JP2011101234A - Signal detector, signal detection method, processor for reception and receiver - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a signal detector which has a small circuit scale and a high accuracy by correctly distinguishing continuation noise and normal PDU (Protocol Data Unit) through an autocorrelation type operation logic. <P>SOLUTION: The signal detector 111 includes: a first calculating part 11 for calculating a first autocorrelation value C1(t) being a correlation value between an input signal d(t) and a first delay signal d(t-16×Δt) obtained by delaying the input signal by an integer multiple of one period of a predetermined repetition signal; a second calculating part 12 for calculating a second autocorrelation value C2(t) being a correlation value between the input signal d(t) and a second delay signal d(t-8×Δt) obtained by delaying the input signal by a time that is not an integer multiple of one period; and a determining part 13 for determining whether the input signal d(t) is the normal PDU on the basis of the first autocorrelation value C1(t) and the second autocorrelation value C2(t). <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a signal detector for determining whether or not an input signal is a normal protocol data unit (hereinafter referred to as “PDU”) based on the autocorrelation value of the input signal and the signal thereof. The present invention relates to a detection method, and a reception processor and receiver using the signal detector.

Conventionally, orthogonal frequency division multiplexing (OFDM) has been used as a communication method for terrestrial digital broadcasting, wireless LAN (Local Area Network) compliant with IEEE802.11a / g / p, cellular phones, power line modems, and the like. It is used.
In such OFDM, a large number of orthogonal subcarriers (carrier waves) arranged at predetermined frequency intervals are digitally modulated at respective frequencies and data is allocated, so that band utilization efficiency is improved as compared with normal frequency division multiplexing. It is improving.

This digital modulation and demodulation in OFDM is realized by executing IFFT (Inverse Fast Fourier Transform) processing on a plurality of symbol data on the transmitting side and executing FFT (Fast Fourier Transform) processing on the received signal on the receiving side. Is done.
At this time, since it is necessary for the receiver to determine the window timing when the FFT processing is performed on the received signal, the arrival of the packet signal (OFDM signal) by OFDM is determined, and the symbol timing of the signal is detected. There is a need (see Patent Document 1).

Therefore, for example, in IEEE802.11a / g / p, a pattern signal (repetitive signal) that repeats at a predetermined period called PLCP (Physical Layer Convergence Protocol) preamble is provided in the first part of the transmission packet signal, and the periodicity of this pattern signal is set. Is used to detect the arrival of the packet signal and the symbol timing on the receiving side.
As such detection methods, there are specifically known two types of autocorrelation type and cross-correlation type.

Among these, the autocorrelation type calculates an autocorrelation value between the received signal and a delayed signal obtained by delaying the received signal by the repetition interval (one period) of the pattern signal, and the peak time of the autocorrelation value (maximum autocorrelation value). The time corresponding to the point) is estimated as a temporary symbol timing.
On the other hand, the cross-correlation type calculates a cross-correlation value between a received signal and a preamble signal waveform (reference signal) stored in advance in the receiver, and estimates a peak time of the cross-correlation value as a temporary symbol timing. (See Non-Patent Document 1).

JP 2005-318512 A

OFDM / OFDMA textbook Takeshi Hattori, Impress R & D pages 173 to 177, pages 183 to 185

Since the autocorrelation type integrates the multiplication result of the received signal and its delayed signal, the correlation value is slightly more gradual and the timing detection accuracy is likely to deteriorate compared to the cross-correlation type. There is an advantage that it can be implemented.
In contrast, the cross-correlation type integrates the multiplication result of the received signal and the preamble signal (reference signal) stored in advance in the receiver, so that the timing detection accuracy is higher than that of the auto-correlation type. On the other hand, a multiplier for each sample is required, but there is a disadvantage that the circuit scale becomes large.

