the calculation result, outputs the phase difference between clock signal Fclk1 and clock signal Fclk2 as the phase error information PSI to the interpolators 15a and 15b. The configuration of the clock phase error operator 18 is the same as shown in Fig. 11 described above, so that detailed description of it will be omitted.

<b>Figs. 2A and 2B</b> are diagrams showing an example of the configuration of the interpolators 15a and 15b according to the present embodiment. Each of the interpolators 15a and 15b includes an FIR (finite impulse response) filter 21 as shown in Fig. 2A, and a coefficient operator 26 as shown in Fig. 2B.

The FIR filter 21, as shown in Fig. 2A, includes a group of registers 23 composed of a plurality of registers 22 in cascade connection, multipliers 24 each of which multiplies the output of each register 22 (or input signal data IDT) by each coefficient (tap coefficient) Ci (i is a subscript equal to an integer in the range minus 4 to plus 4), and an adder 25 which adds the multiplying results of each multiplier 24 and outputs them as output signal data ODT. The FIR filter 21 is here the same as the FIR filter 121 as shown in Fig. 12A, so that detailed description of it will be omitted. In addition, though Fig. 2A illustrates an example of the FIR filter 21 having nine taps, it is not limited to this, the number of taps being arbitrary.

The coefficient operator 26 shown in Fig. 2B includes coefficient memories 27-1 and 27-2, and a selector 28.

The coefficient memories 27-1 and 27-2 are correspondingly provided for each state which is indicated by control information CI, that is, one coefficient memory is provided corresponding to one state. Each of the coefficient memories 27-1 and 27-2 has previously stored in it a tap coefficient table in which tap coefficients Ci corresponding to the amount of a phase error are defined, and outputs tap coefficients Ci corresponding to a phase error indicated by input phase error information PSI to the selector 28.

The tap coefficients Ci are supplied from each of the coefficient memories 27-1 and 27-2 to the selector 28, which then selects the tap coefficients Ci from one of them on the basis of the control information CI, and outputs them to the FIR filter 21.

That is, the coefficient operator 26 according to the present embodiment shown in Fig. 2B selects a set of tap coefficients Ci based on the input phase error information PSI and control information CI, and supplies the tap coefficients Ci corresponding to the selected phase error information and control information to the FIR filter 21. In this first embodiment, the control information CI is reception rate information SRI which is input from a receiving controller, not shown.

In addition, though Fig. 2B illustrates one example of the coefficient operator 26 including two coefficient memories 27-1 and 27-2, it is not limited to this: an arbitrary number of coefficient memories may be provided corresponding to the control information, etc. Further, instead of the configuration providing the plurality of coefficient memories, a configuration in which a memory area of a coefficient memory is divided into plural areas can be used. Furthermore, it may be configured to calculate the tap coefficients Ci with calculation on an as-needed basis of the input phase error information, the control information, etc.

Next, the operation will be described.

A modulated wave which is input from a cable or an antenna, after its frequency is converted into a demodulatable frequency by the RF unit 11, is quadrature demodulated into base band signals of I/Q channels by the MIX unit 12. Subsequently, they are converted into the digital formatted base band signals of the I/Q channels by the A/D converters 13a and 13b.

The base band signals of the I/Q channels digitized by the A/D converters 13a and 13b are given a rough frequency adjustment in order to synchronize with the carrier by the rotator (AFC) 14. Then, clock synchronism is performed with the phase adjustment in the interpolators 15a and 15b. Subsequently, perfect synchronism is performed for the carrier through the fine adjustment of the frequency in the rotator (CR) 16. The phase adjustment in the interpolators 15a and 15b is performed by the FIR filter 21, the tap coefficients Ci being supplied according to the phase error information PSI and the reception rate information SRI. The phase error information PSI is calculated by the clock phase error operator 18 at any time as described above, and is output on the basis of the calculation result.

The base band signals of the I/Q channels, which are finally synchronized with the carrier in the rotator (CR) 16, are wave-shaped by the roll-off filters 17a and 17b, and are decoded by the decoder 19, and output.

