Demodulator, demodulation method and demodulation program

Abstract

To suppress a demodulation error rate while utilizing a fractional spreading factor in LoRa modulation.SOLUTION: A demodulator (30) comprises a sampling unit (302), a correction unit (306), an estimation unit (307), and an output unit (308). The sampling unit performs oversampling on a received signal subjected to FSF-LoRa modulation that uses a predetermined spreading factor and a predetermined fractional coefficient, at a sampling interval obtained by dividing a chip length used in the FSF-LoRa modulation by a predetermined oversampling coefficient, thereby generating an equivalent low-pass signal. The correction unit (306) performs correction, which is based on the spreading factor, the fractional coefficient and the oversampling coefficient, on a de-spread DFT signal obtained by performing de-spreading processing and DFT processing on the equivalent low-pass signal, thereby generating a correction signal. The estimation unit (307) estimates a frequency index having a maximum amplitude among a plurality of frequency indices included in the correction signal as an index of a desired signal.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a demodulator, a demodulation method, and a demodulation program, which can be suitably used for demodulating, for example, LoRa-modulated signals. [Background technology]

[0002] With the development of IoT (Internet of Things), LPWAN (Low Power Wide Area Network), typified by LoRaWAN (Long Range Wide Area Network), is attracting attention. As a communication method for such networks, a modulation and demodulation method using chirp spread spectrum in the physical layer is known.

[0003] In a modulation scheme using chirp spread spectrum, arbitrary data can be expressed by combining multiple types of chirp symbols. Each chirp symbol changes frequency over time. The timing at which the frequency instantaneously switches from maximum to minimum (or from minimum to maximum) varies depending on the type of chirp symbol. Chirp spread spectrum enables communication over relatively long distances with relatively low power consumption. Note that chirp spread spectrum has a trade-off: the higher the spreading factor, the better the noise resistance, but the lower the transmission rate.

[0004] In standardized LoRaWAN, only integer values ​​within a specified range (for example, from 7 to 12) can be selected as the spreading factor. Under these conditions, when two transmission signals sent by two transmitters are simultaneously received by one receiver as a single received signal, it is known that the receiver can receive only one transmission signal if the spreading factors of each transmission signal are the same and the SIR (Signal-to-Interference Ratio) is a specified value (for example, 6 dB) or higher. In this case, if interference cancellation technology is used in the receiver, it will be possible to detect multiple transmission signals, but this will result in the receiver becoming significantly more complex.

[0005] Furthermore, under the above conditions, when two transmission signals transmitted by two transmitters are simultaneously received by one receiver as a single received signal, if the spreading factors of the transmission signals are different and the SIR exceeds a predetermined threshold, the receiver can receive both transmission signals. In this case, the transmission rate will be lower than when the minimum spreading factor is selected for both transmission signals.

[0006] In relation to the above, Non-Patent Document 1 (LoRa Alliance, Inc., "LoRaWAN™ 1.0.3 Specification", 2018, Internet<URL:https: / / lora-alliance.org / sites / default / files / 2018-07 / lorawan1.0.3.pdf> ) discloses the LoRaWAN specification version 1.0.3.

[0007] Furthermore, Non-Patent Document 2 (Saito Ryoya, Adachi Koichi, Fujii Takeo, Kumada Ryota, "Study on Fractional Spreading Factors in LoRa Signals and Proposal of a Demodulation Method," IEICE Technical Report, Kagawa, October 2023) discloses a demodulation method using chirp spread spectrum, which achieves an equivalent non-integer spreading factor. This demodulation method can improve the transmission rate by correcting the previously estimated frequency index taking into account the inclusion of a self-interference signal in the despread DFT signal. However, with this demodulation method, the demodulation error rate varies significantly depending on the symbol number of the desired signal. [Prior art documents] [Non-patent literature]

[0008] [Non-Patent Document 1] LoRa Alliance, Inc., "LoRaWAN(TM) 1.0.3 Specification," 2018, Internet<URL:https: / / lora-alliance.org / sites / default / files / 2018-07 / lorawan1.0.3.pdf> [Non-patent document 2] Ryoya Saito, Koichi Adachi, Takeo Fujii, Ryota Kumada, "Study on fractional spreading factor in LoRa signals and proposal of demodulation method," IEICE Technical Report, Kagawa, October 2023 Summary of the Invention [Problem to be solved by the invention]