For this reason, the autocorrelation type is preferable from the viewpoint of reducing the manufacturing cost of the receiver. However, this autocorrelation type detection method has the following problems.
That is, in the natural world, continuous noise (for example, see the broken lines in FIGS. 5B and 5C) may occur in which a level close to a certain level continues for a longer time than the time of one cycle of the pattern signal. However, in the autocorrelation type, the autocorrelation value between the received signal and the delayed signal delayed by one cycle is calculated from the received signal. Therefore, even in the case of such continuous noise, the autocorrelation value becomes a large value exceeding the threshold value. May end up.

Therefore, although continuous noise has actually arrived, it may be mistaken for the arrival of a normal packet signal. In this case, since it takes time until it is determined as continuous noise in the subsequent digital demodulation, there is a possibility that the next regular packet signal that arrives during that time will be missed.
On the other hand, in the case of the cross-correlation type, since the correlation with the reference signal is taken, there is almost no such inconvenience.

  In view of the above-described problems, the present invention provides a signal detector having a small circuit scale and high accuracy so that continuous noise and regular PDU can be accurately distinguished by an autocorrelation type arithmetic logic. With the goal.

  (1) The signal detector of the present invention calculates a first autocorrelation value that is a correlation value between an input signal and a first delayed signal obtained by delaying the input signal by an integral multiple of one cycle of a predetermined repetitive signal. A second autocorrelation value that is a correlation value between the input signal and a second delayed signal obtained by delaying the input signal by a time that is not an integral multiple of one period of the repetitive signal; And a determination unit that determines whether or not the input signal is a regular PDU based on the first autocorrelation value and the second autocorrelation value.

According to the signal detector of the present invention, not only the first calculation unit calculates the first autocorrelation value but also the second calculation unit calculates the second autocorrelation value.
Since the second autocorrelation value is a correlation value between the input signal and the second delayed signal obtained by delaying the input signal by a time that is not an integral multiple of one cycle of the predetermined repetitive signal, the input signal is based on the continuous noise described above. If it is, it takes a large value, but becomes very small for a signal having a large change such as a preamble of 802.11a / g / p.

For this reason, in the determination unit, for example, by identifying that only the first autocorrelation value becomes a large value based on the first autocorrelation value and the second autocorrelation value, the input signal is a normal PDU. It is possible to determine whether or not continuous noise and regular PDU are more accurately distinguished.
Usually, “frame” is a PDU name used in layer 2 communication, and “packet” is often a name used in layer 3 communication. Unconsciously, both “packets” and “frames” are treated as just one type of PDU.

(2) In the signal detector according to the present invention, specifically, the determination unit determines the input signal when a ratio of the first autocorrelation value to the second autocorrelation value is greater than a predetermined threshold. The regular PDU can be determined.
The reason is that in the case of the first autocorrelation value, a peak is likely to occur for both the regular PDU and the continuous noise, and in the case of the second autocorrelation value, a peak is unlikely to occur for the regular PDU. This is because since a peak is likely to occur for continuous noise, it can be estimated that the input signal is a regular PDU when the ratio is larger than a predetermined threshold.

(3) In the signal detector of the present invention, the delay time for the input signals of the first and second delay signals is arbitrarily set, but the delay of the first and second delay signals for the input signal is arbitrary. As time increases, the timing at which each autocorrelation value is obtained is delayed, and detection of a regular PDU by the determination unit is delayed.
Therefore, in the signal detector of the present invention, the first delay signal is a signal obtained by delaying the input signal by one cycle of the repetitive signal, and the second delay signal is a signal obtained by delaying the input signal by 1 of the repetitive signal. The signal is preferably delayed by a time less than the period.

In the signal detector of the present invention, it is preferable that the second delayed signal is a signal obtained by delaying the input signal by a half cycle of the repetitive signal.
The reason is that as the delay time of the second delay signal is closer to an integral multiple of one cycle of the repetitive signal, the difference between both correlation values disappears, and conversely, the difference between both correlation values increases as the distance from the integral multiple increases. This is because, if the delay time of the second delay signal is set to a half cycle, it is easier to determine the continuous noise based on both correlation values.

