Diversity reception apparatus

Abstract

A demodulator 1 obtains demodulated data in a plurality of channels. Estimating portions 2a and 2b estimate and output the numbers of erroneous symbols and error locations thereof in the demodulated data. A data comparator 3 compares the demodulated data corresponding to the error locations with the demodulated data in the corresponding locations in other channels to determine whether the error location is correct, and it outputs a decision signal in response to the determination. A data selector 4 selects one of the demodulated data in the plurality of channels on the basis of the numbers of erroneous symbols and the decision signals and outputs the data as selected data. It is then possible to maintain the reliability of error detection even when a less redundant short error detecting code is used, and also to accurately select a channel of good quality even when the demodulated data in all channels contain the same extent of errors.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to a diversity reception apparatus which, in digital radio communications and the like, takes out demodulated data in two or more channels from one or a plurality of modulated signals and selects and combines the data to reduce transmission errors.

2. Description of the Background Art

In radio communications, especially in mobile communication, the diversity reception is widely used to reduce transmission errors caused by fading and interference.

In a typical method of the diversity reception, a signal at the highest reception signal strength is selected from among a plurality of channels of signals received at a plurality of antennas. However, in the presence of interference or multipath, error may occur even if the reception signal strength is high. Then a channel of signal of good quality cannot be correctly selected only with the reception signal strength. Conventionally, in order to solve this problem, Japanese Patent Laid-Open No. 1-265739 and Japanese Patent Laid-Open No. 4-8031 propose diversity reception apparatus which can detect errors in received data to select a channel containing the smallest number of errors. Such conventional diversity reception apparatus will now be described referring to the drawing.

FIG. 17 is a block diagram showing the structure of a conventional diversity reception apparatus. As shown in FIG. 17 , this diversity reception apparatus has antennas 101 a and 101 b, receivers 102 a and 102 b, error detectors 103 a and 103 b, an error comparator 104 , and a data selector 105 . The receivers 102 a and 102 b demodulate signals received at the antennas 101 a and 101 b to output demodulated data d 100 a and d 100 b, respectively. The error detectors 103 a and 103 b respectively detect errors in the demodulated data d 100 a and d 100 b and count the numbers of errors. The error comparator 104 receives and compares the numbers of errors outputted from the error detectors 103 a and 103 b and outputs a selection signal s 100 for selecting one of the demodulated data d 100 a and d 100 b containing less errors. The data selector 105 selects the one with less errors from the demodulated data d 100 a and d 100 b on the basis of the selection signal s 100 from the error comparator 104 and outputs it as a selected signal d 101 . With this operation, the diversity reception apparatus shown in FIG. 17 can select demodulated data with less errors from the signals received at the two channels of antennas.

However, the diversity reception apparatus with the above-described structure requires use of a highly redundant error detecting code or a long error detecting code to correctly detect errors. The use of a long error detecting code reduces the selecting rate and then it is impossible to follow a rapid change on the transmission path caused by fast fading, for example. However, the use of a highly redundant error detecting code reduces the data transmission efficiency. Further, if error correcting coding is used to improve the reliability, the redundancy further increases. Moreover, if the demodulated data in all channels include errors to the same extent, it is impossible to determine which channel of demodulated data has the best quality.

SUMMARY OF THE INVENTION

Accordingly, an object of the present invention is to provide a diversity reception apparatus which performs error estimation by using-an error correcting code so as to maintain the reliability of the error estimation even when a less redundant and short code is used, and to accurately compare qualities of demodulated data even when demodulated data in all channels contain errors to a similar extent, so that a channel of good quality can accurately be selected. Another object of the present invention is to improve the reliability by error correction without increasing the redundancy, together with diversity selection.

To achieve the objects above, the present invention has the features shown below.

A first aspect of the present invention is directed to a diversity reception apparatus receiving one or a plurality of modulated signals modulated with error-correctably coded data. According to the first aspect, the diversity reception apparatus comprises:

a demodulator for demodulating the modulated signal(s) to obtain demodulated data in a plurality of channels corresponding to the coded data;

an error estimating portion for estimating the number of erroneous symbols and error locations thereof in the demodulated data in each channel;

a data comparator for comparing the data in the error locations estimated in the demodulated data in each channel with the data in the corresponding locations in the demodulated data in other channels to determine whether the error locations are correct or not; and

a data selector for selecting one of the plurality of channels on the basis of the numbers of erroneous symbols and results of the determination made by the data comparator.

In accordance with the above-described first aspect, the demodulated data are compared among channels to determine whether the error locations are correct. Accordingly, the reliability of the error estimation can be maintained even when a less redundant and short code is used, and a channel of good quality can be selected even when the demodulated data in all channels contain the same extent of errors.

According to a second aspect of the present invention, in the diversity reception apparatus of the first aspect,

when all data in the error locations estimated in the demodulated data in one channel of the plurality of channels respectively differ from the data in the corresponding locations in the demodulated data in all other channels, the data comparator determines that the error locations in the one channel are correct.

In accordance with the above-described second aspect, it is possible to determine whether the estimated error locations in each channel are correct on the basis of data comparison with simple structure.

According to a third aspect of the present invention, in the diversity reception apparatus of the first aspect,

when all data in the error locations estimated in the demodulated data in one channel of the plurality of channels respectively differ from the data in the corresponding locations in the demodulated data in at least one of other channels, the data comparator determines that the error locations in the one channel are correct.

In accordance with the above-described third aspect, it is possible to determine whether the estimated error locations in each channel are correct by data comparison with simple structure.

According to a fourth aspect of the present invention, in the diversity reception apparatus of the first aspect,

when data whose number is equal to or larger than a predetermined threshold in the data in the error locations estimated in the demodulated data in one channel of the plurality of channels differ from the data in the corresponding locations in the demodulated data in at least one of other channels, the data comparator determines that the error locations in the one channel are correct.

In accordance with the above-described fourth aspect, the determination by data comparison about the correctness of the error estimation can be made more flexibly by setting the threshold value.

According to a fifth aspect of the present invention, in the diversity reception apparatus of the first aspect,

the error estimating portion estimates the error locations in the demodulated data in each channel bit by bit, and

when all bits in the error locations estimated in the demodulated data in one channel of the plurality of channels respectively differ from the bits in the corresponding locations in the demodulated data in at least one of other channels, the data comparator determines that the error locations in the one channel are correct.

According to a sixth aspect of the present invention, in the diversity reception apparatus of the first aspect,

the error estimating portion estimates the error locations in the demodulated data in each channel bit by bit, and

when bits whose number is equal to or larger than a predetermined threshold in the bits in the error locations estimated in the demodulated data in one channel of the plurality of channels differ from the bits in the corresponding locations in the demodulated data in at least one of other channels, the data comparator determines that the error locations in the one channel are correct.