[0009] In view of the above circumstances, one object of the present disclosure is to provide a demodulator, a demodulation method, and a demodulation program that suppress the demodulation error rate while using a fractional spreading factor for LoRa modulation. Other objects and novel features will become apparent from the description of this specification and the accompanying drawings. [Means for solving the problem]

[0010] The following describes the means for solving the problems using the numbers and symbols used in the description of the invention. These numbers and symbols are added in parentheses for reference purposes to show an example of the correspondence between the claims and the description of the invention. Therefore, the claims should not be interpreted as being limited by the parenthetical descriptions.

[0011] According to one embodiment, the demodulator (30) includes a sampling unit (302), a correction unit (306), an estimation unit (307), and an output unit (308). The sampling unit (302) oversamples a received signal that has been subjected to FSF-LoRa modulation using a predetermined spreading factor and a predetermined fractional coefficient, at a sampling interval obtained by dividing the chip length used in the FSF-LoRa modulation by a predetermined oversampling coefficient, to generate an equivalent low-pass signal. The correction unit (306) performs despreading and DFT processing on the equivalent low-pass signal to generate a despread DFT signal by performing correction based on the spreading factor, fractional coefficient, and oversampling coefficient, to generate a corrected signal. The estimation unit (307) estimates, from among multiple frequency indexes included in the corrected signal, the frequency index with the largest amplitude as the index of the desired signal. The output unit (308) outputs the desired signal demodulated based on the estimated index to the outside.

[0012] According to one embodiment, a demodulation method includes (S02) oversampling a received signal that has been subjected to FSF-LoRa modulation using a predetermined spreading factor and a predetermined fractional coefficient, using a sampling interval obtained by dividing the chip length used in the FSF-LoRa modulation by a predetermined oversampling factor. The demodulation method further includes (S05) generating a corrected signal by performing a correction based on the spreading factor, fractional coefficient, and oversampling factor on the despread DFT signal obtained by performing despreading and DFT processing on the equivalent lowpass signal. The demodulation method further includes (S06) estimating the frequency index with the largest amplitude from among multiple frequency indexes included in the corrected signal as the index of the desired signal, and (S07) externally outputting the demodulated desired signal based on the estimated index.

[0013] According to one embodiment, the demodulation program is executed by the arithmetic unit (32) to perform a predetermined process. This process includes oversampling a received signal that has been subjected to FSF-LoRa modulation using a predetermined spreading factor and a predetermined fractional coefficient, using a sampling interval obtained by dividing the chip length used in the FSF-LoRa modulation by a predetermined oversampling factor (S02), to generate an equivalent low-pass signal. This process further includes performing despreading and DFT processing on the equivalent low-pass signal to generate a despread DFT signal, performing correction based on the spreading factor, fractional coefficient, and oversampling factor, to generate a corrected signal (S05). This process further includes estimating the frequency index with the largest amplitude from among multiple frequency indexes included in the corrected signal as the index of the desired signal (S06), and outputting the desired signal demodulated based on the estimated index to the outside (S07). [Effects of the Invention]

[0014] According to one embodiment, the demodulation error rate can be reduced while utilizing fractional spreading factors for LoRa modulation. [Brief explanation of the drawings]