(4) Furthermore, in the signal detector of the present invention, the second calculation unit can calculate a plurality of second autocorrelation values using the plurality of second delay signals having different delay times. Is preferred.
In this case, if the determination unit determines whether the input signal is a regular PDU based on the first autocorrelation value and the plurality of second autocorrelation values, the first Compared to the case where there is only one two autocorrelation value, it becomes possible to more accurately distinguish between continuous noise and regular PDUs.

(5) The signal detection method of the present invention is a regular PDU detection method performed by the signal detector of the present invention, and has the same effects as the signal detector.
(6) A receiving processor according to the present invention includes the signal detector according to the present invention and a demodulating unit that digitally demodulates the PDU determined to be legitimate, and has the same effects as the signal detector. Play.

  (7) Further, the radio signal receiver of the present invention includes a radio reception unit that receives an analog signal from an antenna, an A / D converter that converts the received analog signal into a digital signal, and the converted digital signal. The present invention includes a receiving processor of the present invention to which a signal is input, and has the same effects as the receiving processor.

  As described above, according to the present invention, since continuous noise and regular PDU can be accurately distinguished using autocorrelation type arithmetic logic, a highly accurate signal detector with a small circuit scale can be obtained.

It is a road top view for showing the whole structure of an intelligent road traffic system. It is a functional block diagram which shows the internal structure of the receiver which concerns on embodiment of this invention. It is a functional block diagram which shows the internal structure of the timing detection part of the said receiver. It is a graph which shows the time transition of the reception level (RSSI) of the said receiver, a 1st autocorrelation value, and a 2nd autocorrelation value when a regular packet comes after noise. (A) is a frame format of an OFDM signal, and (b) and (c) are waveform diagrams of a short preamble and continuous noise of the OFDM signal.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
In the present embodiment, a specific example of the present invention will be described on the assumption that the technique of the present invention is applied to a wireless communication system that constitutes an intelligent transport system (ITS). However, the present invention can also be applied to other communication systems other than the intelligent transportation system.

[Overall system configuration]
FIG. 1 is a road plan view showing the overall configuration of an intelligent road traffic system.
As shown in FIG. 1, the intelligent road traffic system of the present embodiment is composed of a roadside communication device 1 provided near an intersection and a plurality of in-vehicle communication devices 3 that can communicate with the device 1. A roadside communication device 1 shown in the vicinity of the center of 1 is installed, for example, on a column of a traffic light 2 at an intersection.

Each vehicle traveling on the road is equipped with an in-vehicle communication device 3 that can communicate with the roadside communication device 1. The roadside communication device 1 can communicate with a large number (for example, about 200) of the vehicle-mounted communication devices 3 in the communicable area.
The roadside communication device 1 is connected to the central device 4 of the traffic control center, and the central device 4 and the roadside communication device 1 are connected by wire (or wirelessly). Moreover, the road-to-road communication between the roadside communication apparatuses 1 located at each intersection, the road-to-vehicle and road-to-vehicle communication between the roadside communication apparatus 1 and the in-vehicle communication apparatus 3, and the inter-vehicle communication between the in-vehicle communication apparatuses 3 For this, wireless communication is used. Among these, it is presumed that CAMA / CA is used for inter-vehicle communication between the in-vehicle communication devices 3.

In this embodiment, the OFDM method is adopted as a modulation method for wireless communication between the communication apparatuses 1 and 3.
This method is a multi-carrier digital modulation method in which transmission data is carried on a large number of carrier waves (subcarriers). Since the subcarriers are orthogonal to each other, the data can be arranged so densely as to overlap on the frequency axis. Because there is an advantage.

Furthermore, the in-vehicle communication device 3 of the present embodiment transmits a radio signal in a format compliant with IEEE 802.11a / g / p using an access control method based on CSMA / CA.
For this reason, the regular frames (for example, about 1 msec in length) transmitted by each in-vehicle communication device 3 are transmitted in time series without overlapping each other, and for example, about 40 μsec is transmitted between the frames. There is a time gap. In principle, the number of frames transmitted from the transmitter of each in-vehicle communication device 3 is one except in the so-called burst mode.