According to a seventh aspect of the present invention, the diversity reception apparatus of the fifth aspect further comprises a bit inverter for inverting the bits in the error locations in the demodulated data in the channel selected by the data selector when the number of erroneous symbols is equal to or smaller than a given value.

In accordance with the above-described seventh aspect, the bit inversion is performed to correct errors on the basis of the estimated result about the number of erroneous symbols and the error locations thereof which were used for selection. Accordingly transmission errors can be reduced by effects of both of diversity selection and error correction with low redundancy and simple structure.

According to an eighth aspect of the present invention, in the diversity reception apparatus of the seventh aspect,

the bit inverter does not perform the bit inversion when the number of erroneous symbols exceeds the given value.

In accordance with the above-described eighth aspect, the bit inversion is not performed when the number of erroneous symbols exceeds a given value, which prevents an increase of errors due to miscorrection.

According to a ninth aspect of the present invention, in the diversity reception apparatus of the first aspect,

the data selector selects a channel in which the number of erroneous symbols is the smallest.

In accordance with the above-described ninth aspect, more reliable selection can be made than in diversity selection using error detection.

According to a tenth aspect of the present invention, in the diversity reception apparatus of the first aspect,

when the numbers of erroneous symbols are the smallest in a plurality of the channels, the data selector selects a channel in which the error locations are determined to be correct by the data comparator from among the plurality of the channels in which the numbers of erroneous symbols are the smallest.

In accordance with the above-described tenth aspect, a more reliable channel can be selected when a plurality of channels contain the smallest number of erroneous symbols.

According to an eleventh aspect of the present invention, in the diversity reception apparatus of the first aspect,

when the numbers of erroneous symbols are equal in all channels, the data selector selects a channel in which the error locations are determined to be correct by the data comparator.

In accordance with the above-described eleventh aspect, the selection can be made reliably even when there is no difference among the numbers of erroneous symbols.

According to a twelfth aspect of the present invention, in the diversity reception apparatus of the first aspect,

the number of the channels of the demodulated data is two, and

when the numbers of erroneous symbols are equal in the two channels and the error locations in the two channels are both determined to be correct by the data comparator, the data selector maintains the preceding selection.

In accordance with the above-described twelfth aspect, the selection made in the previous time is maintained when it is impossible to determine which channel has the best quality. A reliable selection can be made by using the previous information when an error is occurring due to a factor having continuity in time such as fading and continual noise.

According to a thirteenth aspect of the present invention, in the diversity reception apparatus of the first aspect,

the number of the channels of the demodulated data is two, and

when the numbers of erroneous symbols are equal in the two channels and the error locations in the two channels are both determined to be incorrect by the data comparator, the data selector maintains the preceding selection.

In accordance with the above-described thirteenth aspect, like in the twelfth aspect, the selection made in the previous time is maintained when a determination cannot be made as to which channel has the best quality. Then it is possible to make reliable selection by using the previous information when an error is occurring due to a factor having continuity in time such as fading and continual noise.

According to a fourteenth aspect of the present invention, in the diversity reception apparatus of the first aspect,

when the data comparator determines that the error locations in a plurality of the channels are correct, the data selector selects a channel in which the number of erroneous symbols is the smallest from among the plurality of the channels in which the error locations are determined to be correct.

In accordance with the above-described fourteenth aspect, more reliable selection can be made when the estimated error locations are determined to be correct in a plurality of channels.

According to a fifteenth aspect of the present invention, in the diversity reception apparatus of the first aspect,

when the data comparator determines that all of the error locations in each of the channels are not correct, the data selector selects a channel in which the number of erroneous symbols is the smallest.

In accordance with the above-described fifteenth aspect, a reliable channel can be selected even when the error locations estimated in all of the channels are incorrect.

According to a sixteenth aspect of the present invention, in the diversity reception apparatus of the first aspect,

the error estimating portion estimates the number of erroneous symbols and error locations thereof for each block defined by dividing the demodulated data in each channel to given length, and

the data selector selects one of the plurality of channels for each block.

In accordance with the above-described sixteenth aspect, the selection is made block by block. Accordingly, even if an error remains, the error affects only within the block, without affecting other blocks.

According to a seventeenth aspect of the present invention, the diversity reception apparatus of the first aspect further comprises a data detector for determining whether the demodulated data in each of the plurality of channels is valid or invalid, and

the data selector selects one of the channels determined to be valid by the data detector from among the plurality of channels.

In accordance with the above-described seventeenth aspect, a determination is made as to whether data is valid. This prevents the malfunction of accidentally determining received noise as error-free data.

According to an eighteenth aspect of the present invention, in the diversity reception apparatus of the seventeenth aspect,

the data detector is a unique word detector for detecting particular data pattern in the demodulated data.

In accordance with the above-described eighteenth aspect, it is determined whether demodulated data in a plurality of channel are valid or invalid by using particular data pattern. Accordingly, similarly to the seventeenth aspect, it is possible to prevent the malfunction of accidentally determining received noise to be error-free data.

According to a nineteenth aspect of the present invention, in the diversity reception apparatus of the first aspect,

the modulated signal is a chirp PSK signal obtained by multiplying the phase-shift keying signal by a chirp signal whose frequency is swept at the same period as the symbol period, and

the demodulator includes a subband filter for extracting part of the band of the modulated signal and a differential detector for differentially detecting an output of the subband filter.

In accordance with the above-described nineteenth aspect, it is possible to perform diversity reception in which a plurality of channels of demodulated data are obtained from the same modulated signal received at the same antenna.

A twentieth aspect of the present invention is directed to a diversity reception apparatus receiving one or a plurality of modulated signals modulated with error correctably coded data. According to the twentieth aspect, the diversity reception apparatus comprises:

a demodulator for demodulating the modulated signal(s) to obtain demodulated data in a plurality of channels corresponding to the coded data;

an error correcting portion for estimating the number of erroneous symbols and error locations thereof in the demodulated data in each channel, and for correcting an error in the demodulated data in each channel on the basis of the error locations and outputting the error-corrected demodulated data as decoded data;

a data comparator for comparing the data in the error locations estimated in the decoded data in each channel with the data in the corresponding locations in the decoded data in other channels to determine whether the error locations are correct or not; and

a data selector for selecting one of the plurality of channels on the basis of the numbers of erroneous symbols and results of the determination made by the data comparator.

In accordance with the above-described twentieth aspect, decoded data is obtained by performing error correction at the same time as error estimation, and whether the error locations are correct is determined by comparing the decoded data among channels. Accordingly, even when a less redundant and short code is used, the reliability of the error estimation can be maintained, and even if all channels of demodulated data contain the same extent of errors, a channel of good quality can be selected.

According to a twenty-first aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

when all data in the error locations estimated in the decoded data in one channel of the plurality of channels are respectively equal to the data in the corresponding locations in the decoded data in all other channels, the data comparator determines that the error locations in the one channel are correct.