[0015] [Figure 1] FIG. 1 is a diagram illustrating an example of a configuration of a communication system according to an embodiment. [Figure 2] FIG. 2 is a block circuit diagram showing an example of the configuration of a receiver according to an embodiment. [Figure 3] FIG. 3 is a block circuit diagram showing an example of the configuration of a demodulator according to an embodiment. [Figure 4] FIG. 4 is a flowchart showing an example of a configuration of the process of the demodulation method according to an embodiment. [Figure 5] FIG. 5 is a diagram for explaining the LoRa signal and the FSF-LoRa signal. [Figure 6] FIG. 6 is a diagram for explaining the basic chirp signal. [Figure 7] FIG. 7 is a diagram for explaining the relationship between a desired signal and a self-interference signal. [Figure 8] FIG. 8 is a block circuit diagram showing an example of the configuration of a demodulator according to the related art. [Figure 9] FIG. 9 is a graph showing an example of the results of a computer simulation. [Figure 10] FIG. 10 is a partially enlarged view of FIG. [Figure 11] FIG. 11 is a graph showing an example of a comparison between a demodulator according to an embodiment and a demodulator according to the related art. [Figure 12] FIG. 12 is a diagram showing an example of the configuration of a communication system in a computer simulation. [Figure 13] FIG. 13 is a graph showing an example of the average SER from the results of the computer simulation of FIG. [Figure 14] FIG. 14 is a graph showing an example of throughput from the results of the computer simulation of FIG. DETAILED DESCRIPTION OF THE INVENTION

[0016] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS With reference to the accompanying drawings, embodiments for implementing a demodulator, a demodulation method, and a demodulation program according to the present disclosure will be described below.

[0017] (Embodiment) 1, a communication system 1 according to one embodiment includes a transmitter 2 that transmits a transmission signal and a receiver 3 that receives the transmission signal as a reception signal. The transmitter 2 includes a modulator 20 that modulates a desired signal into a transmission signal. The receiver 3 includes a demodulator 30 that demodulates the reception signal to generate a demodulated signal.

[0018] As shown in Fig. 2, the receiver 3 according to one embodiment may be configured like a so-called computer. In the example of Fig. 2, the receiver 3 includes a bus 31, an arithmetic unit 32, a storage device 33, a communication device 34, and an input / output device 35. As an example, the bus 31 is configured to connect the arithmetic unit 32, the storage device 33, the communication device 34, and the input / output device 35 so that they can communicate with each other.

[0019] The arithmetic device 32 includes a receiving unit 321, a demodulating unit 322, and an output unit 323. The storage device 33 includes a program storage unit 331 and a data storage unit 332. The arithmetic device 32 executes a receiver program stored in the program storage unit 331 to realize the processing of the receiving unit 321, the demodulating unit 322, and the output unit 323. The receiving unit 321, the demodulating unit 322, and the output unit 323 are virtual functional blocks that realize processing in cooperation with the arithmetic device 32 and the storage device 33. The receiving unit 321 receives a transmission signal from the transmitter 2 via the communication device 34 as a received signal. The demodulating unit 322 operates as a demodulator 30 that demodulates the received signal to generate a demodulated signal. The output unit 323 outputs the demodulated signal to the outside.

[0020] The receiver program may be read from the recording medium 330 and stored in the program storage unit 331. A part of the receiver program that realizes the processing of the demodulation unit 322 is called a demodulation program. The recording medium 330 may be a non-transitory and tangible medium.

[0021] The communication device 34 receives a transmission signal transmitted by a transmitting terminal as a received signal. The communication device 34 may be further connected to a network (not shown). A receiver program including a demodulation program may be obtained from an external source via the communication device 34 and stored in the program storage unit 331.

[0022] The input / output device 35 outputs information to the user and accepts operations input by the user. As an example, the input / output device 35 includes a display device that outputs images, a keyboard that accepts key input operations, and the like.

[0023] As shown in FIG. 3, the demodulator 30 according to one embodiment includes an input unit 301, a sampling unit 302, a basic chirp signal generating unit 303, a despreading unit 304, a DFT (Discrete Fourier Transform) unit 305, a correction unit 306, an estimation unit 307, and an output unit 308.

[0024] The input unit 301 supplies the received signal received by the receiving unit 321 in FIG. 2 to the sampling unit 302. The sampling unit 302 performs sampling processing on the received signal to generate an equivalent low-pass signal. The basic chirp signal generation unit 303 generates a basic chirp signal. The despreading unit 304 performs despreading processing on the equivalent low-pass signal using the basic chirp signal to generate a despread signal. The DFT unit 305 performs DFT processing on the despread signal to generate a despread DFT signal. The correction unit 306 corrects the despread DFT signal including the desired signal and the self-interference signal to generate a corrected signal. The estimation unit 307 estimates the index of the desired signal from multiple frequency indexes included in the corrected signal. The output unit 308 outputs the signal demodulated based on the estimated index to the outside as a demodulated signal.