In the first embodiment, symbol rate information (symbol frequency set signal) of a receiving channel is supplied to the interpolators 15a and 15b as the reception rate information SRI. The bandwidth (cut-off frequency) of the interpolators 15a and 15b is controlled by changing the tap coefficients Ci of the FIR filter 21 which are supplied by the coefficient operator 26 according to the symbol rate (symbol frequency) indicated by the symbol rate information.

Since the bandwidth of the transmitting signal changes with the symbol rate, changing the bandwidth of the interpolators 15a and 15b according to the symbol rate and changing that according to the bandwidth of the transmitting signal are the same. That is, in the first embodiment, the bandwidth of interpolators 15a and 15b is controlled and changed according to the bandwidth of the transmitting signal.

In addition, in the case of normal reception, the symbol rate (the bandwidth of the transmitting signal) of a receiving channel is previously known, and this information is included in the receiving controller, not shown, etc. Accordingly, the symbol rate information supplied to the interpolators 15a and 15b can be created and output on the basis of the information included in the receiving controller, etc.

As described above, the bandwidth of the interpolators 15a and 15b preferably has a conveying characteristic without affecting the received signal, and is generally a wider band than that of the received signal. Thereby, as an example described in Fig. 3A, the bandwidth of the interpolators 15a and 15b is wider by a constant or given frequency amount (e.g. 5MHz) than the symbol rate (signal bandwidth).

Fig. 3A is a diagram showing a controlling example of the bandwidth of the interpolators 15a and 15b according to the present embodiment. In Fig. 3A, each of WF1, WF2 and WF3 indicates the band where each symbol rate is SR1, SR2, and SR3; and each of BF1, BF2, and BF3 indicates the bandwidth of interpolators 15a and 15b corresponding to each receiving signal band WF1, WF2, and WF3.

As shown in Fig. 3A, by controlling the bandwidth of the interpolators 15a and 15b according to the amount of the symbol rate i.e., the band of the receiving signal, for example, the bandwidth of the interpolators can be set to the bandwidth which is slightly wider than the band of the receiving signal and is hardly affected by the carrier shift, corresponding to the receiving signal. In addition, as shown in Fig. 3A, the tap coefficients Ci, by which each of the bandwidths BF1, BF2, and BF3 of the interpolators 15a and 15b is set, are previously stored in the coefficient memories included in the coefficient operator 26 of the interpolators 15a and 15b, so that they may appropriately be selected according to each symbol rate SR1, SR2, and SR3 indicated by the reception rate information SRI.

In the example above, the bandwidth of the interpolators 15a and 15b is set to be wider by the constant number of frequencies (frequency amount) than the symbol rate (signal bandwidth). In the example shown in <b>Fig. 3B,</b> it can be wider by a constant ratio (e.g. 1.3 times) in comparison to the symbol rate (signal bandwidth). In this case also, symbol rate information (symbol frequency set signal) of a receiving channel may be supplied to the interpolators 15a and 15b as receive rate information; thus the bandwidth (cut-off frequency) of the interpolators 15a and 15b may be controlled by changing the tap coefficients Ci of the FIR filter 21 which are supplied by the coefficient operator 26.

Fig. 3B is a diagram showing another controlling example of the bandwidth of the interpolators 15a and 15b according to the present embodiment. In Fig. 3B, each of WF1, WF2 and WF3 indicates the signal band where each symbol rate is SR1, SR2, and SR3; and each of BF11, BF12, and BF13 indicates the bandwidth of the interpolators 15a and 15b corresponding to each receiving signal band WF1, WF2, and WF3.

In this way, if the bandwidth of the interpolators 15a and 15b is set to be wider by k times (k is an arbitrary value) in comparison to the symbol rate (signal bandwidth), the bandwidth, for example, can be of higher accuracy and have a better characteristic than being wider by the constant number of frequencies in the case where the dynamic range of the symbol rate is wide.

As described above, in accordance with the first type of embodiment, the bandwidth of the interpolators 15a and 15b (FIR filter 21) can be controlled to a suitable bandwidth corresponding to the bandwidth of the receiving signal, by changing the tap coefficients which are supplied to the FIR filter 21 of the interpolators 15a and 15b according to the symbol rate information (bandwidth of receiving signal) of the receiving channel.

Next, a second embodiment according to the present invention will be described.

