JP2006279254A - Timing detection method, and timing detector and receiver utilizing the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect switching timing from a single carrier signal to a multi-carrier signal. <P>SOLUTION: A matched filter 40 calculates a correlation value between a received burst signal and a prescribed signal pattern. A symbol timing detection part 44 detects start timing when the prescribed signal pattern is repeated in a series of the known single carrier signal arranged in the burst signal on the basis of the correlation value calculated by the matched filter 40. When the start timing is detected, the matched filter 40 starts to calculate a correlation value between the received burst signal and a series of the known multi-carrier signal. The symbol timing detection part 44 detects the timing of switching from the series of the known single carrier signal to the series of the known multi-carrier signal near the start timing on the basis of the correlation value calculated by the matched filter 40. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a timing detection technique, and more particularly, to a timing detection method for detecting timing with respect to a received signal, a timing detection apparatus and a reception apparatus using the timing detection method.

In the field of wireless communication, the spread spectrum communication system (SS) has been studied conventionally. The spread spectrum communication system includes a direct spread system (DS) and a frequency hopping system (FH). In the FH system, spread spectrum communication is performed by hopping the carrier frequency one after another based on a code sequence. Therefore, the spectrum distribution in the FH system occupies a wide band when observed for a long time, but is a signal that occupies only a specific frequency band when observed in units of one bit or symbol, and is narrower than the DS system. Signal. Therefore, since it can be said that it is an interference avoidance type SS, there is an advantage that the probability that a plurality of users communicate at the same frequency at the same time becomes small. In general, the frequency band of a transmission signal is determined according to a predetermined frequency hopping pattern. However, a plurality of frequency hopping patterns are provided, and the receiving side determines in advance which frequency hopping pattern is used for which timing. Since it is not known, it is necessary to establish both time and frequency synchronization by synchronization acquisition (for example, see Patent Document 1).
JP 2001-285247 A

  The burst signal in the MB-OFDM system includes a preamble part, a header part, and a payload part. In addition, each part is configured by a predetermined number of symbols according to the communication mode. A preamble part is used for the synchronization acquisition process. The preamble part is divided into timing synchronization and channel estimation, the timing synchronization preamble pattern is a single carrier signal defined in the time domain, and the channel estimation preamble pattern is a multicarrier signal defined in the frequency domain. is there. Communication data is arranged in the header part and the payload part, both of which are multicarrier signals. There are two types of timing synchronization. One is detection of symbol boundaries, and the other is detection of switching timing from a single carrier signal to a multicarrier signal. If one symbol of the time synchronization preamble is {Spi} (i = 1, 2,..., N), {Spi} is followed by l symbols, and {−Spi} is followed by m symbols. The preamble is composed. Here, n = 1 + m.

  Normally, an N-tap matched filter is used in the period of l symbols for symbol boundary detection. Further, the switching timing is detected in a period of m symbols. In the m symbol period, the matched filter operated in the l symbol period is continuously operated to detect a symbol whose peak value of the correlation value output is inverted. However, when the determination is performed using the signed correlation value, if the frequency offset is incomplete, the detection accuracy decreases as the SNR deteriorates, so that the switching timing detection accuracy decreases, resulting in a channel or the like. Will not function properly.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide a timing detection technique for detecting the timing of switching from a single carrier signal to a multicarrier signal.

  In order to solve the above-described problem, a timing detection device according to an aspect of the present invention includes a known single carrier signal sequence configured by repetition of a predetermined signal pattern, and the same bandwidth as the known single carrier signal sequence. Calculating a first correlation value between a received burst signal and a predetermined signal pattern at a receiving unit that receives a burst signal in which at least a sequence of known multicarrier signals defined by is continuously arranged And when a predetermined signal pattern is repeated among a series of known single carrier signals arranged in the received burst signal based on the matched filter calculated and the first correlation value calculated by the matched filter. A detection unit for detecting the start timing. In the matched filter, when the detection timing is detected by the detection unit, calculation of the second correlation value between the burst signal received by the reception unit and the known multicarrier signal sequence is started. In the vicinity of the timing, a timing for switching from a known single carrier signal sequence to a known multicarrier signal sequence is detected based on the second correlation value calculated by the matched filter.

  According to this aspect, since the switching timing also appears at the timing at which the start timing should appear, the switching timing is detected with high accuracy by detecting the switching timing while using a known multicarrier signal after the start timing is detected. It can be detected.

  In the matched filter, when calculating the second correlation value, a time-domain value for a known multicarrier signal sequence may be used. In this case, since the second correlation value can be calculated with the same circuit configuration as that for calculating the first correlation value, an increase in circuit scale can be suppressed.

  The detection unit detects a peak for the first correlation value a plurality of times, detects a start timing based on the peak for the first correlation value detected a plurality of times, and a peak for the second correlation value And a means for detecting timing for switching from a known single carrier signal sequence to a known multicarrier signal sequence based on a peak for one second correlation value. In this case, since the start timing is detected from the first correlation values detected a plurality of times, the detection accuracy of the start timing can be improved. Further, since the switching timing is detected from the second correlation value detected once, the delay due to the switching timing detection process can be reduced.

  When detecting the peak with respect to the first correlation value, the detection unit compares the ratio of the average value of the first correlation value and the peak with respect to the first correlation value with a threshold value, and detects the peak with respect to the second correlation value. When detecting the peak, the ratio between the average value of the second correlation values and the peak with respect to the second correlation value may be compared with a threshold value. In this case, since the relative value is compared with the threshold value, the influence of noise or the like can be reduced.

  Another aspect of the present invention is a receiving device. In this apparatus, a sequence of known single carrier signals configured by repetition of a predetermined signal pattern and a sequence of known multicarrier signals defined by the same bandwidth as the sequence of known single carrier signals are continuous. Receiving a burst signal arranged in the receiver, a matched filter for calculating a first correlation value between the burst signal received by the receiving unit and a predetermined signal pattern, and a first filter calculated by the matched filter Based on the correlation value, among the known single carrier signal sequences arranged in the received burst signal, a detection unit for detecting a start timing when a predetermined signal pattern is repeated, and detection of the start timing by the detection unit And a processing unit that performs predetermined processing on the burst signal. In the matched filter, when the detection timing is detected by the detection unit, calculation of the second correlation value between the burst signal received by the reception unit and the known multicarrier signal sequence is started. In the vicinity of the timing, based on the second correlation value calculated by the matched filter, a switching timing for switching from a known single carrier signal sequence to a known multicarrier signal sequence is detected. When the switching timing is detected by the detection unit, processing for the multicarrier signal is executed.