According to a twenty-second aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

when all data in the error locations estimated in the decoded data in one channel of the plurality of channels are respectively equal to the data in the corresponding locations in the decoded data in at least one of other channels, the data comparator determines that the error locations in the one channel are correct.

According to a twenty-third aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

when data whose number is equal to or larger than a predetermined threshold in the data in the error locations estimated in the decoded data in one channel of the plurality of channels are equal to the data in the corresponding locations in the decoded data in at least one of other channels, the data comparator determines that the error locations in the one channel are correct.

According to a twenty-fourth aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

the error correcting portion does not perform the error correction when the number of erroneous symbols exceeds a given value.

According to a twenty-fifth aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

the data selector selects a channel in which the number of erroneous symbols is the smallest.

According to a twenty-sixth aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

when the numbers of erroneous symbols are the smallest in a plurality of the channels, the data selector selects a channel in which the error locations are determined to be correct by the data comparator from among the plurality of the channels in which the numbers of erroneous symbols are the smallest.

According to a twenty-seventh aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

when the numbers of erroneous symbols are equal in all channels, the data selector selects a channel in which the error locations are determined to be correct by the data comparator.

According to a twenty-eighth aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

the number of the channels of the demodulated data is two, and

when the numbers of erroneous symbols are equal in the two channels and the error locations in the two channels are both determined to be correct by the data comparator, the data selector maintains the preceding selection.

According to a twenty-ninth aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

the number of the channels of the demodulated data is two, and

when the numbers of erroneous symbols are equal in the two channels and the error locations in the two channels are both determined to be incorrect by the data comparator, the data selector maintains the preceding selection.

According to a thirtieth aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

when the data comparator determines that the error locations in a plurality of the channels are correct, the data selector selects a channel in which the number of erroneous symbols is the smallest from among the plurality of the channels in which the error locations are determined to be correct.

According to a thirty-first aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

when the data comparator determines that the error locations are incorrect in all channels, the data selector selects a channel in which the number of erroneous symbols is the smallest.

According to a thirty-second aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

the error correcting portion estimates the number of erroneous symbols and error locations thereof for each block defined by dividing the demodulated data in each channel to given length, and

the data selector selects one of the plurality of channels for each block.

According to a thirty-third aspect of the present invention, the diversity reception apparatus of the twentieth aspect further comprises a data detector for determining whether the demodulated data in each of the plurality of channels is valid or invalid, and

the data selector selects one of the channels determined to be valid by the data detector from among the plurality of channels.

According to a thirty-fourth aspect of the present invention, in the diversity reception apparatus of the thirty-third aspect,

the data detector is a unique word detector for detecting particular data pattern in the demodulated data.

According to a thirty-fifth aspect of the present invention, in the diversity reception apparatus of the twentieth aspect,

the modulated signal is a chirp PSK signal obtained by multiplying a phase-shift keying signal by a chirp signal whose frequency is swept at the same period as the symbol period, and

the demodulator includes a subband filter for extracting part of the band of the modulated signal and a differential detector for differentially detecting an output of the subband filter.

A thirty-sixth aspect of the present invention is directed to a diversity reception apparatus receiving one or a plurality of modulated signals modulated with error correctably coded data. According to the thirty-sixth aspect, the diversity reception apparatus comprises:

a demodulator for demodulating the modulated signal(s) to obtain demodulated data in a plurality of channels corresponding to the coded data;

an error estimating portion for estimating an error location in the demodulated data in each channel;

a data comparator for comparing the data in the error location estimated in the demodulated data in each channel with the data in the corresponding location in the demodulated data in other channels to determine whether the error location is correct or not; and

a data selector for selecting a channel in which the error location is determined to be correct by the data comparator from among the plurality of channels.

In accordance with the above-described thirty-sixth aspect, demodulated data are compared among channels to determine whether the error locations are correct. Accordingly, the reliability of the error estimation can be maintained even when a less redundant and short code is used, which enables reliable selection of channels.

These and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram showing the structure of a diversity reception apparatus according to a first embodiment of the present invention.

FIG. 2 is a diagram showing the data structure of data for modulating a carrier into a modulated signal in the first embodiment.

FIG. 3 is a block diagram showing a structure of the data comparator in the first embodiment.

FIG. 4 is a diagram showing the contents of the data memory portions shown in FIG. 3 in the first embodiment.

FIG. 5 is a timing chart showing operation made by the decision address generator shown in FIG. 3 in the first embodiment.

FIG. 6 is a block diagram showing a structure of the data comparator which is used when three or more channels of demodulated data are used in the first embodiment.

FIG. 7 is a block diagram showing another structure of the data comparator in the first embodiment.

FIG. 8 is a block diagram showing the structure of the data selector in the first embodiment.

FIG. 9 is a flowchart showing an example of the decision procedure of the decision logic operation portion in the diversity reception apparatus of the first embodiment.

FIG. 10 is a flowchart showing another example of the decision procedure of the decision logic operation portion in the diversity reception apparatus of the first embodiment.

FIG. 11 is a block diagram showing the structure of the bit inverter in the first embodiment.

FIG. 12 is a block diagram showing the structure of a diversity reception apparatus according to a second embodiment of the present invention.

FIG. 13 is a block diagram showing the structure of a diversity reception apparatus according to a third embodiment of the present invention.

FIG. 14 is a block diagram showing a structure of the data comparator in the third embodiment.

FIG. 15 is a block diagram showing the structure of the data comparator which is used when three or more channels of demodulated data are used in the third embodiment.

FIG. 16 is a block diagram showing the structure of the data selector in the third embodiment.

FIG. 17 is a block diagram showing the structure of a conventional diversity reception apparatus.

FIG. 18 is a diagram showing results obtained by computer simulation about the selection error probability in the diversity reception apparatus of the first embodiment.

FIG. 19 is a diagram showing results obtained by computer simulation and actual measurements about the bit error rate characteristic in error-correcting and combining in the first embodiment and CRC selecting and combining under static environment without multipath.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

(First Embodiment)

FIG. 1 is a block diagram showing a diversity reception apparatus according to a first embodiment of the present invention. As shown in FIG. 1 , this diversity reception apparatus has a demodulator 1 , error estimating portions 2 a and 2 b, a data comparator 3 , a data selector 4 , a bit inverter 5 , and a data detector 8 . The demodulator 1 receives a modulated signal s 0 as input. The modulated signal s 0 is a chirp PSK signal obtained by multiplying a phase-shift keying (PSK) signal by a chirp signal whose frequency is swept at the same period as the symbol period, which is the same as the modulated signal disclosed in FIG. 12 in U.S. Pat. No. 5,504,774, which is incorporated herein by reference. The demodulator 1 operates similarly to the demodulator in the receiving apparatus disclosed in U.S. Pat. No. 5,504,774, which is a subband demodulator including at least two channels of subband filters and at least two channels of differential detectors. In the description hereinafter, it is assumed that the demodulator 1 includes two channels of subband filters 1 fa and 1 fb and two channels of differential detectors 1 da and 1 db.