[Receiver of roadside communication device]
Next, the internal configuration of the receiver mounted on the roadside communication device 1 will be described.
FIG. 2 is a functional block diagram showing an internal configuration of the receiver 1R according to the embodiment of the present invention.
In the following, for the sake of convenience, the description will proceed assuming that the roadside communication device 1 is the “receiver 1R” and the in-vehicle communication device 3 is the “transmitter 3s”, but the receiver 1R of the present embodiment is It can also be mounted on the in-vehicle communication device 3.

The receiver 1R of the roadside communication device 1 is configured by connecting the components illustrated in FIG. 2 as illustrated.
That is, the receiver 1R includes an antenna 101, a low-noise amplifier 102, a mixer 103, a variable amplifier 104 including an AGC circuit, an orthogonal detector 105, A / D converters 106 and 107, and a reception processor 110 in order from the left side. ing. Among these, the antenna 101, the low noise amplifier 102, the mixer 103, and the quadrature detector 105 constitute a wireless reception unit that receives an analog signal.

First, the radio signal (analog OFDM signal) transmitted from each transmitter 3S is received by the antenna 101, and this received signal is narrowed down to a predetermined band by a band pass filter (not shown) and is sent to the low-noise amplifier 102 at the subsequent stage. Entered.
The received signal input to the low noise amplifier 103 is amplified by the amplifier 102, then mixed by the mixer 103 at a frequency output from a radio frequency (RF) oscillator 108, further amplified by the variable amplifier 104, The signal is input to the subsequent quadrature detector 105.

The output signal of the variable amplifier 104 is quadrature detected at a frequency output from an intermediate frequency (IF) oscillator 109 by the quadrature detector 105 and demodulated into an in-phase component I signal and a quadrature component Q signal. Is done.
The signal of each component is converted into a digital signal by the A / D converters 106 and 107 in the next stage and input to the reception processor 110.

The receiving processor 110 according to the present embodiment is configured by, for example, an FPGA (Field Programmable Gate Array) having one or a plurality of memories and a CPU therein.
In the FPGA, configuration information for various logic circuits can be set (configured) in advance at the time of shipment of the processor, manufacture of the receiver 1R, and the like, and each function shown in FIG. Portions 111 to 120 are configured.

  That is, as shown in FIG. 2, the reception processor 110 of this embodiment includes, in order from the left, a timing detection unit 111, an AFC (Automatic Freqency Control) unit 112, a guard interval removal unit 113, a Fourier transform unit 114, and channel estimation Unit 115, channel equalization unit 116, phase rotation correction unit 117, subcarrier demodulation unit 118, deinterleave processing unit 119, and Viterbi decoding unit 120.

Among them, the timing detection unit 111 performs a short preamble SP (which is a repetitive signal at the head of the OFDM signal) with respect to the input signals (I component and Q component digital values) from the A / D converters 106 and 107. By executing the auto-correlation type arithmetic logic using FIG. 5A), the normal packet arrival of the OFDM signal and its symbol timing are detected. Details of the timing detection unit 111 will be described later.
When detecting a regular packet, the timing detection unit 111 outputs the timing information (symbol timing time information) St to the AFC unit 112 and the guard interval removal unit 113.

The AFC unit 112 roughly adjusts the carrier frequency error for the OFDM signal based on the timing information St from the timing detection unit 11, and the guard interval removal unit 113 detects the guard interval GI ( 5 (a)) is removed and input to the Fourier transform unit 114.
The Fourier transform unit 114 performs a fast Fourier transform process on the OFDM signal from which the guard interval GI is removed, converts the OFDM signal into a time domain signal, and outputs the signal.

An output signal from the frame conversion unit 114 is input to the channel estimation unit 115 and the channel equivalent unit 116.
The channel estimator 115 compares each subcarrier by comparing the long preamble LF (see FIG. 5A) included in the input signal with the reference signal of the long preamble LP without transmission path distortion stored in advance. The channel equalization unit 116 removes the estimated distortion from each subcarrier and equalizes it to the state at the time of transmission.