  According to this aspect, since the switching timing also appears at the timing when the start timing should appear, the switching timing is detected with high accuracy by detecting the switching timing while using a known multicarrier signal after the start timing is detected. Can be detected.

  Yet another embodiment of the present invention is a timing detection method. In this method, a sequence of known single carrier signals configured by repetition of a predetermined signal pattern and a sequence of known multicarrier signals defined by the same bandwidth as the sequence of known single carrier signals are continuous. Receiving a burst signal arranged in the step, calculating a first correlation value between the received burst signal and a predetermined signal pattern, and receiving a burst based on the calculated first correlation value A step of detecting a start timing when a predetermined signal pattern is repeated in a sequence of known single carrier signals arranged in the signal, and when the start timing is detected, a received burst signal and a known multicarrier Calculating a second correlation value with the sequence of signals, and in the vicinity of the start timing, Based on the function value, comprising the steps of: detecting the timing of switching from the series of the known single carrier signal into a sequence of a known multi-carrier signal.

  The step of calculating the second correlation value may use time domain values for a known sequence of multicarrier signals. The step of detecting the start timing detects the peak for the first correlation value a plurality of times, detects the start timing based on the peak for the first correlation value detected a plurality of times, and detects a known single carrier signal The step of detecting the timing of switching from a sequence of known multicarrier signals to a sequence of known multicarrier signals detects one peak with respect to the second correlation value, and based on the peak with respect to one second correlation value, a known single carrier You may detect the timing which switches from the signal series to the known multicarrier signal series. The step of detecting the start timing compares the ratio of the average value of the first correlation value and the peak of the first correlation value with a threshold when detecting the peak for the first correlation value, The step of detecting the timing of switching from a single carrier signal sequence to a known multicarrier signal sequence comprises detecting an average value of the second correlation value and a second correlation value when detecting a peak with respect to the second correlation value. The ratio of peak to may be compared with a threshold value.

  It should be noted that any combination of the above-described constituent elements and a conversion of the expression of the present invention between a method, an apparatus, a system, a recording medium, a computer program, etc. are also effective as an aspect of the present invention.

  According to the present invention, it is possible to detect the timing of switching from a single carrier signal to a multicarrier signal.

  Before describing the present invention in detail, an outline will be described. Embodiments of the present invention relate to a communication system in which frequency hopping is performed on a symbol basis. In the burst signal in the communication system according to the present embodiment, a preamble is added to the head portion. The preamble includes two types of preambles. One is a preamble signal of a single carrier signal (hereinafter referred to as “first preamble signal”), and the other is a preamble signal of a multicarrier signal (hereinafter referred to as “second preamble signal”). At this time, it is assumed that the bandwidths of the first preamble signal and the second preamble signal are equal. The receiving apparatus according to the present embodiment operates as follows in order to detect the timing of switching from the first preamble signal to the second preamble signal. It is assumed that the frequency hopping pattern can be recognized by a predetermined method.

  The first preamble is configured by repeating a predetermined signal pattern a plurality of times. The receiving apparatus sets a predetermined signal pattern in the matched filter. The matched filter detects a correlation value between the received burst signal and a predetermined signal pattern. The receiving device detects the start timing of a predetermined signal pattern (hereinafter simply referred to as “start timing”) by detecting the peak of the correlation value. When detecting the start timing, the receiving apparatus sets the second preamble signal in the matched filter. The receiving device detects a correlation value between the received burst signal and the second preamble signal. The first preamble signal and the second preamble signal are continuously arranged, and the first preamble signal includes a plurality of start timings. Therefore, after the appearance of a predetermined number of start timings, the next start timing becomes the switching timing from the first preamble signal to the second preamble signal (hereinafter simply referred to as “switching timing”). As a result, the receiving apparatus detects the switching timing by detecting the peak of the correlation value in the vicinity of the start timing.

  Moreover, in order to detect a frequency hopping pattern, the receiving apparatus performs the following process. A plurality of types of first preamble patterns are provided, and each preamble pattern corresponds to a frequency hopping pattern. That is, if the receiving apparatus specifies a preamble pattern in a preamble section of a burst signal, it can also acquire a frequency hopping pattern corresponding to the specified preamble pattern. The receiving apparatus specifies a preamble pattern by a matched filter, but includes only one matched filter without including a matched filter corresponding to the number of preamble patterns. Therefore, the number of taps included in one matched filter and corresponding to the number of samples included in one symbol is divided by the number of preamble patterns, and the divided taps Are each set as a group. Further, according to the number of taps of each group, a part of the preamble pattern is extracted, and it is applied to all the groups, and each of the extracted preamble patterns is combined to form a new preamble pattern (hereinafter referred to as “preamble pattern”). , “Specification signal series”).

  The receiving apparatus performs a cross-correlation process between the received signal and the signal sequence for identification, but adds the cross-correlation processes in units of groups and outputs a plurality of partial correlation values in units of groups. By comparing the magnitudes of a plurality of output correlation values, the number of groups is limited to a number smaller than the number of groups provided in advance. Further, the receiving device generates a new signal sequence for identification according to the limited number of groups, repeatedly executes the same processing, and finally selects one group. The receiving apparatus specifies a preamble pattern corresponding to one selected group, and accordingly specifies a frequency hopping pattern. By limiting the number of groups, the length of each preamble included in one specifying signal sequence is increased, and the accuracy of the correlation value as a result is improved.

  FIG. 1 shows a configuration of a communication system 100 according to the embodiment. The communication system 100 includes a transmission device 10 and a reception device 12. The transmission device 10 includes a baseband modulation unit 14, an up converter 16, a code generation unit 18, a frequency synthesizer 20, and a transmission antenna 22. The reception device 12 includes a reception antenna 24, a down converter 26, and a synchronization acquisition unit. 28, a code generator 30, a frequency synthesizer 32, a baseband demodulator 34, and a controller 36. The signal includes a baseband signal 200, a synchronization pattern signal 202, and a synchronization timing signal 204.