Fig. 4 is a block diagram showing an example of a configuration of a demodulator 40 according to the second embodiment. In Fig. 4, the blocks having functions being identical to those of the blocks shown in Fig. 1, are indicated by the same numbers/characters, so that duplicated description will be omitted. In Fig. 4, interpolators 41a and 41b correspond to the interpolators 15a and 15b shown in Fig. 1, and are configured as well as those in Figs. 2A and 2B.

The demodulator 40 according to the second embodiment shown in Fig. 4 is different from the demodulator 10 according to the first embodiment shown in Fig. 1 with respect to the information input to the interpolators. In the second embodiment, multi-value information QAMI (QAM set information) is input from a receiving controller (not shown) to the interpolators 41a and 41b of the demodulator 40, instead of the reception rate information SRI.

The multi-value information QAMI becomes the control information CI for the interpolators 41a and 41b in the second embodiment. That is, in the second embodiment, the bandwidth (cut-off frequency) of the interpolators 41a and 41b is changed according to a multi-value relating to the quadrature amplitude modulation.

In this, the multi-value relating to the quadrature amplitude modulation is one of the parameters to which error occurrence by causing the carrier shift or signal cut-off affects. In the quadrature amplitude modulation, distortion and/or noise strength differs depending upon the multi-value, while as for the cut-off distortion of the self-wave with the carrier shift, its strength also differs depending upon the multi-value. As the multi-value is greater, the strength against the influence of noise, etc. from the adjacent channel decreases; for example, even if it sufficiently endures against the multi-value 4 of 4PSK, there is a possibility that it does not endure at all against the multi-value 256 of 256QAM.

For this reason, in the second embodiment, the bandwidth (cut-off frequency) of the interpolators 41a and 41b is controlled by changing the tap coefficients Ci supplied to the FIR filter according to the phase error information PSI and the multi-value information QAMI. The tap coefficients Ci supplied to the FIR filter are stored into each coefficient memory provided corresponding to each multi-value, and the tap coefficients Ci corresponding to the multi-value indicated by the multi-value information QAMI are supplied to the FIR filter, appropriately exchanging the selector on the basis of the multi-value information QAMI. In addition, as described above, since, as the multi-value is greater, the strength decreases, as the multi-value is greater, the bandwidth (cut-off frequency) of the interpolators 41a and 41b is controlled to be narrower. That is, the bandwidth of the interpolators 41a and 41b is changed according to the multi-value, in accordance with the relation that the bandwidth of the interpolators 41a and 41b is inversely proportional to the multi-value.

According to the second embodiment, the bandwidth of the interpolators 41a and 41b can be suitably controlled in a way corresponding to the multi-value of the receiving signal, by changing the tap coefficients Ci which are supplied to the FIR filter of the interpolators 41a and 41b according to the multi-value of the modulation system of the receiving signal.

Next, a third embodiment according to the present invention will be described.

Fig. 5 is a block diagram showing an example of a configuration of a demodulator 50 according to the third embodiment. In Fig. 5, the blocks having functions identical to those of the blocks shown in Fig. 1 are indicated by the same numbers/characters, so that overlapping description will be omitted.

In Fig. 5, interpolators 51a and 51b correspond to the interpolators 15a and 15b shown in Fig. 1, and are configured as well as those in Figs. 2A and 2B. The numeral 52a is a comparator (subtracter) where the base band signal input to the interpolator 51a and the base band signal output from the interpolator 51a are input, then the difference between them is output. Similarly with the comparator (subtracter) 52b, the base band signal input to the interpolator 51b and the base band signal output from the interpolator 51b are input, then the difference of them is output. The outputs from the comparators 52a and 52b are input to a controller 53, which outputs control information CI on the basis of them to the interpolators 51a and 51b.

In this, if the influence of the signal of the adjacent channel on the signal of the receiving channel is small, the influence of the self-wave reduction with the carrier shift can be decreased by making the bandwidth (cut-off frequency) of the interpolators wide. For this reason, the demodulator 50 shown in Fig. 5 according to the third embodiment changes the bandwidth (cut-off frequency) of the interpolators 51a and 51b corresponding to the influence of the adjacent channel.