The output signal from the channel equalization unit 116 is input to the phase rotation correction unit 117 and the subcarrier demodulation unit 118.
The phase rotation correction unit 117 uses a known pilot signal (not shown) included in the OFDM signal to generate a steady phase rotation amount due to phase noise generated during frequency conversion in the transmitter 3S and the receiver 1R. Is detected and the rotation amount of each subcarrier is corrected, and the subcarrier demodulator 118 demodulates each corrected subcarrier into digital data.

  The demodulated digital data is returned to the same state as that at the time of transmission in the deinterleave processing unit 119, and then subjected to error correction decoding in the Viterbi decoding unit 120 and extracted as decoded data.

[OFDM signal frame format]
FIG. 5 (a) shows the frame format of an OFDM signal compliant with IEEE 802.11a / g / p, and FIGS. 5 (b) and 5 (c) show an example of a short preamble SP and continuous noise of the OFDM signal. It is a waveform diagram.
As shown in FIG. 5 (a), the frame of the OFDM signal is composed of the short preamble SP, guard interval GI, long preamble LP, guard interval GI, SIGNAL, guard interval GI, SIGNAL part, guard interval GI and DATA in order from the left. It consists of parts.

Among these, the short preamble SP is a fixed pattern signal (repetition signal) having a period of 1.6 μs, and the same signal is repeated 10 times. This short preamble SP is mainly used for detection of an OFDM signal, detection of its symbol timing, coarse adjustment of the AFC unit 112, and the like.
The long preamble LP is a fixed pattern signal (repetition signal) having a period of 3.2 μs, and the same signal is repeated twice. This long preamble LP is used for fine adjustment of a carrier frequency error, channel estimation for each subcarrier, and the like.

A signal from the ten short preambles SP to two long preambles LP across the guard interval GI is referred to as a “PLPC preamble”.
The SIGNAL part includes important physical layer header information such as the transmission rate and data length of subsequently transmitted data, and the DATA part contains actual data based on the transmission rate set by the SIGNAL part. Is included.

  The short preamble SP of the OFDM signal has a repetitive waveform of a predetermined pattern with a period of 1.6 μs as shown in FIGS. 5B and 5C. However, as shown by a broken line in FIG. May cause “continuous noise” in which the level continues almost constant in a time period longer than the period of 1.6 μs of the short preamble SP. The problem due to the continuous noise and the details of the solution will be described later.

[Configuration of timing detector]
FIG. 3 is a functional block diagram showing an internal configuration of the timing detection unit 111 in the receiver 1R of the present embodiment.
As shown in FIG. 3, the timing detection unit 111 according to the present embodiment mainly includes a first calculation unit 11, a second calculation unit 12, and a determination unit 13. Among these, the first calculation unit 11 includes a first delay unit 14, complex multipliers 15A and 15B, integration circuits 16A and 16B, and a divider 18 in this order from the left.

The second calculation unit 12 includes a second delay unit 20, complex multipliers 21A and 21B, integration circuits 22A and 22B, and a divider 24 in this order from the left.
In FIG. 3, d (t) is an input signal input from the A / D converters 106 and 107 to the timing detection unit 111, and an in-phase component (I signal) and a quadrature component with the discrete time t as a variable. It consists of a complex digital signal having (Q signal).

In the following, it is assumed that one cycle of the short preamble SP is T0, and the sampling cycle Δt in the A / D converters 106 and 107 is 1/16 of the cycle T0. That is, T0 = 16 · Δt.
Therefore, the input signal d (t) includes a total of 16 time-series discrete values within one period T0. Furthermore, in FIG. 3, the mark “*” represents a conjugate complex number.

In the first calculation unit 11, the input signal d (t) is input as it is to one complex multiplier 15A, and the input signal d (t) is input to the short preamble SP by the first delay unit 14. A first delay signal d (t-16 · Δt) delayed by one period T0 is input.
Therefore, the complex multiplier 15A multiplies the input signal d (t) by the complex conjugate of the first delay signal d (t-16 · Δt), and outputs the result to the integrating circuit 16A. The integrating circuit 16A integrates the result of the complex multiplication with the number of samples k = 0 to N−1 (in this case, N = 16). The result of this integration indicates a value corresponding to the power of the input signal d (t), and this is given to the subsequent divider 18.