  The baseband modulation unit 14 modulates the data signal based on a modulation scheme such as PSK, MSK, or OFDM. In addition, the baseband modulation unit 14 arranges the first preamble signal and the second preamble signal at the head part of the burst signal. The format of the burst signal and the configuration of the first preamble signal and the second preamble signal will be described later. The code generator 18 generates a pseudo random code signal, and the frequency synthesizer 20 generates a carrier wave to be randomly hopped based on the pseudo random code signal. The up-converter 16 frequency-hops the modulated signal using a carrier wave that is randomly hopped. The transmitting antenna 22 transmits a frequency hopped signal. The receiving antenna 24 receives a signal transmitted from the transmitting antenna 22. The frequency synthesizer 32 generates a carrier wave to be randomly hopped similarly to the frequency synthesizer 20, and the down converter 26 converts the frequency of the received signal by the randomly hopped carrier wave. The frequency-converted signal is output as a baseband signal 200.

  Here, if the frequency hopping pattern of the carrier wave generated by the frequency synthesizer 20 matches the frequency hopping pattern of the carrier wave generated by the frequency synthesizer 32, the down converter 26 can accurately convert the frequency of the received signal. If you don't, you can't convert the frequency. Therefore, the synchronization acquisition unit 28 synchronizes the frequency hopping pattern of the carrier wave generated by the frequency synthesizer 32 with the frequency hopping pattern of the received signal so that the received signal can be accurately frequency-converted. An instruction signal related to the synchronization of the hopping pattern is output as the synchronization pattern signal 202. Further, the synchronization acquisition unit 28 also executes symbol timing synchronization of the received signal, and outputs an instruction signal related to symbol timing synchronization as the synchronization timing signal 204.

  Furthermore, the synchronization acquisition unit 28 detects the start timing of the first preamble signal and the switching timing from the first preamble signal to the second preamble signal as symbol timing. These detected timings are also output as the synchronization timing signal 204. The baseband demodulator 34 performs demodulation processing on the burst signal whose start timing is detected by the synchronization acquisition unit 28. The demodulation process corresponds to the modulation process in the baseband modulation unit 14. Further, when the switching timing is detected in the synchronization acquisition unit 28, the demodulation processing for the multicarrier signal is executed. The demodulation process for the multicarrier signal includes, for example, FFT.

  FIGS. 2A to 2E show the structure of a burst format according to the embodiment. FIG. 2A shows a burst format in the MB-OFDM system. The horizontal axis is time. A frame is roughly divided into a preamble part, a header part, and a data part, and each frame is composed of a number of symbol data defined according to the communication mode. In the preamble portion, “PS / FS” and “CE” are continuously arranged. “PS / FS” corresponds to the aforementioned first preamble signal, and “CE” corresponds to the aforementioned second preamble signal. The switching timing described above corresponds to the leading timing of the second preamble signal. The first preamble signal is a single carrier signal, and the second preamble signal is a multicarrier signal, but both bandwidths are defined to be equal. The header part and the data part are also composed of multicarrier signals.

  FIG. 2B shows the configuration of the first preamble signal. Here, the first preamble signal is composed of n symbols, and one symbol is composed of a signal of 128 samples. That is, the first preamble signal is configured by repeating a signal of 128 samples as a predetermined signal pattern. As described above, in the first symbol signal of n symbols, the value is inverted between the l symbol portion and the m symbol portion. The aforementioned start timing corresponds to the start timing for one symbol. Further, since frequency hopping is performed on a symbol basis, the same frequency is used in one symbol.

  FIG. 2C shows a pattern of the first preamble signal. A preamble part is used for the synchronization acquisition process, and four types of patterns of the first preamble are provided while maintaining an orthogonal relationship with each other. Here, the four types of first preamble patterns are referred to as a first pattern to a fourth pattern. Further, as described above, four types of frequency hopping patterns are defined corresponding to each of the four types of first preamble patterns.

  FIG. 2D shows the configuration of the second preamble signal. Here, the second preamble signal is composed of x symbols. FIG. 2E shows the configuration of one symbol of the second preamble signal. As shown in the figure, one symbol is constituted by a signal of 165 samples. Here, the second preamble signal is a multicarrier signal, and a signal of 128 samples out of a signal of 165 samples corresponds to a signal obtained by IFFT of the multicarrier signal. Of the 165 sample signals, 32 sample signals correspond to the guard interval, and 5 samples correspond to the guard time. Note that, as described above, since the bandwidths of the first preamble signal and the second preamble signal are equal, the interval between both samples is also equal.

  FIGS. 3A to 3E show hopping frequencies and hopping patterns according to the embodiment. Here, as a wireless network having a narrower range than the wireless LAN, a WPAN (Wireless Personal Area Network) that is a short-range wireless network between PDAs and peripheral devices is targeted. In WPAN, higher speed is required compared with USB, Wireless 1394, or Bluetooth (registered trademark), and one of the methods for realizing this is the MB-OFDM method. FIG. 3A shows the target hopping frequency. Here, the frequencies “f1”, “f2”, and “f3” are used. FIG. 3B shows a first hopping pattern, which is a hopping pattern corresponding to the first pattern of FIG. In the period of 6 symbols, frequency hopping is performed in the order of “f1” → “f2” → “f3” → “f1” → “f2” → “f3”. Here, the timing of each symbol is indicated by “S1” to “S3”.

  FIG. 3 (c) shows a second hopping pattern, which is a hopping pattern corresponding to the second pattern of FIG. 2 (c). In the period of 6 symbols, frequency hopping is performed in the order of “f1” → “f3” → “f2” → “f1” → “f3” → “f2”. FIG. 3 (d) shows a third hopping pattern, which is a hopping pattern corresponding to the third pattern of FIG. 2 (c). In the period of 6 symbols, frequency hopping is performed in the order of “f1” → “f1” → “f2” → “f2” → “f3” → “f3”. FIG. 3 (e) shows a fourth hopping pattern, which is a hopping pattern corresponding to the fourth pattern of FIG. 2 (c). In the period of 6 symbols, frequency hopping is performed in the order of “f1” → “f1” → “f3” → “f3” → “f2” → “f2”.