The subband filters 1 fa and 1 fb both receive the modulated signal s 0 as their inputs, extract signals in different subbands, and output them as subband signals sba and sbb, respectively. The differential detectors 1 da and 1 db respectively receive the subband signals sba and sbb and perform differential detection and data demodulation to output first-channel demodulated data d 1 a and second-channel demodulated data d 1 b, respectively. The data for modulating a carrier into the modulated signal is coded by BCH coding in advance, as will be described later.

The error estimating portion 2 a receives the demodulated data d 1 a, and estimates the number of erroneous symbols and the error locations in the demodulated data d 1 a by using properties of the BCH code. Then, the error estimating portion 2 a outputs an erroneous symbol number e 1 a and error locations e 2 a. Each of the error locations e 2 a represents the position of an estimated erroneous symbol in the demodulated data d 1 a. Similarly, the error estimating portion 2 b receives the demodulated data d 1 b, estimates the number of erroneous symbols and the error locations in the demodulated data d 1 b, and then outputs thus estimated erroneous symbol number e 1 b and error locations e 2 b.

The data detector 8 detects unique words described later for the demodulated data d 1 a and d 1 b to determine whether they include valid data and outputs the detection result d 8 . The detection result d 8 is a signal indicating three states: a state in which both of first and second channels are valid, a state in which only the first channel is valid, and a state in which only the second channel is valid.

The data comparator 3 compares data corresponding to the error locations e 2 a in the demodulated data d 1 a with data in the corresponding locations in the demodulated data d 1 b to output a decision signal s 1 a, and also compares data corresponding to the error locations e 2 b in the demodulated data d 1 b with data in the corresponding locations in the demodulated data d 1 a to output a decision signal s 1 b.

When the detection result d 8 outputted from the data detector 8 indicates the state in which the first and second channels are both valid, the data selector 4 selects one of the demodulated data d 1 a and d 1 b on the basis of the erroneous symbol numbers e 1 a and e 1 b and the decision signals s 1 a and s 1 b, and then outputs it as selected data d 2 . It also selects the erroneous symbol number and error locations of the same channel as the selected demodulated data and outputs them as a selected erroneous symbol number e 3 and selected error locations e 4 , respectively. It also outputs address “ra” described later. When the detection result d 8 outputted from the data detector 8 indicates a state in which only one of the channels is valid, the channel which contains valid data is selected, independently of the erroneous symbol numbers and error locations. The bit inverter 5 inverts erroneous bits in the selected data d 2 on the basis of the selected erroneous symbol number e 3 , selected error locations e 4 , and address ra, to correct errors in the selected data d 2 and then it outputs decoded data d 3 .

The data for modulating a carrier into the modulated signal is structured on the basis of the unit frame structure shown in FIG. 2 . That is to say, this frame has a unique word 10 for detection of the head of valid data at its beginning, which is followed by j blocks 11 , 12 , . . . , 1 j each formed of n-bit data. Each of the j blocks is a binary BCH code capable of 2-bit error correction, which includes k bits of information data and 2 m bits of redundant data. The numbers n, k and m are integers in the relation of n=k+2 m. It is of course possible to use a code capable of correction of 3 or more bits as the error correcting code.

The error estimating portions 2 a and 2 b shown in FIG. 1 each perform syndrome calculation of the BCH code and estimate the numbers of erroneous symbols and the error locations from the calculated results, which can be realized with the same structure as part of the BCH error correcting device shown in U.S. Pat. No. 5,216,676, for example. As is well known, the 2-bit error correcting BCH code can detect four kinds of error states: no error, 1-bit error, 2-bit error, and 3-or more-bit error, from its properties. In the case of an error of 2 bits or less, it is possible to calculate the locations of the erroneous bits. However, when an error of 3 or more bits occurs, it may not be distinguished from an error of 2 bits or less, and then it may be miscorrected.

FIG. 3 is a block diagram showing a structure of the data comparator 3 shown in FIG. 1 . As shown in FIG. 3 , the data comparator 3 has data memory portions 31 a and 31 b, a decision address generator 32 , an exclusive OR operation portion 33 , a shift register 34 , AND operation portions 35 a and 35 b, and latches 36 a and 36 b. FIG. 4 is a diagram showing the contents of the data memory portions shown in FIG. 3 . The contents of the data are represented, from the leading bit in a block of demodulated data, as 0th bit, first bit, . . . , (n−1)th bit. FIG. 5 is a timing chart showing operation of the decision address generator 32 shown in FIG. 3 . The data memory portions 31 a and 31 b shown in FIG. 3 store one block of data in the demodulated data d 1 a and d 1 b, respectively, as shown in FIG. 4 . The decision address generator 32 receives the first-channel error locations e 2 a and the second-channel error locations e 2 b and arrange the first-channel first error location, first-channel second error location, second-channel first error location, and second-channel second error location in a time series manner as shown in FIG. 5 , and outputs them to the data memory portions 31 a and 31 b as decision address a 32 for data reading. It also outputs a clock c 32 for instructing the shift register 34 to read outputs corresponding to the respective error locations sent from the exclusive OR operation portion 33 , and also outputs a latch signal s 32 to the latches 36 a and 36 b. The exclusive OR operation portion 33 calculates exclusive OR of the outputs from the data memory portions 31 a and 31 b (corresponding to comparison between the data of two channels). As the result, when the data in the two channels are equal, it provides a comparison result at 0, and when the data are different, it provides a comparison result at 1, to the shift register. The shift register 34 captures the outputs from the exclusive OR operation portion 33 while shifting according to the rise of the clock c 32 to store the results of exclusive OR corresponding to the first-channel first error location, first-channel second error location, second-channel first error location, and second-channel second error location. The AND operation portion 35 a calculates AND of the comparison results corresponding to the first-channel first error location and the first-channel second error location stored in the shift register 34 . As the result, when the data in the two channels corresponding to the first-channel error locations both differ, it outputs an operation result at 1, and otherwise it outputs an operation result at 0. The AND operation portion 35 b ANDs the comparison results corresponding to the second-channel first error location and the second-channel second error location stored in the shift register 34 . As the result, when the data in the two channels corresponding to the second-channel error locations both differ, it outputs an operation result at 1, and otherwise it outputs an operation result at 0. The latches 36 a and 36 b respectively capture, hold, and output the outputs from the AND operation portions 35 a and 35 b in accordance with the timing of the latch signal s 32 from the decision address generator 32 . Accordingly, the data comparator in FIG. 3 outputs “1” as the first-channel decision signal s 1 a when data in the two channels corresponding to the first-channel error locations differ in both locations, and outputs “1” as the second-channel decision signal s 1 b when data in the two channels corresponding to the second-channel error locations differ in both locations.