[0025] The process of the demodulation method according to one embodiment will be described with reference to the flowchart of Fig. 4. The process of the flowchart of Fig. 4 may start when the receiver 3 starts up. When the process of the flowchart of Fig. 4 starts, step S01 is executed.

[0026] In step S01 in Fig. 4, the input unit 301 in Fig. 3 supplies a received signal to the sampling unit 302. More specifically, the received signal has been subjected to FSF (Fractional Spreading Factor)-LoRa (Long Range) modulation.

[0027] After step S01 in Fig. 4, step S02 is executed. In step S02, the sampling unit 302 in Fig. 3 performs oversampling processing on the received signal to generate an equivalent low-pass signal. Oversampling processing is sampling processing using a sampling period shorter than the period corresponding to the frequency of the received signal.

[0028] Step S03 is executed after step S02 in Fig. 4. In step S03, the despreading unit 304 in Fig. 3 performs despreading processing on the equivalent low-pass signal using the basic chirp signal generated by the basic chirp signal generating unit 303 to generate a despread signal.

[0029] Step S04 is executed after step S03 in Fig. 4. In step S04, DFT unit 305 in Fig. 3 performs DFT processing on the despread signal to generate a despread DFT signal.

[0030] Step S05 is executed after step S04 in Fig. 4. In step S05, the corrector 306 in Fig. 3 corrects the inverse-corrected DFT signal to generate a corrected signal.

[0031] After step S05 in Fig. 4, step S06 is executed. In step S06, the estimation unit 307 in Fig. 3 estimates an index of the desired signal from among the multiple frequency indexes included in the corrected signal. More specifically, the estimation unit 307 estimates the frequency index with the largest amplitude from among the multiple frequency indexes included in the corrected signal as the index of the desired signal.

[0032] Step S07 is executed after step S06 in Fig. 4. In step S07, output unit 308 in Fig. 3 outputs the demodulated signal demodulated based on the estimated index to the outside as a desired signal.

[0033] After step S07 in FIG. 4, the processing of the flowchart ends.

[0034] LoRa modulation and FSF-LoRa modulation will be described in relation to step S01 in Fig. 4. In LoRa modulation, a LoRa signal as shown in Fig. 5 is generated. The LoRa signal in the example of Fig. 5 is a combination of up-chirp signals whose frequency increases linearly over time. The LoRa signal is generated by combining a number of symbols, the total number of which is equal to 2 raised to the power of the spreading factor s, obtained by modifying the basic up-chirp signal. The length of each symbol on the time axis is the same symbol length T s The symbol length T s The length obtained by dividing by the power of 2 to the diffusion rate s is the tip length T c The frequency length of the range in which the frequency of each symbol changes is the same bandwidth W, and the chip length T c Each of the 2 to the power of spreading factor s types of symbols is distinguished by the timing at which the frequency switches from the maximum value to the minimum value of the bandwidth W. Therefore, one symbol included in a LoRa signal can represent any of the 2 to the power of spreading factor s types of data.

[0035] In the standardized LoRaWAN (Long Range Wide Area Network), an integer between 7 and 12 is used as the spreading factor s. However, in the example of FIG. 5, for easier understanding, the case where the spreading factor s is equal to 2 will be described.

[0036] In one embodiment, instead of a LoRa signal generated based on a spreading factor s, an FSF-LoRa signal as shown in Fig. 5 is used as a transmission signal and a reception signal. The FSF-LoRa signal is generated based on a fractional spreading factor, which is an extension of the spreading factor s in the LoRa signal to a fraction other than an integer. More specifically, s The chip length T for each of the symbols c and symbol length T s are multiplied by a factor 1+ρ, where the value ρ is called the fractional factor. The fractional factor ρ is greater than or equal to 0 and less than 1. Here, the factor 1+ρ may be a rational number expressed as a ratio of two integers. That is, the chip length T c and symbol length T s In the FSF-LoRa signal, which is a LoRa signal using the c (s,ρ) is (1+ρ)T c and the symbol length T s (s,ρ) is (1+ρ)T s is.