  FIG. 4 shows the configuration of the synchronization acquisition unit 28. The synchronization acquisition unit 28 includes a matched filter 40, a hopping pattern detection unit 42, a symbol timing detection unit 44, and a synchronization control unit 46. The signal includes a correlation value 206, a matched filter control signal 208, a hopping pattern detection unit control signal 210, a determination result 212, a symbol timing detection unit control signal 214, and a symbol timing 216.

  The matched filter 40 has a plurality of taps, and calculates a correlation value between the baseband signal 200 and the specifying signal sequence. The matched filter 40 receives a baseband signal 200 including a predetermined sequence of reference signals, that is, a first preamble signal having a predetermined pattern. At this time, only a signal corresponding to only a predetermined hopping frequency, for example, “f2”, among a plurality of hopping frequencies as shown in FIG.

  The matched filter 40 derives a signal sequence for identification based on the first pattern of the preamble that is a candidate for the pattern of the first preamble signal included in the baseband signal 200, based on the fourth pattern. Here, the plurality of taps included in the matched filter 40 are divided by the number of preamble patterns “4”, “4” groups are set, and the first patterns respectively corresponding to the “4” groups To a fourth pattern are combined to generate a signal sequence for identification. Since the length of the specifying signal sequence is matched with the number of taps of the matched filter 40, when generating the specifying signal sequence, each of the first pattern to the fourth pattern has a length of 1/4. Only the part corresponding to this is used. Further, the matched filter 40 calculates four types of correlation values corresponding to the groups, and outputs each of them as a correlation value 206.

The hopping pattern detection unit 42 averages the correlation value 206 output from the matched filter 40 in symbol units, generates and analyzes a delay profile, and detects a frequency hopping pattern. Specifically, the hopping pattern detection unit 42 selects a part of the groups corresponding to the preamble pattern incorporated in the signal sequence for identification based on the correlation value 206 calculated by the matched filter 40. Select. For example, when “4” groups are included, “2” groups are selected from the groups. Further, the selected result is output as a determination result 212 to the synchronization control unit 46, and the synchronization control unit 46 controls the number of taps of the matched filter 40 and the signal sequence for identification based on the determination result 212. Here, the number N of taps set in one group at the initial stage is set as N = α / β, where α is the number of samples included in one symbol and β is the number of frequency hopping patterns. Is done. In the second stage, N = 2α / β, and in the kth stage, N = 2 ^k−1 α / β. In the following description, α is assumed to be “128” and β is assumed to be “4”.

  As described above, the hopping pattern detection unit 42 causes the matched filter 40 to re-execute processing for the selected group through the synchronization control unit 46, and sets the preamble pattern included in the identification signal in stages. The pattern of the first preamble signal included in the baseband signal 200 is specified from one finally selected group. Based on the identified pattern of the first preamble signal, the synchronization control unit 46 identifies the frequency hopping pattern.

  At that time, the matched filter 40 calculates a correlation value 206 between the first preamble signal corresponding to one pattern and the baseband signal 200 in order to detect the start timing. This corresponds to the signal pattern included in the first preamble signal being set in the matched filter 40.

  The symbol timing detection unit 44 averages the correlation value 206 output from the matched filter 40 for each symbol, generates and analyzes a delay profile, and detects the symbol timing stepwise. Here, it is assumed that the processing of the symbol timing detection unit 44 is executed after the hopping pattern detection unit 42 limits the pattern of the first preamble signal included in the baseband signal 200 to one. In the first stage, the symbol timing detection unit 44 detects the start timing of the first preamble signal based on the correlation value 206 calculated by the matched filter 40. At this time, it is assumed that the ratio of the average value of the correlation value 206 and the ratio of the peak to the correlation value 206 is compared with a threshold value, and the symbol timing detection unit 44 detects a peak when the ratio becomes larger than the threshold value. Furthermore, the symbol timing detection unit 44 detects such a peak a plurality of times, and detects the start timing based on the peaks detected a plurality of times. For example, the start timing is detected by performing statistical processing on the peaks detected a plurality of times.

  Thus, when the start timing is detected by the symbol timing detection unit 44, the matched filter 40 starts calculating the correlation value 206 between the synchronization pattern signal 202 and the second preamble signal. At this time, “CE” in FIG. 2A is set in the matched filter 40. Note that “CE” is defined as a time-domain signal. That is, in the matched filter, the time domain value for the second preamble signal is used when calculating the correlation value 206.

  As a second stage, the symbol timing detection unit 44 detects the switching timing based on the correlation value 206 calculated by the matched filter 40 in the vicinity of the start timing. At this time, it is assumed that the ratio of the average value of the correlation value 206 and the ratio of the peak to the correlation value 206 is compared with a threshold value, and the symbol timing detection unit 44 detects a peak when the ratio becomes larger than the threshold value. Further, it is assumed that the symbol timing detection unit 44 detects one such peak and detects the switching timing when one peak is detected. Information such as the detected timing is sequentially input to the synchronization control unit 46 as the determination result 212 and the symbol timing 216. Based on these information, the synchronization control unit 46 matches the matched filter 40, the hopping pattern detection unit 42, the symbol timing. The detection unit 44 is controlled. For simplicity, the number of oversamples of the baseband signal 200 is “1”.

  This configuration can be realized in terms of hardware by a CPU, memory, or other LSI of any computer, and in terms of software, it is realized by a program having a communication function loaded in the memory. Describes functional blocks realized by collaboration. Accordingly, those skilled in the art will understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof.

  FIG. 5 shows the configuration of the matched filter 40. The matched filter 40 includes a first buffer 50a, a second buffer 50b, an M-th buffer 50m, collectively referred to as a buffer 50, a first multiplier 52a, a second multiplier 52b, and an M-th multiplier, collectively referred to as a multiplier 52. 52m, an addition unit 54, a storage unit 56, a selection unit 58, and a reference code buffer 60.