When the number of errors is less than two, the data comparator 3 operates as shown below. First, the data memory portions 31 a and 31 b contain dummy data previously stored in region specified by dummy address, other than the region for storing received data, where the dummy data for the two channels are set to different values. Specifically, for example, the dummy address is set as n, the first-channel dummy data as “0,” and the second-channel dummy data as “1.” When the number of errors is one, the decision address generator 32 outputs the dummy address in place of the second error location, and when the number of errors is zero, it outputs the dummy address in place of the first error location and the second error location. As the result, when the number of errors is one, the exclusive OR corresponding to the second error location is always “1,” which is provided to the AND operation portion as one of its inputs. Then the result of comparison about the data in the first error location is intactly provided as the output from the AND operation portion. That is to say, when the number of errors is one, the data comparator 3 generates the decision signal on the basis of the data in the first error location only. When the number of errors is zero, all outputs from the exclusive OR are at “1,” and hence the two inputs to the AND operation portion are both at “1,” and therefore the output from the AND operation portion is always at “1.” That is to say, when the number of errors is zero, the data comparator 3 always generates “1” as the decision signal.

When a code capable of error correction of 3 or more bits is used as the error correcting code, a similar structure can be applied by increasing the number of stages of the shift register 34 and providing inputs of 3 or more bits to the AND operation portions 35 a and 35 b.

While FIG. 3 shows a structure for use with two channels of demodulated data, a data comparator having the structure shown in FIG. 6 , for example may be used in place of the data comparator shown in FIG. 3 . The data comparator can thus be easily expanded so that the selection can be made from three or more channels of inputs. The data comparator shown in FIG. 6 uses a data anticoincidence detector 331 in place of the exclusive OR operation portion 33 in the data comparator shown in FIG. 3 . The other parts operate in the same way as those in FIG. 3 . The data anticoincidence detector 331 outputs “0” when all inputs are equal, and outputs “1” when at least one of them differs. This is due to the fact that when all inputs are equal, the possibility that the bit is wrong is very small, and it is then expected that the estimated error location is incorrect. The data anticoincidence detector 331 may be structured to output “1” only when the value of the bit in the error location in the demodulated data of the target channel differs from values of the bits in the corresponding location in all other channels.

Although the data comparator in FIG. 3 outputs “1” as the decision signals when the data in the two channels corresponding to the error locations differ in both of the two locations, the data comparator may be structured as shown in FIG. 7 so that the decision signals are outputted when the data differ in at least one of the two locations. The data comparator in FIG. 7 has counting portions 37 a and 37 b and comparators 38 a and 38 b, in place of the logic operation portions 35 a and 35 b. The counting portions 37 a and 37 b count the numbers of different bits in the data in the two channels corresponding to the error locations. The comparators 38 a and 38 b compare the results counted by the counting portions 37 a and 37 b with a predetermined threshold (“1” in this case) and output “1” when they are equal to or larger than the threshold, and output “0” in other cases. The other structural elements and operation are the same as those described referring to FIG. 3 . When a 3-bit error correctable code is used as the error correcting code, an arbitrary number which is equal to or larger than 1 and smaller than the number of correctable bits can be used as the threshold, and the criterion for decision can thus be flexibly set.

FIG. 8 is a block diagram showing the structure of the data selector 4 shown in FIG. 1 . As shown in FIG. 8 , the data selector 4 has an erroneous symbol number comparator 61 , a decision logic operation portion 62 , selectors 64 , 65 , and 66 , an address generator 67 , and data memory portions 68 a and 68 b. The erroneous symbol number comparator 61 compares the erroneous symbol numbers e 1 a and e 1 b to determine which is larger, or whether they are equal, and outputs a signal showing the result of the decision as a comparison result “re” about the numbers of erroneous symbols. The decision logic operation portion 62 receives the result of the decision, re, from the erroneous symbol number comparator 61 , the decision signals s 1 a and s 1 b, and the detection result d 8 from the data detector 8 , and decides which of the channels should be selected and generates a selection signal s 62 . The selectors 64 , 65 , and 66 respectively select one of the data read from the data memory portions, one of the input error locations, and one of the erroneous symbol numbers, in accordance with the selection signal s 62 from the decision logic operation portion 62 , and output them as the selected data d 2 , the selected error location e 4 , and the selected erroneous symbol number e 3 , respectively. The data memory portions 68 a and 68 b respectively store one block of data in the demodulated data d 1 a and d 1 b and read the data according to addresses from the address generator 67 . The address generator 67 sequentially generates the address ra from 0 to (n−1) for one block of data, which are provided to the data memory portions 68 a and 68 b and also to the outside.

FIG. 9 is a flowchart showing an example of the decision procedure made by the decision logic operation portion 62 shown in FIG. 8 . When the detection result d 8 from the data detector 8 indicates that one of the channels is invalid, the valid channel is selected independently of the comparison result re about the erroneous symbol numbers and the decision signals s 1 a and s 1 b. When the detection result d 8 from the data detector indicates that both channels are valid, a decision is made first on the basis of the comparison result re from the erroneous symbol number comparator 61 . That is to say, when the numbers of erroneous symbols differ, a channel with a less number of erroneous symbols is selected. When the erroneous symbol numbers are equal, one of the channels of which decision signal is at “1” is selected on the basis of the decision signals s 1 a (the decision signal for the first channel) and s 1 b (the decision signal for the second channel). When the decision signals for the two channels are both at “1” or both at “0,” it is impossible to decide which is better. Then one of the channels which was selected in the previous block is maintained. The decision logic operation portion 62 makes this decision for each block of the BCH code and updates the output.

FIG. 10 is a flowchart showing another example of the decision procedure of the decision logic operation portion 62 . In this decision procedure, when the detection result d 8 from the data detector 8 indicates that one of the channels is invalid, the valid channel is selected independently of the comparison result re about the numbers of erroneous symbols and the decision signals s 1 a and s 1 b, similarly to the decision procedure shown in FIG. 9 . However, unlike the decision procedure shown in FIG. 9 , when the detection result d 8 from the data detector indicates the state in which both channels are valid, the decision signals s 1 a (the first-channel decision signal) and s 1 b (the second-channel decision signal) are checked first and one of the channels of which decision signal is at “1” is selected. When the decision signals for the two channels are both at “1” or both at “0,” the decision is made on the basis of the comparison result re from the erroneous symbol number comparator 61 . That is to say, a channel with a smaller number of erroneous symbols is selected when the numbers of erroneous symbols differ. When the numbers of erroneous symbols are equal, it is impossible to decide which channel is superior, so the channel selected in the previous block is maintained. The decision logic operation portion 62 makes the decision for each block of BCH code and updates the output.