  On the other hand, the input signal d (t) is input as it is without delay to the other complex multiplier 15B. The complex multiplier 15B multiplies the input signal d (t) by the complex conjugate of the input signal d (t) and outputs the result to the integration circuit 16B. The integrating circuit 16A integrates the result of the complex multiplication with the number of samples k = 0 to N−1. The result of this integration is also given to the subsequent divider 18.

  The divider 18 performs normalization by dividing the output result of the integration circuit 16A by the output result of the integration circuit 16B, and outputs the first autocorrelation value C1 (t) as a result to the determination unit 13. From the above calculation procedure, the first autocorrelation value C1 (t) is expressed by the following equation (1).

On the other hand, in the second calculation unit 12, the input signal d (t) is input as it is to one complex multiplier 22A, and the input signal d (t) is short-preamble by the second delay unit 20. The second delayed signal d (t−8 · Δt) delayed by the SP half cycle (0.5 · T0) is input.
Therefore, the complex multiplier 21A multiplies the complex conjugate of the input signal d (t) and the second delay signal d (t-8 · Δt), and outputs the result to the integrating circuit 16A. The integrating circuit 22A integrates the result of the complex multiplication with the number of samples k = 0 to N−1. The result of this integration is given to the divider 24 at the subsequent stage.

  On the other hand, the input signal d (t) is input as it is without delay to the other complex multiplier 21B. The complex multiplier 21B multiplies the input signal d (t) by the complex conjugate of the input signal d (t) and outputs the result to the integration circuit 22B. The integration circuit 22B integrates the result of the complex multiplication with the number of samples k = 0 to N-1. The result of this integration is also given to the subsequent divider 24.

  The divider 24 divides the output result of the integrating circuit 22A by the output result of the integrating circuit 22B, normalizes the result, and outputs the second autocorrelation value C2 (t) that is the result to the determining unit 13. From the above calculation procedure, the second autocorrelation value C2 (t) is expressed by the following equation (2).

The first autocorrelation value C1 (t) is obtained by shifting the input signal d (t) by the signal length (period T0) of the short preamble SP and multiplying the complex conjugate of the input signal d (t) by the original signal. This is a long-running moving average.
In this case, in the time zone in which the short preamble SP is repeatedly transmitted, the shifted signal d (t−16 · Δt) and the original input signal d (t) are the same signal, so the first autocorrelation value C1 (t) shows a large value in the time zone.

Thereafter, when the reception of the short preamble SP is completed and the time when the long preamble P is received, the periodicity of the short preamble SP is lost, so the first autocorrelation value C1 (t) does not show a large value.
Therefore, the arrival of the OFDM signal can be detected by determining whether or not the first autocorrelation value C1 (t) exceeds a predetermined threshold value. In addition, since the time (peak time) corresponding to the maximum point of the correlation value of the first autocorrelation value C1 (t) is at least within the period in which the short preamble SP is repeatedly transmitted, the peak time is a temporary symbol. It can be estimated that it is timing.

However, as described above, for example, “continuous noise” in which the level continues for a longer time than the cycle T0 (= 1.6 μs) of the short preamble SP, as shown by the broken lines in FIGS. May occur.
In this case, the first autocorrelation value C1 (t) is an autocorrelation value between the input signal d (t) and the first delay signal d (t−16 · Δt) delayed by one period T0. Even if the input signal is a continuous noise, a large value is generated.

  Therefore, if the OFDM signal is detected and the symbol timing is estimated using only the first autocorrelation value C1 (t), even if the received signal is continuous noise as shown in FIGS. 5B and 5C. There is a risk that this will be erroneously detected as a regular packet, and meaningless symbol timing that cannot be used for data demodulation will be generated.