  The storage unit 56 stores in advance a plurality of identifying signal sequences or preamble patterns in which preambles having different patterns are combined. Here, “first pattern”, “second pattern”, “third pattern”, “fourth pattern”, “CE” are stored. Further, “a combination of the first pattern, the second pattern, the third pattern, and the fourth pattern”, “a combination of the first pattern and the second pattern”, “a combination of the first pattern and the third pattern”, “a first pattern” And 4th pattern ”,“ 2nd and 3rd pattern combination ”,“ 2nd and 4th pattern combination ”,“ 3rd and 4th pattern combination ”,“ 1st pattern ”, “Second pattern”, “third pattern”, and “fourth pattern” may be stored. In addition, the storage unit 56 selects groups so that the number of groups included in the signal sequence for identification decreases stepwise based on the instruction included in the matched filter control signal 208. The specifying signal sequence is selected such that the length of each preamble combined with the specifying signal sequence is increased stepwise.

  Based on an instruction included in the matched filter control signal 208, the selection unit 58 selects a sequence of identification signals including a corresponding group from among a plurality of identification signal sequences stored in the storage unit 56. . That is, the selection unit 58 selects the “combination of the first pattern, the second pattern, the third pattern, and the fourth pattern” from the storage unit 56 in the first stage, and the second stage. In “a combination of the first pattern and the second pattern”, “a combination of the first pattern and the third pattern”, “a combination of the first pattern and the fourth pattern”, “a combination of the second pattern and the third pattern”, “ The selection is made so that either “the combination of the second pattern and the fourth pattern” or “the combination of the third pattern and the fourth pattern” is formed, and the “first pattern” and “the second pattern” are selected in the third stage. ”,“ Third pattern ”, or“ fourth pattern ”. Thereafter, the selection unit 58 selects “CE”.

  The baseband signal 200 is sequentially stored in the buffer 50. As described above, since M is 128, the buffer 50 is composed of a 128-step shift register. On the other hand, the reference code buffer 60 is loaded with the signal sequence for identification selected at the search stage according to the instruction included in the matched filter control signal 208. That is, as described above, in the first stage, a partial sequence of preamble patterns corresponding to the first pattern to the fourth pattern of the preamble is set in order of 32 chips, and in the second stage, the determination result in the first stage is displayed. The preamble pattern partial series corresponding to the X1 pattern and X2 pattern of the preamble selected as the top two candidates (X1, X2 are either 1 to 4 and X1 and X2 are different) are 64 chips each. Set in order. Finally, as the third stage, 128 chips are set for the preamble corresponding to the Xth pattern (X is any one of 1 to 4) of the preamble specified in the second stage.

  A correlation value between the identifying signal sequence stored in the reference code buffer 60 and the 128 baseband signals 200 stored in the buffer 50 is calculated by the multiplier 52 and the adder 54. The multiplier 52 multiplies the 128 baseband signals 200 stored in the buffer 50 by the signal sequence for identification, and the adder 54 adds the multiplication results. Before the multiplication, if the signal sequence for identification is “0”, it is converted to “1”, and if it is “1”, it is converted to “−1”. The adder 54 is controlled to change the addition range as follows according to the first to third stages.

  In the first stage, (1) among the buffers 50, the output vector of the multiplier 52 corresponding to the 32nd 50 (not shown) from the first buffer 50a (here, the first buffer 50, the same applies hereinafter) is obtained. (2) Addition of the output vector of the multiplication unit 52 corresponding to the 64th buffer 50 (not shown) from the 33rd buffer 50 (not shown) of the buffer 50, (3) 65th (not shown) of the buffer 50 The output vector of the multiplication unit 52 corresponding to the 96th buffer 50 (not shown) is added from the buffer 50. (4) Of the buffers 50, multiplication corresponding to the 128th (M) th buffer 50 from the 97th buffer 50 (not shown). The output vectors of the unit 52 are added, and four correlation values that are the addition results of (1) to (4) are output as the correlation value 206.

  In the second stage, (1) among the buffers 50, the output vector of the multiplier 52 corresponding to the 64th buffer 50 (not shown) is added from the first buffer 50, and (2) the 65th (not shown) of the buffer 50. The output vectors of the multiplication unit 52 corresponding to the 128th buffer 50 are added from the first buffer 50, and two correlation values, which are the addition results of (1) and (2), are output as the correlation value 206. In the third stage, the output vectors of all the multiplication units 52 are added, and one correlation value is output as the correlation value 206.

  When a start timing is detected by a symbol timing detection unit 44 (not shown) based on the output correlation value 206, the selection unit 58 receives a signal indicating completion of start timing detection from the symbol timing detection unit 44. This signal is included in the matched filter control signal 208. Upon receiving the signal, the selection unit 58 selects “CE” and causes the reference code buffer 60 to set the selected “CE”. Since the processing of the buffer 50, the multiplication unit 52, and the addition unit 54 is as described above, description thereof is omitted. As a result of the above processing, a correlation value 206 for “CE” is output.

  FIG. 6 shows the configuration of the hopping pattern detection unit 42. The hopping pattern detection unit 42 includes a separation unit 70, a first calculation unit 72 a collectively called a calculation unit 72, a second calculation unit 72 b, a third calculation unit 72 c, a fourth calculation unit 72 d, and a determination unit 74. The first calculator 72a includes an adder 76, a first selector 78, a delay profile memory 80, a second selector 82, a switch 84, a delay profile buffer 86, a peak detector 88, an intensity calculator 90, and an averaging unit 92. including.

The separation unit 70 separates the input correlation value 206 from the first pattern to the fourth pattern. The adder 76, the first selector 78, the delay profile memory 80, and the second selector 82 add the correlation values output from the separator 70. At this time, since the preamble has a periodicity in units of 16 samples, the correlation values are added every 16 samples based on the periodicity. The first selector 78 outputs the signal output from the adder 76 as data for updating the delay profile memory 80 in accordance with the hopping pattern detector control signal 210. The delay profile memory 80 is sequentially updated with data that is synchronously averaged every 16 samples in this way. Assuming that the time series data of the correlation value 206 immediately after the start of the effective correlation value 206 input is {Ci} (i = 1, 2, 3,...), The output signal vector from the delay profile memory 80 { AVE1, AVE2,..., AVE16} are as follows.
AVE1 = C1 + C17 + ... + C113
AVE2 = C2 + C18 + ... + C114
...
AVE16 = C16 + C32 + ... + C128 (1)

  When the above averaging is completed, the delay profile buffer 86 is turned on, and the delay profile data (AVE1-16) is loaded into the delay profile buffer 86. Further, the delay profile buffer 86 is turned OFF, the delay profile memory 80 is further reset, and the process proceeds to delay profile generation for the second stage. Strictly speaking, the delay profile data calculated here is a sum of 128 propagation paths divided into 16 paths. For this reason, it does not have accurate information on delay dispersion, but timing information is not necessary for frequency hopping pattern detection in this embodiment.