FIG. 11 is a block diagram showing the structure of the bit inverter 5 shown in FIG. 1 . As shown in FIG. 11 , the bit inverter 5 has a coincidence detector 71 , a comparator 72 , an AND operation portion 73 , and an exclusive OR operation portion 74 . The coincidence detector 71 outputs “1” only at an instant the input selected error location e 4 coincides with the address ra. The comparator 72 compares the selected erroneous symbol number e 3 with a given threshold and outputs “1” only when the selected erroneous symbol number e 3 does not exceed the threshold. The exclusive OR operation portion 74 inverts the bit in the selected data d 2 only when the output from the coincidence detector 71 and the output from the comparator 72 are both at “1.” Accordingly when the selected erroneous symbol number does not exceed the given threshold, the bit inverter 5 inverts the bit corresponding to the selected error location in the selected data, so as to correct the error. For the 2-bit error correcting BCH code, an error of 3 or more bits cannot be corrected. Therefore it is adequate to set the given threshold to 2.

Although the above-described data comparator 3 and data selector 4 have their respective data memory portions, a single memory can be shared for their functions. This enables reduction in memory capacity.

The above-described structure realizes a diversity reception apparatus which performs highly reliable selection for each block on the basis of the number of erroneous symbols and reliability of error locations in demodulated data, and also performs error correction at the same time.

In the description above, the input modulated signal is a chirp PSK signal and the demodulator is a subband demodulator having subband filters and differential detectors for two channels. However, the present invention can be applied to other applications, i.e., generally to diversity reception using a plurality of demodulated data. For example, as shown in the second embodiment of the present invention described below, the invention can be applied to a structure in which a plurality of demodulated data are obtained from a plurality of modulated signals obtained through reception at a plurality of antennas.

(Second Embodiment)

FIG. 12 is a block diagram showing a diversity reception apparatus according to a second embodiment of the present invention. As shown in FIG. 12 , this diversity reception apparatus differs from the diversity reception apparatus of the first embodiment shown in FIG. 1 in that it has a demodulator 1 ′ including a first detector 1 a and a second detector 1 b. Other structural elements are the same as those in the first embodiment shown in FIG. 1 . The demodulator 1 ′ obtains two channels of demodulated data d 1 a and d 1 b from two channels of modulated signals obtained by receiving radio waves at two channel of antennas. The operation is the same as that of the diversity reception apparatus shown in FIG. 1 in other respects, and therefore it is not described again.

(Third Embodiment)

FIG. 13 is a block diagram showing a diversity reception apparatus according to a third embodiment of the present invention. As shown in FIG. 13 , this diversity reception apparatus has error correcting portions 6 a, 6 b, . . . , 6 c in place of the error estimating portions 2 a, 2 b, . . . , 2 c, a data comparator 3 ′ in place of the data comparator 3 , and a data selector 4 ′ in place of the data selector 4 . The other structural elements and operation are the same as those in the first embodiment shown in FIG. 1 . Similarly to the error estimating portions 2 a, 2 b, . . . , 2 c shown in FIG. 1 , the error correcting portions 6 a, 6 b, . . . , 6 c estimate the numbers of erroneous symbols and the error locations from the demodulated data d 1 a, d 1 b, . . . , d 1 c. Further, when the number of errors is estimated to be 2 or less, they correct the erroneous bits in the demodulated data and output decoded data d 10 a, d 10 b, . . . , d 10 c. When the number of errors is estimated to be 3 or more, they output the intact demodulated data d 1 a, d 1 b, . . . ; d 1 c as the decoded data d 10 a, d 10 b, . . . , d 10 c without correcting. These error correcting portions can be realized with the same structure as the BCH error correcting device shown in U.S. Pat. No. 5,216,676, which is incorporated herein by reference, for example. The data comparator 3 ′ has the same function as the data comparator 3 shown in FIG. 1 . The data selector 4 selects one of the decoded data d 10 a, d 10 b, . . . , d 10 c on the basis of the erroneous symbol numbers e 1 a, e 1 b, . . . e 1 c and the decision signals s 1 a, s 1 b, . . . , s 1 c and outputs it as final decoded data d 10 .

FIG. 14 is a block diagram showing the structure of the data comparator 3 ′ shown in FIG. 13 in the case in which two channels of demodulated data are used. While it has almost the same structure as the data comparator shown in FIG. 3 , an exclusive NOR operation portion 33 ′ is used in place of the exclusive OR operation portion 33 , since the decoded data d 10 a and d 10 b are inputted in place of the demodulated data d 1 a and d 1 b. That is to say, the shift register 34 receives a comparison result at 1 when bits in error location in the two channels are equal, and at 0 when they are different. The other structural elements and operation are the same as those of the data comparator shown in FIG. 3 .

It is possible to easily expand the structure so that it can make the selection from three or more channels of inputs by using a data comparator having the structure shown in FIG. 15 , for example, in place of the data comparator shown in FIG. 14 . The data comparator of FIG. 15 includes a data coincidence detector 331 ′ in place of the data anticoincidence detector 331 in the data comparator of FIG. 6 . The other parts operate in the same way as those in the data comparator of FIG. 6 . The data coincidence detector 331 ′ outputs “1” when all inputs are equal and outputs “0” when even only one differs. This is due to the fact that it is expected that the results of error correction are correct when all inputs are equal. The data coincidence detector 331 ′ may be constructed to output “0” only when the value of the bit in the error location in the decoded data of the target channel differs from values of the bits in the corresponding location in all other channels.

FIG. 16 is a block diagram showing the structure of the data selector 4 ′ shown in FIG. 13 . Although it has almost the same structure as the data selector shown in FIG. 8 , it differs in that the decoded data d 10 a and d 10 b are inputted in place of the demodulated data d 1 a and d 1 b and that the decoded data d 10 is outputted in place of the selected data d 2 . Its operation is similar to that of the data selector shown in FIG. 8 . The data selector 4 ′ does not output the selected erroneous symbol number e 3 , selected error location e 4 , and address ra, nor receive the error locations e 2 a and e 2 b. Hence the data selector 4 ′ does not have the selectors 65 and 66 shown in FIG. 8 .

(Characteristic Evaluation in Error-Correcting and Combining)

As has been described above, diversity combining using an error correcting code is performed in the embodiments of the invention. According to the diversity combining, accuracy of error estimation is checked on the basis of comparison with other channels, so that the reliability of the error estimation can be maintained even when a less redundant and short code is used. It is also possible to select a channel of good quality even when all channels of demodulated data have the same extent of errors. For such diversity combining using an error correcting code (hereinafter referred to as “error-correcting and combining”), characteristic evaluation in the diversity combining according to the first embodiment is shown by the inventors of the present invention in “A Consideration of ‘Error-Correcting and Combining’ Method in SR-chirp Scheme,” Technical Report of IEICE (the Institute of Electronics, Information and Communication Engineers), CS98-33, May 1998. Now, on the basis of the contents of the technical report, characteristic evaluation in the error-correcting and combining of the first embodiment will be described. In the first embodiment, the present invention is applied to a diversity reception apparatus based on a modulating/demodulating scheme called “SR-chirp PSK scheme.” The “SR-chirp PSK scheme” is a modulating/demodulating scheme in which a chirp PSK signal obtained by multiplying a differential-PSK-modulated primary modulated signal by a chirp signal is transmitted and part of the band, subband, is extracted on the receiving side and demodulated by differential detection.