[0037] As another example, the LoRa signal and the FSF-LoRa signal may be a combination of down-chirp signals whose frequency decreases linearly over time.

[0038] The oversampling process will be described with reference to step S02 in FIG. 4. In the oversampling process in step S02, the number of samples (2 s ) multiplied by the oversampling coefficient q, s) is used to perform sampling processing of the FSF-LoRa signal. In addition, the oversampling period used in oversampling is the sampling period ((1+ρ)T c ) divided by the oversampling factor q (((1+ρ) / q)T c )

[0039] The equivalent low-pass discrete-time signal obtained as a result of the oversampling process is given by the following formula 1.

number

[0040] The basic chirp signal used in the despreading process will be described in relation to step S03 in Fig. 4. When the received signal is a combination of up-chirp signals, as in the example of Fig. 5, a down-chirp signal as shown in Fig. 6 is used as the basic chirp signal for the despreading process. Conversely, when the received signal is a combination of down-chirp signals, an up-chirp signal is used as the basic chirp signal for the despreading process.

[0041] The DFT process will be described with reference to step S04 in Fig. 4. The despread DFT signal is calculated as shown in the following "Equation 2".

number

[0042] Figure 7 shows the absolute amplitude of the despread DFT signal |d[n] m 7 is a graph showing an example of |d[n]|. In the example of FIG. 7, the spreading factor s is 7, the fractional coefficient ρ is 0.25, the oversampling coefficient q is 2, and the symbol number m of the desired signal is 10. In this case, the absolute amplitude |d[n] of the despread DFT signal m Three local peaks P1, P2, and P3 are observed in the graph of |. These local peaks P1, P2, and P3 include the desired signal and self-interference signals that occur because the oversampling process does not satisfy the Nyquist criterion. The first local peak P1 appears at the frequency index n=10 and corresponds to the desired signal with symbol number m=10. The second local peak P2 appears at the frequency index n=77 and corresponds to the first self-interference signal. The third local peak P3 appears at the frequency index n=205 and corresponds to the second self-interference signal.

[0043] In one embodiment, the relative distance from the frequency index of the first local peak P1 corresponding to the desired signal to the frequency index of the second local peak P2 corresponding to the first self-interference signal is calculated as shown in the following formula 3.

number

[0044] Furthermore, the relative distance from the frequency index of the first local peak P1 corresponding to the desired signal to the frequency index of the third local peak P3 corresponding to the second self-interference signal is calculated as shown in the following "Equation 4".

number

[0045] In this way, the relative distance Δn from the frequency index of the first local peak P1 corresponding to the desired signal to the frequency index of the second local peak P2 corresponding to the first self-interference signal si,1 and the relative distance Δn from the frequency index of the first local peak P1 corresponding to the desired signal to the frequency index of the third local peak P3 corresponding to the second self-interference signal. si,2 are calculated as functions of the oversampling factor q, the spreading factor s, and the fractional factor ρ.

[0046] The correction process will be described in relation to step S05 in Fig. 4. In the subsequent step S06, the absolute amplitude |d[n] of the despread DFT signal is calculated. mThe frequency index n for which | is largest is estimated as the index of the desired signal. Therefore, in one embodiment, in step S05, a correction process is performed so that the absolute amplitude of the first local peak P1 corresponding to the desired signal is reliably larger than the absolute amplitudes of the local peaks P2 and P3 corresponding to the self-interference signal. However, since the frequency index of the desired signal is unknown at the time of step S05, the relative distance Δn si,1 Specifically, the absolute amplitude of the corrected despread DFT signal is calculated as in the following equation (5).

number

[0047] As described above, the relative distance Δn si,2 is also calculated as a function of the oversampling factor q, the spreading factor s, and the fractional factor ρ, but the second self-interference signal is not used in the correction process because it can be considered as a mirroring of the desired signal.