The peak detector 88 obtains the maximum value PEAK_k (= max (AVE1, AVE2,..., AVE16)) of the delay profile data loaded in the delay profile buffer 86. Note that k is 2 or 4. The averaging unit 92 calculates an average value AVE_k (= (AVE1 + AVE2 +... + AVE16) / 16) of delay profile data. The intensity calculator 90 calculates the relative peak intensity as follows.
Relative peak intensity = max (AVE1, AVE2,..., AVE16) / ((AVE1 + AVE2 +... + AVE16) / 16)
= 16 * max (AVE1, AVE2,..., AVE16) / (AVE1 + AVE2 +... + AVE16) (2)

  The determination unit 74 compares the relative peak intensities output from the calculation unit 72 and selects a predetermined number of preamble patterns. Eventually, one preamble pattern is selected. Specifically, the selection is made as follows. In the first stage, the top two systems with relative peak intensities are selected, and their numbers (two of 1-4) are output as the determination result 212. In the second stage, the one with the larger relative peak intensity is selected, and the number (one of 1 to 4) is output as the determination result 212. The synchronization control unit 46 specifies a corresponding hopping pattern based on the determination result 212. The identified hopping pattern is output as the synchronization pattern signal 202.

  FIG. 7 shows the configuration of the symbol timing detection unit 44. The symbol timing detection unit 44 includes a relative peak level calculation unit 48 and a determination unit 98. The relative peak level calculation unit 48 includes a peak detection unit 94 and a moving average unit 96. The symbol timing detection unit 44 receives the correlation value 206 corresponding to the symbol pattern specified by the determination unit 74, and detects the timing of the baseband signal 200 based on the correlation value 206. As described above, the timing of the baseband signal 200 includes the start timing and the switching timing.

  The correlation value 206 is input to the relative peak level calculation unit 48. The start of the third stage is notified by the symbol timing detection unit control signal 214, and the relative peak level calculation unit 48 starts operation. The peak detector 94 detects the maximum value of the input correlation value 206 and holds the timing information together with the maximum value. Next, when the inputted correlation value 206 is larger than the retained maximum value, the maximum value register is updated with the value and the timing information is also updated.

  The correlation value 206 is also input to the moving average unit 96, and the moving average unit 96 performs a moving average of the correlation values 206 for 128 samples. When the processing for 128 samples is completed, the determination unit 98 calculates the relative peak level (= maximum value / moving average value), and determines the detection of the symbol timing by comparing with the threshold value. As shown in FIG. 2B, since the first preamble signal includes a plurality of symbols, the first preamble signal includes a plurality of start timings. When detecting the start timing, the determination unit 98 determines the detection of the symbol timing by a plurality of times equal to or less than l. Statistical processing such as averaging is performed on these results, and the start timing is detected. The determination unit 98 outputs the detected start timing as the symbol timing 216.

  After detection of the start timing, the symbol timing detection unit 44 executes switching timing detection based on an instruction included in the symbol timing detection unit control signal 214. The processing of the relative peak level calculation unit 48 and the determination unit 98 is the same as the start timing. However, when switching timing is detected, the determination unit 98 determines symbol timing detection only once, and then detects switching timing.

  FIG. 8 shows the operation timing of the synchronization acquisition unit 28. Here, the relationship between the baseband signal 200 and each procedure of the synchronization acquisition process is shown with the horizontal axis as the time axis. The first preamble signal is displayed by being divided for each symbol, but is based on the start timing of the operation of the matched filter 40. For this reason, the boundaries of the displayed symbols are virtual. The synchronization acquisition unit 28 fills the buffer 50 included in the matched filter 40 in the period “t1”. In the period “t2”, the hopping pattern detector 42 and the symbol timing detector 44 detect the frequency hopping pattern and the start timing. Further, in the period “t3”, the symbol timing detection unit 44 performs switching timing detection. The detection of switching timing is completed when the second symbol of the second preamble signal is received, but a symbol buffer for storing the first symbol is provided in advance for FFT processing. Therefore, it can be used in common and the circuit scale does not increase significantly even by the present invention.

  FIG. 9 shows the operation timing of the synchronization acquisition unit 28. FIG. 9 shows the relationship between the received preamble signal and each procedure of the synchronization acquisition process, with the horizontal axis as the time axis, as in FIG. Here, in particular, the period “t1” in FIG. 8 and the timing for detecting the frequency hopping pattern are shown. Although the first preamble signal is divided and displayed for each symbol, it is based on the start timing of the operation of the matched filter 40, and the symbol boundary of the display is virtual.

  “T1” is a filling period of the buffer 50 in the matched filter 40. “T2” is an operation (outputting a valid correlation value 206) period of the matched filter 40 in the first stage. “T3” is a candidate narrowing period of frequency hopping patterns in the first stage by the hopping pattern detection unit 42. Two candidates are selected. “T4” is a period during which the matched filter 40 operates (outputs a valid correlation value 206) in the second stage. “T5” is a frequency hopping pattern detection period in the second stage. “T6” is an operation (outputting a valid correlation value 206) period of the matched filter 40 in the third stage. The symbol timing detection by the symbol timing detection unit 44 is performed for each symbol during “T6”.

  With the above processing, processing up to symbol timing detection is completed in a preamble period of 19 symbols. As shown in FIG. 2C, the preamble used for synchronization acquisition is composed of n symbols. In this embodiment, since the synchronization acquisition is completed with the number of symbols less than n symbols, the processing within the period of the first preamble signal can be completed.