However, it is very difficult to directly measure the signal strength of the adjacent channel. In the demodulator 50 according to the third embodiment, the influence of the adjacent channel is detected by the difference between the front and back signal levels of the interpolators 51a and 51b. That is, by detecting the difference between the front and back signal levels (input and output) of the interpolators 51a and 51b with the comparators 52a and 52b, the noise component with the signal of the adjacent channel is detected. Then, on the basis of the detected result, the control information CI is output from the controller 53, and the bandwidth (cut-off frequency) of the interpolators 51a and 51b is controlled. In this way, the bandwidth (cut-off frequency) of the interpolators 51a and 51b is controlled by utilizing the difference between the front and back signal levels of the interpolators 51a and 51b.

Concretely, if the component of the adjacent channels is cutoff by the interpolators 51a and 51b, the difference between the front and back signal levels of these is little, so that it is possible to make the bandwidth (cut-off frequency) of the interpolators 51a and 51b wide. If on the contrary the difference between the front and back signal levels of the interpolators 51a and 51b is large, it is considered that the component of the adjacent channels is not cutoff, so that the control is changed to make the bandwidth (cut-off frequency) of the interpolators 51a and 51b narrow.

Fig. 6 is a block diagram showing a configuration of the controller 53 shown in Fig. 5.

In Fig. 6, the numeral 61 is an amplitude comparator which calculates the difference between the front and back signal levels of the interpolators 51a and 51b, and which is composed of the comparators 52a and 52b. The numeral 62 is a loop filter (LPF) which accumulates (integrates) the outputs (comparison results of the signal levels) of the amplitude comparator 61. The numeral 63 is an evaluation unit which compares the output from the loop filter 62 with a previously set threshold, and outputs control information CI according to the comparison result. The number of the previously set threshold is arbitrary; if a plurality of thresholds are set, it is possible to finely control the bandwidth (cut-off frequency) of the interpolators 51a and 51b.

Next, a fourth embodiment according to the present invention will be described.

Fig. 7 is a block diagram showing an example of a configuration of a demodulator 70 according to the fourth embodiment. In Fig. 7, the blocks having functions identical to those of the blocks shown in Fig. 1, are indicated by the same numbers/characters, so that overlapping description will be omitted.

In Fig. 7, interpolators 71a and 71b correspond to the interpolators 15a, 15b shown in Fig. 1, and are configured as well as those in Figs. 2A and 2B. The numeral 72 is a controller which outputs control information CI on the basis of signals output from the low-pass filters 17a and 17b.

Fig. 8 is a block diagram showing a configuration of the controller 72 shown in Fig. 7.

In Fig. 8, the numeral 81 is an error calculator which calculates a distance between a signal point of a signal output from the low-pass filters 17a and 17b and a basic signal point previously determined according to the modulation system, i.e. calculates an error. The numeral 82 is a loop filter (LPF) which accumulates (integrates) outputs from the error calculator 81, and the numeral 83 is an evaluation unit which outputs control information CI according to the output from the loop filter 82.

That is, the controller 72 according to the fourth embodiment observes a constellation of the output signal from the low-pass filters 17a and 17b, and outputs control information CI according to its result, and controls the bandwidth (cut-off frequency) of the interpolators 71a and 71b. Concretely, if the signal of the adjacent channel affects the signal of the receiving channel, since the error of the demodulating signal, which is calculated by the error calculator 81 where the constellation is inflated, becomes large, the controller 72 changes the bandwidth (cut-off frequency) of the interpolators 71a and 71b narrowly.

In addition, though the first to fourth embodiments are described individually, these may coordinate or be combined as desired in whole or part, so that the same effect can be obtained from the coordinating as that from each embodiment.

Further, the foregoing embodiments are only one example for performing the present invention concretely, and should not be intended to limit the technological scope. That is, the present invention can be provided as various alternative embodiments without departing from the technology spirit or main feature thereof.

According to the present invention, by changing the tap coefficients supplied to the FIR filter included in the interpolation processor according to the receiving signal, the filter band of the FIR filter can be set with a wider bandwidth than that of the receiving signal and with hard effect from the carrier shift in each receiving channel, and the filter bandwidth characteristic of the FIR filter can appropriately be controlled according to the receiving signal.