In the characteristic evaluation, a single-bit correcting code of binary BCH code (hereinafter referred to as BCH SEC code or simply as SEC code) and a double-bit correcting code of binary BCH code (hereinafter referred to as BCH DEC code, or simply as DEC code) are used as the error correcting code. In the description below, when one block is n-bit long and the information data is k-bit long as shown in FIG. 2 , the BCH SEC and BCH DEC are preceded by the numbers (n, k).

In error-correcting and combining, when the error rate on the transmission path is large to a certain extent, erroneous estimation may occur, and then a channel with a larger number of errors may be selected (hereinafter selecting a channel with more errors is referred to as selection error). FIG. 18 shows results obtained by computer simulations about the selection error probability in the first embodiment when ( 63 , 57 ) BCH SEC code and ( 63 , 51 ) BCH DEC code were used. In FIG. 18 , the abscissa shows the ratio of signal energy per one bit of information data to the noise power density, Eb/No, and the ordinate shows the selection error probability. The characteristic evaluation used the following modulation/demodulation parameters:

(a) The modulation/demodulation scheme is SR-chirp QPSK.

(b) The number of reception subbands is two channels.

(c) The reception subband frequency is f 0 ±1.75 MHz.

(d) For subband width, BT=2.4.

(e) Spreading rate is 10.8.

The transmission path is a static environment AWGN channel without multipath (hereinafter referred to as static AWGN). In FIG. 18 , the solid lines show characteristics in the case of error-correcting and combining, i.e , characteristics in the case where the error estimation accuracy is determined on the basis of comparison with other channels, and the dotted lines show characteristics in the case where the selection is made only on the basis of comparison among the estimated numbers of erroneous bits, without making comparison with other channels. In FIG. 18 , the thin lines show characteristics in the case where the SEC code is used, and the thick lines show characteristics in the case where the DEC code is used.

It is seen from FIG. 18 that, generally, the selection error probability is larger when the SEC code is used than when the DEC code is used. Comparing the solid lines and the dotted lines in FIG. 18 shows that determining the accuracy by comparison with other channels contributes to reduction of the selection errors. In the case of the SEC code, the selection errors are reduced by 10 times or more, and in the case of the DEC code, they are reduced by almost 10 times.

FIG. 19 shows the bit error rate characteristic (hereinafter referred to as BER characteristic) in the error-correcting and combining and the CRC selecting and combining under static environment without multipath. In the CRC selecting and combining (shown as CRC in FIG. 19 ), data is framed by using CRC error detecting code with unique word, information data, and CRC (Cyclic Redundancy Check code) arranged in order, and data in a channel with no error are selected and combined frame by frame. In the error-correcting and combining (shown as ECC in FIG. 19 ), ( 63 , 51 ) BCH DEC code was used. For a reference, the diagram shows characteristic of one subband (no correcting/combining), too. In FIG. 19 , the solid line and the dotted line show results of computer simulation, and the points show actual measurements with an artificial transmission path and prototype modem.

It is seen from FIG. 19 that the required Eb/No is improved by about 2 to 3 dB in the error-correcting and combining, as compared with the CRC selecting and combining. In the CRC selecting and combining, no improvement is seen as compared with that with one subband in the region with larger bit error rate (BER, 10 −3 or higher). This is due to the fact that frames in both channels suffer errors at the same time and the selection cannot be made appropriately. In contrast, in the error-correcting and combining, it is improved also in the region with relatively large error rate in which the bit error rate (BER) is about 10 −3 to 10 −2 .

While the invention has been described in detail, the foregoing description is in all aspects illustrative and not restrictive. It is understood that numerous other modifications and variations can be devised without departing from the scope of the invention.

Claims

1. A diversity reception apparatus receiving one or a plurality of modulated signals modulated with error-correctably coded data, comprising:demodulating means for demodulating said modulated signal(s) to obtain demodulated data in a plurality of channels corresponding to said coded data; error estimating means for estimating the number of erroneous symbols and error locations thereof in the demodulated data in each channel; data comparing means for comparing the data in said error locations estimated in the demodulated data in each channel with the data in the corresponding locations in the demodulated data in other channels to determine whether said error locations are correct or not; and data selecting means for selecting one of said plurality of channels on the basis of said numbers of erroneous symbols and results of the determination made by said data comparing means.

2. The diversity reception apparatus according to claim 1, wherein when all data in said error locations estimated in the demodulated data in one channel of said plurality of channels respectively differ from the data in the corresponding locations in the demodulated data in all other channels, said data comparing means determines that said error locations in said one channel are correct.

3. The diversity reception apparatus according to claim 1, wherein when all data in said error locations estimated in the demodulated data in one channel of said plurality of channels respectively differ from the data in the corresponding locations in the demodulated data in at least one of other channels, said data comparing means determines that said error locations in said one channel are correct.

4. The diversity reception apparatus according to claim 1, wherein when data whose number is equal to or larger than a predetermined threshold in the data in said error locations estimated in the demodulated data in one channel of said plurality of channels respectively differ from the data in the corresponding locations in the demodulated data in at least one of other channels, said data comparing means determines that said error locations in said one channel are correct.

5. The diversity reception apparatus according to claim 1, wherein said error estimating means estimates error locations in the demodulated data in each channel bit by bit, andwhen all bits in said error locations estimated in the demodulated data in one channel of said plurality of channels respectively differ from the bits in the corresponding locations in the demodulated data in at least one of other channels, said data comparing means determines that said error locations in said one channel are correct.

6. The diversity reception apparatus according to claim 5, further comprising bit inverting means for inverting the bits in said error locations in the demodulated data in the channel selected by said data selecting means when said number of erroneous symbols is equal to or smaller than a given value.

7. The diversity reception apparatus according to claim 6, wherein said bit inverting means does not perform the bit inversion when said number of erroneous symbols exceeds the given value.

8. The diversity reception apparatus according to claim 1, wherein said error estimating means estimates said error locations in the demodulated data in each channel bit by bit, andwhen bits whose number is equal to or larger than a predetermined threshold in the bits in said error locations estimated in the demodulated data in one channel of said plurality of channels respectively differ from the bits in the corresponding locations in the demodulated data in at least one of other channels, said data comparing means determines that said error locations in said one channel are correct.

9. The diversity reception apparatus according to claim 1, wherein said data selecting means selects a channel in which said number of erroneous symbols is the smallest.

10. The diversity reception apparatus according to claim 1, wherein when said numbers of erroneous symbols are the smallest in a plurality of said channels, said data selecting means selects a channel in which said error locations are determined to be correct by said data comparing means from among the plurality of said channels in which said numbers of erroneous symbols are the smallest.