On the other hand, the second autocorrelation value C2 (t) shifts the input signal d (t) by half the signal length of the short preamble SP (half cycle 0.5 · T0), and converts it into a complex conjugate thereof. This is obtained by multiplying the original signal and taking a moving average for the signal length of the preamble SP.
In this case, since the shifted signal d (t−8 · Δt) and the original input signal d (t) are not the same signal even in the time zone in which the short preamble SP is repeatedly transmitted, the second autocorrelation is performed. The value C2 (t) does not indicate a large value.

Further, since the second autocorrelation value C2 (t) is an autocorrelation value with the second delayed signal d (t−8 · Δt) delayed by a half cycle of the short preamble SP, the first autocorrelation value C1 ( As in the case of t), a large value is generated when the received signal is the continuous noise shown in FIGS. 5B and 5C.
FIG. 4 shows temporal transitions of the reception level (RSSI), the first autocorrelation value C1 (t), and the second autocorrelation value C2 (t) of the receiver 1R when a regular packet arrives after noise. It is a graph.

As indicated by ellipses A and B in FIG. 4, in the time zone in which the received signal is noise, the first autocorrelation value C1 (t) and the second autocorrelation value C2 (t) have substantially the same peak waveform at the same time. It can be estimated that this is the time when the above-mentioned “continuation noise” occurs.
On the other hand, as indicated by an ellipse C in FIG. 4, in the time zone in which the received signal is a regular packet, a high peak appears in the first autocorrelation value C1 (t) during the reception period of the short preamble SP. No peak appears in the two autocorrelation value C2 (t).

Thus, in the case of the first autocorrelation value C1 (t), a peak is likely to occur for both the normal packet and the continuous noise, but in the case of the second autocorrelation value C2 (t), It can be seen that peaks are unlikely to occur, but peaks are likely to occur for continuous noise.
Therefore, in the present embodiment, the determination unit 13 calculates a ratio of the first autocorrelation value C1 (t) to the second autocorrelation value C2 (t), and the input is performed when the ratio is larger than a predetermined threshold. It is determined that the signal d (t) is a regular packet (a regular OFDM signal).
Note that it is preferable to determine whether the packet is a regular packet or noise by comparing values obtained by further integrating the autocorrelation values C1 (t) and C2 (t) over a maximum of 10 periods, because accuracy is further improved.

  The determination unit 13 searches for the peak time point of the first autocorrelation value C1 (t) only when it is determined that the input signal d (t) is a normal packet (normal OFDM signal), and this peak time point The timing information St consisting of the time is generated, and this information St is notified to the AFC unit 112 and the guard interval removal unit 113.

[Effect of the timing detection unit of this embodiment]
As described above, according to the timing detection unit 111 of the present embodiment, in addition to the first autocorrelation value C1 (t), the input signal d (t) and the second signal obtained by delaying the input signal d (t) by a half cycle of the short preamble SP. A second autocorrelation value C2 (t) that is a correlation value with the delayed signal d (t−8 · Δt) is calculated, and based on these correlation values C1 (t) and C2 (t), the input signal d ( Since it is determined whether or not t) is a regular packet, continuous noise and regular packets can be more accurately distinguished.
Further, unlike the case of the cross-correlation type, it is not necessary to store the waveform of the short preamble SP as a reference signal, so that there is an advantage that the circuit scale can be reduced and it is economical.

[Other variations]
The embodiment disclosed this time is illustrative and not restrictive. The scope of the right of the present invention is defined by the claims, and includes all modifications within the scope equivalent to the configuration.
For example, in the above embodiment, the first delayed signal for obtaining the first autocorrelation value C1 (t) is the signal d (t−16 · Δt) obtained by delaying the input signal d (t) by one cycle of the short preamble SP. However, the first delay signal may be generated by being delayed by an integral multiple of one period T0, and the delay time of the first delay signal may be two or more periods.

  In the above embodiment, the second delayed signal for obtaining the second autocorrelation value C2 (t) is the signal d (t−8 · Δt) obtained by delaying the input signal d (t) by a half cycle of the short preamble SP. However, the second delay signal may be generated by being delayed by a time that is not an integral multiple of one cycle T0, and the delay time of the second delay signal is not limited to a half cycle.