  FIGS. 10A to 10D show correlation values calculated in the first stage in the hopping pattern detection unit 42. FIG. These are examples of output waveforms of the delay profile buffer 86 in the first stage. FIGS. 10A to 10D show correlation values between the first to fourth patterns of the preamble corresponding to the frequency hopping pattern of the transmitted signal and the baseband signal 200, respectively. Here, the correlation value is a partial correlation for 32 samples. The relative peak intensity in FIGS. 10 (a) and 10 (d) is large. The determination unit 74 outputs an instruction signal to select the first pattern and the fourth pattern as a result of selection in the first stage.

  11A to 11B show the correlation values calculated in the second stage in the hopping pattern detection unit 42. FIG. These are examples of output waveforms of the delay profile buffer 86 in the second stage. FIGS. 11A to 11B respectively show correlation values between the first pattern, the fourth pattern, and the baseband signal 200 of the preamble corresponding to the frequency hopping pattern of the transmitted signal. Here, the correlation value is a partial correlation for 64 samples. The relative peak intensity in FIG. 11A is larger. The determination unit 74 outputs an instruction signal to select the first pattern as a result of the selection in the second stage.

  FIG. 12 shows the correlation values calculated in the third stage by the hopping pattern detection unit 42. This shows an output waveform example of the 128-tap matched filter 40. A correlation value peak is observed in one symbol period, and the symbol boundary timing can be calculated from this peak position.

  The operation of the synchronization acquisition unit 28 having the above configuration will be described. The reference code buffer 60 receives a signal sequence for identification that is a combination of four patterns. The matched filter 40 calculates four types of correlation values between the baseband signal 200 and the identifying signal sequence, and outputs a correlation value 206. The hopping pattern detection unit 42 derives the relative sizes of the four types of correlation values, and selects two patterns based on the relative sizes. The reference code buffer 60 receives a signal sequence for identification that is a combination of two patterns. The matched filter 40 calculates two types of correlation values between the baseband signal 200 and the identifying signal sequence, and outputs a correlation value 206.

  The hopping pattern detection unit 42 derives the relative sizes of the two types of correlation values, and selects one pattern based on the relative sizes. The reference code buffer 60 receives a signal sequence for identification that is a combination of one pattern. The matched filter 40 calculates one type of correlation value between the baseband signal 200 and the specifying signal sequence, and outputs a correlation value 206. The symbol timing detection unit 44 detects a start timing by detecting a peak having a relative magnitude based on the magnitude of the correlation value 206. A frequency hopping pattern is also derived based on the selected pattern.

  After detecting the start timing, the matched filter 40 sets the second preamble signal in the reference code buffer 60. The matched filter 40 calculates a correlation value between the baseband signal 200 and the second preamble signal and outputs the correlation value 206. The symbol timing detection unit 44 detects a peak of a relative magnitude based on the magnitude of the correlation value 206, and detects a switching timing.

  According to the embodiment of the present invention, since the signal obtained by converting the second preamble signal into the time domain is used for correlation processing, the switching timing can be detected not by detecting the sign inversion of the correlation value but by detecting the peak of the correlation value. . In addition, since the switching timing is detected by detecting the peak of the correlation value, the accuracy of frame synchronization can be improved. Further, although completion of frame synchronization is delayed by one symbol, an increase in circuit scale can be suppressed because the symbol buffer that is originally provided for FFT processing is shared.

  In addition, since the switching timing also appears at the timing when the start timing appears, the switching timing can be detected with high accuracy by detecting the switching timing while using the second preamble signal after the detection of the start timing. In addition, since the signal obtained by converting the second preamble signal into the time domain is used for correlation processing, the correlation value for detecting the switching timing with the same circuit configuration as that for calculating the correlation value for detecting the start timing. Can be calculated. In addition, an increase in circuit scale can be suppressed. In addition, since the start timing is detected from the correlation values detected a plurality of times, the detection accuracy of the start timing can be improved. In addition, since the switching timing is detected from the correlation value detected once, the delay due to the switching timing detection process can be reduced. Moreover, since the peak is detected by comparing the relative value with the threshold value, the influence of noise or the like can be reduced.

  In addition, the matched filter is divided and used to narrow down the pattern candidates by determining the relative level of the correlation value with a short correlation period, while gradually increasing the correlation period step by step. Since one preamble pattern is detected by the relative level determination, it is possible to realize a synchronization acquisition method with excellent detection accuracy and detection time while maintaining one matched filter resource. In addition, the circuit scale and power consumption can be reduced as compared with a method using a plurality of matched filters.

  The present invention has been described based on the embodiments. This embodiment is an exemplification, and it will be understood by those skilled in the art that various modifications can be made to combinations of the respective constituent elements and processing processes, and such modifications are also within the scope of the present invention. is there.

  In the embodiment of the present invention, the storage unit 56, the selection unit 58, and the reference code buffer 60 derive the identification signal sequence that combines the four types of preamble patterns, and then specify the identification signal sequence that combines the two types of preamble patterns. And a signal sequence for identification composed of one kind of preamble pattern. However, the present invention is not limited thereto, and for example, the preamble pattern may be specified by a pattern different from the embodiment by using the periodicity of the preamble pattern. That is, a group of a plurality of representative preamble patterns selected from a plurality of preamble patterns is set, one representative preamble pattern is selected by the same processing as the embodiment, and one representative One preamble pattern may be selected from a plurality of preamble patterns represented by a simple preamble pattern.

  Specifically, the first pattern and the second pattern are selected as representative patterns, and one of these is selected, and then a selection is made between the patterns represented by the selected pattern. For example, if the first pattern is selected, one of the first pattern and the fourth pattern is selected. In other words, after deriving a signal sequence for identification that combines two types of preamble patterns, a signal sequence for identification that combines two types of preamble patterns is derived again, and then a signal sequence for identification that consists of one type of preamble pattern is derived. To do. According to this modification, the number of preamble pattern samples combined with the signal sequence for identification in the first stage can be increased, so that the accuracy of the correlation value is increased. That is, one pattern may be selected step by step from a plurality of patterns.

In the embodiment of the present invention, the intensity calculator 90 derives the relative peak intensity using division. However, the present invention is not limited to this, and the relative peak intensity may be derived, for example, by subtraction. For example, the relative peak intensity is also shown as:
Relative peak intensity = max (AVE1, AVE2,..., AVE16) − ((AVE1 + AVE2 +... + AVE16) / 16) (3)
According to this modification, the circuit scale can be reduced. That is, it is only necessary to be able to derive a relative intensity that takes into account the influence of noise or the like.