11. The diversity reception apparatus according to claim 1, wherein when said numbers of erroneous symbols are equal in all said channels, said data selecting means selects a channel in which said error locations are determined to be correct by said data comparing means.

12. The diversity reception apparatus according to claim 1, wherein the number of said channels of said demodulated data is two, andwhen said numbers of erroneous symbols are equal in the two channels and said error locations in the two channels are both determined to be correct by said data comparing means, said data selecting means maintains the preceding selection.

13. The diversity reception apparatus according to claim 1, wherein the number of said channels of said demodulated data is two, andwhen said numbers of erroneous symbols are equal in the two channels and said error locations in the two channels are both determined to be incorrect by said data comparing means, said data selecting means maintains the preceding selection.

14. The diversity reception apparatus according to claim 1, wherein when said data comparing means determines that said error locations in a plurality of said channels are correct, said data selecting means selects a channel in which said number of erroneous symbols is the smallest from among the plurality of said channels in which said error locations are determined to be correct.

15. The diversity reception apparatus according to claim 1, wherein when said data comparing means determines that all of said error locations in each of said channels are not correct, said data selecting means selects a channel in which said number of erroneous symbols is the smallest.

16. The diversity reception apparatus according to claim 1, wherein said error estimating means estimates the number of erroneous symbols and error locations thereof for each block defined by dividing the demodulated data in each channel to given length, andsaid data selecting means selects one of said plurality of channels for each said block.

17. The diversity reception apparatus according to claim 1, further comprising data detecting means for determining whether the demodulated data in each of said plurality of channels is valid or invalid, andsaid data selecting means selects one of the channels determined to be valid by said data detecting means from among said plurality of channels.

18. The diversity reception apparatus according to claim 17, wherein said data detecting means is unique word detecting means for detecting particular data pattern in said demodulated data.

19. The diversity reception apparatus according to claim 1, wherein said modulated signal is a chirp PSK signal obtained by multiplying the phase-shift keying signal by a chirp signal whose frequency is swept at the same period as the symbol period, andsaid demodulating means includes subband filtering means for extracting part of the band of said modulated signal and differential detecting means for differentially detecting an output of said subband filtering means.

20. A diversity reception apparatus receiving one or a plurality of modulated signals modulated with error correctably coded data, comprising:demodulating means for demodulating said modulated signal(s) to obtain demodulated data in a plurality of channels corresponding to said coded data; error correcting means for estimating the number of erroneous symbols and error locations thereof in the demodulated data in each channel, and for correcting an error in the demodulated data in each channel on the basis of said error locations and outputting the error-corrected demodulated data as decoded data; data comparing means for comparing the data in said error locations estimated in the decoded data in each channel with the data in the corresponding locations in the decoded data in other channels to determine whether said error locations are correct or not; and data selecting means for selecting one of said plurality of channels on the basis of said numbers of erroneous symbols and results of the determination made by said data comparing means.

21. The diversity reception apparatus according to claim 20, wherein when all data in said error locations estimated in the decoded data in one channel of said plurality of channels are respectively equal to the data in the corresponding locations in the decoded data in all other channels, said data comparing means determines that said error locations in said one channel are correct.

22. The diversity reception apparatus according to claim 20, wherein when all data in said error locations estimated in the decoded data in one channel of said plurality of channels are respectively equal to the data in the corresponding locations in the decoded data in at least one of other channels, said data comparing means determines that said error locations in said one channel are correct.

23. The diversity reception apparatus according to claim 20, wherein when data whose number is equal to or larger than a predetermined threshold in the data in said error locations estimated in the decoded data in one channel of said plurality of channels are respectively equal to the data in the corresponding locations in the decoded data in at least one of other channels, said data comparing means determines that said error locations in said one channel are correct.

24. The diversity reception apparatus according to claim 20, wherein said error correcting means does not perform the error correction when said number of erroneous symbols exceeds a given value.

25. The diversity reception apparatus according to claim 20, wherein said data selecting means selects a channel in which said number of erroneous symbols is the smallest.

26. The diversity reception apparatus according to claim 20, wherein when said numbers of erroneous symbols are the smallest in a plurality of said channels, said data selecting means selects a channel in which said error locations are determined to be correct by said data comparing means from among the plurality of said channels in which said numbers of erroneous symbols are the smallest.

27. The diversity reception apparatus according to claim 20, wherein when said numbers of erroneous symbols are equal in all of said channels, said data selecting means selects a channel in which said error locations are determined to be correct by said data comparing means.

28. The diversity reception apparatus according to claim 20, wherein the number of said channels of said demodulated data is two, andwhen said numbers of erroneous symbols are equal in the two channels and said error locations in the two channels are both determined to be correct by said data comparing means, said data selecting means maintains the preceding selection.

29. The diversity reception apparatus according to claim 20, wherein the number of said channels of said demodulated data is two, andwhen said numbers of erroneous symbols are equal in the two channels and said error locations in the two channels are both determined to be incorrect by said data comparing means, said data selecting means maintains the preceding selection.

30. The diversity reception apparatus according to claim 20, wherein when said data comparing means determines that said error locations in a plurality of said channels are correct, said data selecting means selects a channel in which said number of erroneous symbols is the smallest from among the plurality of said channels in which said error locations are determined to be correct.

31. The diversity reception apparatus according to claim 20, wherein when said data comparing means determines that said error locations are incorrect in all of said channels, said data selecting means selects a channel in which said number of erroneous symbols is the smallest.

32. The diversity reception apparatus according to claim 20, wherein said error correcting means estimates the number of erroneous symbols and error locations thereof for each block defined by dividing the demodulated data in each channel to given length, andsaid data selecting means selects one of said plurality of channels for each said block.

33. The diversity reception apparatus according to claim 20, further comprising data detecting means for determining whether the demodulated data in each of said plurality of channels is valid or invalid, andsaid data selecting means selects one of the channels determined to be valid by said data detecting means from among said plurality of channels.

34. The diversity reception apparatus according to claim 33, wherein said data detecting means is unique word detecting means for detecting particular data pattern in said demodulated data.

35. The diversity reception apparatus according to claim 20, wherein said modulated signal is a chirp PSK signal obtained by multiplying the phase-shift keying signal by a chirp signal whose frequency is swept at the same period as the symbol period, andsaid demodulating means includes subband filtering means for extracting part of the band of said modulated signal and differential detecting means for differentially detecting an output of said subband filtering means.

36. A diversity reception apparatus receiving one or a plurality of modulated signals modulated with error correctably coded data, comprising:demodulating means for demodulating said modulated signal(s) to obtain demodulated data in a plurality of channels corresponding to said coded data; error estimating means for estimating an error location in the demodulated data in each channel; data comparing means for comparing the data in said error location estimated in the demodulated data in each channel with the data in the corresponding location in the demodulated data in other channels to determine whether said error location is correct or not; and data selecting means for selecting a channel in which said error location is determined to be correct by said data comparing means from among said plurality of channels.