Of course, the longer the delay time of the first delay signal or the second delay signal, the later the timing at which the first autocorrelation value C1 (t) or the second autocorrelation value C2 (t) is obtained. Will be delayed.
Therefore, in order to perform the packet detection at the determination unit 13 earliest, a signal obtained by delaying the input signal d (t) by one cycle T0 of the short preamble SP is used as the first delay signal as in the above embodiment. A signal obtained by delaying the signal d (t) by a time shorter than one cycle T0 of the short preamble SP is preferably used as the second delay signal.

In the above embodiment, only one second autocorrelation value C2 (t) is calculated, but a plurality of second delay signals (for example, d (t−4 · Δt) or d A plurality of second autocorrelation values C2 (t) may be calculated using (t-12 · Δt)).
In this case, the determination unit 13 obtains a ratio of the first autocorrelation value C1 (t) to each second autocorrelation value C2 (t), and inputs depending on whether each ratio is greater than a predetermined threshold value. If it is determined whether or not the signal (t) is a regular packet, it is possible to more accurately distinguish between the continuous noise and the regular packet than when the second autocorrelation value C2 (t) is only one.

Furthermore, in the above embodiment, the periodicity of the short preamble SP is used. However, the present invention can also be implemented using other repetitive signals (for example, the long preamble LP).
In the above-described embodiment, the OFDM receiver 1R is exemplified. However, the present invention is not limited to this, and the communication method is not particularly limited as long as autocorrelation type signal detection using a fixed pattern signal is performed. It is not limited.

11 First Calculation Unit 12 Second Calculation Unit 13 Determination Unit 3S Transmitter 110 Reception Processor 111 Timing Detection Unit (Signal Detector)
C1 (t) First autocorrelation value C2 (t) Second autocorrelation value d (t) Input signal d (t-16 · Δt) First delay signal d (t-8 · Δt) Second delay signal

Claims (7)

A first calculation unit that calculates a first autocorrelation value that is a correlation value between an input signal and a first delayed signal obtained by delaying the input signal by an integral multiple of one cycle of a predetermined repetition signal;
A second calculation unit that calculates a second autocorrelation value that is a correlation value between the input signal and a second delayed signal obtained by delaying the input signal by a time that is not an integral multiple of one cycle of the repetitive signal;
A determination unit configured to determine whether the input signal is a normal protocol data unit (hereinafter referred to as “PDU”) based on the first autocorrelation value and the second autocorrelation value; ,
A signal detector.   The signal detector according to claim 1, wherein the determination unit determines that the input signal is the regular PDU when a ratio of the first autocorrelation value to the second autocorrelation value is greater than a predetermined threshold. .   The first delay signal is a signal obtained by delaying the input signal by one cycle of the repetitive signal, and the second delay signal is a signal obtained by delaying the input signal by a time less than one cycle of the repetitive signal. The signal detector according to claim 1 or 2. The second calculation unit can calculate a plurality of the second autocorrelation values using the plurality of second delay signals having different delay times, respectively.
The said determination part determines whether the said input signal is regular PDU based on the said 1st autocorrelation value and several said 2nd autocorrelation value. A signal detector according to claim 1. Calculating a first autocorrelation value that is a correlation value between an input signal and a first delayed signal obtained by delaying the input signal by an integral multiple of one period of a predetermined repetitive signal;
Calculating a second autocorrelation value that is a correlation value between the input signal and a second delayed signal obtained by delaying the input signal by a time that is not an integral multiple of one cycle of the repetitive signal;
Determining whether the input signal is a regular PDU based on the first autocorrelation value and the second autocorrelation value;
A signal detection method comprising: The signal detector according to any one of claims 1 to 4, which determines whether or not the input signal is a regular PDU based on an autocorrelation value of the input signal;
A demodulator that digitally demodulates the PDU determined to be regular;
A receiving processor. A wireless receiver for receiving an analog signal from an antenna;
An A / D converter for converting the received analog signal into a digital signal;
The receiving processor according to claim 6, wherein the converted digital signal is input;
A radio signal receiver.

Images