  In the embodiment of the present invention, the synchronization acquisition unit 28 uses one matched filter 40. However, the present invention is not limited to this, and for example, two correlation periods may be used to double the correlation period. According to this modification, it is possible to improve accuracy. That is, the number of matched filter divisions may be optimized according to the correlation period (reference code length), the number of frequency hopping patterns, the time taken for the synchronization acquisition process, and the like.

  In the embodiment of the present invention, the matched filter 40 is used to detect the timing. However, the present invention is not limited to this. For example, a sliding correlator may be used. Since the processing is after symbol synchronization, the sliding correlator can be used by fixing the symbols. According to this modification, the matched filter 40 can be stopped during the switching timing detection period, so that power consumption can be reduced. Further, the operation period of the matched filter 40 may be limited to the vicinity of the symbol boundary. According to the present modification, during the switching timing detection period, the operation period of the matched filter 40 can be shortened, and the power consumption can be reduced.

It is a figure which shows the structure of the communication system which concerns on an Example. FIGS. 2A to 2E are diagrams illustrating a configuration of a burst format according to the embodiment. 3A to 3E are diagrams illustrating hopping frequencies and hopping patterns according to the embodiment. It is a figure which shows the structure of the synchronous acquisition part of FIG. It is a figure which shows the structure of the matched filter of FIG. It is a figure which shows the structure of the hopping pattern detection part of FIG. FIG. 5 is a diagram illustrating a configuration of a symbol timing detection unit in FIG. 4. It is a figure which shows the operation | movement timing of the synchronous acquisition part of FIG. It is a figure which shows the operation timing which further detailed the operation timing of FIG. 10A to 10D are diagrams illustrating correlation values calculated in the first stage in the hopping pattern detection unit of FIG. FIGS. 11A and 11B are diagrams illustrating correlation values calculated in the second stage in the hopping pattern detection unit of FIG. It is a figure which shows the correlation value calculated in the 3rd step in the hopping pattern detection part of FIG.

Explanation of symbols

  40 matched filter, 42 hopping pattern detection unit, 44 symbol timing detection unit, 46 synchronization control unit, 48 relative peak level calculation unit, 50 buffer, 52 multiplication unit, 54 addition unit, 56 storage unit, 58 selection unit, 60 reference Code buffer, 70 separation unit, 72 calculation unit, 74 determination unit, 76 addition unit, 78 first selector, 80 delay profile memory, 82 second selector, 84 switch, 86 delay profile buffer, 88 peak detection unit, 90 intensity calculation Unit, 92 averaging unit, 94 peak detection unit, 96 moving average unit, 98 determination unit, 100 communication system.

Claims (6)

A sequence of known single carrier signals configured by repetition of a predetermined signal pattern and a sequence of known multicarrier signals defined by the same bandwidth as the known single carrier signal sequence are arranged at least continuously. Receiving a received burst signal;
A matched filter for calculating a first correlation value between a burst signal received by the receiving unit and a predetermined signal pattern;
Based on the first correlation value calculated by the matched filter, a start timing when a predetermined signal pattern is repeated in a sequence of known single carrier signals arranged in the received burst signal is detected. A detection unit,
In the matched filter, when the detection timing is detected by the detection unit, calculation of a second correlation value between the burst signal received by the reception unit and a known multicarrier signal sequence is started,
The detection unit detects a timing of switching from a known single carrier signal sequence to a known multicarrier signal sequence based on the second correlation value calculated by the matched filter in the vicinity of the start timing. A timing detection apparatus characterized by:   The timing detection apparatus according to claim 1, wherein the matched filter uses a time-domain value for a known multi-carrier signal sequence when calculating the second correlation value.   The detection unit detects a peak for the first correlation value a plurality of times, detects a start timing based on the peak for the first correlation value detected a plurality of times, and a second correlation value Means for detecting one peak and detecting timing for switching from a known single carrier signal sequence to a known multicarrier signal sequence based on a peak corresponding to a second correlation value. The timing detection device according to claim 1 or 2.   When detecting the peak with respect to the first correlation value, the detection unit compares the ratio of the average value of the first correlation value and the peak with respect to the first correlation value with a threshold value, and sets the second correlation value. 4. The timing detection apparatus according to claim 3, wherein a ratio of an average value of the second correlation values and a peak of the second correlation value is compared with a threshold value when detecting a peak for. A sequence of known single carrier signals configured by repetition of a predetermined signal pattern and a sequence of known multicarrier signals defined by the same bandwidth as the known single carrier signal sequence are arranged at least continuously. Receiving a received burst signal;
A matched filter for calculating a first correlation value between a burst signal received by the receiving unit and a predetermined signal pattern;
Based on the first correlation value calculated by the matched filter, a start timing when a predetermined signal pattern is repeated in a sequence of known single carrier signals arranged in the received burst signal is detected. A detection unit;
A processing unit that performs a predetermined process on the burst signal whose start timing is detected by the detection unit;
In the matched filter, when the detection timing is detected by the detection unit, calculation of a second correlation value between the burst signal received by the reception unit and a known multicarrier signal sequence is started,
In the vicinity of the start timing, the detection unit determines a switching timing for switching from a known single carrier signal sequence to a known multicarrier signal sequence based on the second correlation value calculated by the matched filter. Detect and
The receiving device, wherein the processing unit executes processing on a multicarrier signal when switching timing is detected by the detecting unit. A sequence of known single carrier signals configured by repetition of a predetermined signal pattern and a sequence of known multicarrier signals defined by the same bandwidth as the known single carrier signal sequence are arranged at least continuously. Receiving a received burst signal;
Calculating a first correlation value between the received burst signal and a predetermined signal pattern;
Detecting a start timing when a predetermined signal pattern is repeated among a series of known single carrier signals arranged in a received burst signal based on the calculated first correlation value;
When a start timing is detected, calculating a second correlation value between the received burst signal and a sequence of known multi-carrier signals;
Detecting a timing of switching from a known single carrier signal sequence to a known multicarrier signal sequence based on the calculated second correlation value in the vicinity of the start timing;
A timing detection method comprising:

Images