[0048] The estimation process will be described with reference to step S06 in Fig. 4. As described above, the absolute amplitude |d[n] of the despread DFT signal m The frequency index n at which | is maximum is estimated as the index of the desired signal. Specifically, the index of the desired signal is calculated as shown in the following "Equation 6".

number

[0049] The results of a computer simulation performed under the following conditions are described below. In this computer simulation, a communication system 1 includes one transmitter 2 and one receiver 3, as shown in FIG. 1. The transmitter 2 transmits a single FSF-LoRa signal in an AWGN (Additive White Gaussian Noise) channel. The bandwidth W is 125 kHz, the spreading factor s is 7, the fractional coefficient ρ is 0.25, and the oversampling coefficient q is 2. Under the above conditions, a computer simulation was performed to compare the results of demodulation using three types of demodulators.

[0050] The first demodulator 40 is disclosed in Non-Patent Document 2 (Ryoya Saito, Koichi Adachi, Takeo Fujii, Ryota Kumada, "Study on fractional spreading factor in LoRa signal and proposal of demodulation method", IEICE Technical Report, Kagawa, October 2023), and is configured as shown in FIG. 8. The demodulator 40 in FIG. 8 differs from the demodulator 30 in FIG. 3 in the following respects. That is, the correction unit 306 in FIG. 3 is replaced by a correction unit 406 in FIG. 8. Also, the estimation unit 307 in FIG. 8 calculates the absolute amplitude d[n] of the despread DFT signal output by the DFT unit 305. m 8 is supplied with the absolute amplitude d[n] of the despread DFT signal output by the DFT unit 305, and outputs the frequency index with the largest absolute amplitude as the index of the desired signal. m and the estimation result output by the estimation unit 307, and the absolute amplitude d[n] mThe estimation result is corrected for the influence of the self-interference signal based on the above.

[0051] The second demodulator 30 is a demodulator 30 according to one embodiment and is configured as shown in Fig. 3. The third demodulator is a full-search demodulator that demodulates each of all frequency indexes included in the despread DFT signal by correlating it with all other frequency indexes. Therefore, the third demodulator has the theoretically lowest SER (Symbol Error Rate), which represents the rate at which the estimation result of the symbol number of the demodulated signal is incorrect.

[0052] FIG. 9 shows the results of a computer simulation of the symbol error rate characteristics of the first demodulator 40, the second demodulator 30, and the third demodulator. FIG. 10 is a partially enlarged view of FIG. 9. Each of FIG. 9 and FIG. 10 includes a total of three graphs G11, G12, and G13. In common with the graphs G11, G12, and G13, the horizontal axis represents the average SNR (Signal to Noise Ratio), and the vertical axis represents the average SER. The first graph G11 corresponds to the first demodulator 40, the second graph G12 corresponds to the second demodulator 30, and the third graph G13 corresponds to the third demodulator. As shown in FIG. 9 and FIG. 10, when the average SER is 10 -4 , the average SNR of the demodulator 30 according to the embodiment is about 1.5 dB lower than the average SNR of the demodulator 40 according to the related art, and is only about 1.8 dB closer to the theoretically lowest average SNR of the third demodulator.

[0053] FIG. 11 is a graph comparing the SER for each symbol number m between a first demodulator 40 according to the related art and a second demodulator 30 according to one embodiment. FIG. 11 includes two graphs, G21 and G22. In both graphs G21 and G22, the horizontal axis represents symbol number m and the vertical axis represents SER. The first graph G21 corresponds to the first demodulator 40 according to the related art, and the second graph G22 corresponds to the second demodulator 30 according to one embodiment. As shown in FIG. 11, the SER of the first demodulator 40 varies significantly depending on the symbol number m, rising to a maximum of approximately 0.0030, while the SER of the second demodulator 30 is more uniform with respect to the symbol number m, rising only to a maximum of approximately 0.0010.

[0054] Thus, the second demodulator 30 according to the embodiment has better SER characteristics than the first demodulator 40 according to the related art.

[0055] The results of a computer simulation performed under the following conditions are described below. In this computer simulation, as shown in FIG. 12, a communication system 1B includes a transmitter 2A having a modulator 20A, a transmitter 2B having a modulator 20B, and a receiver 3 having a demodulator 30. The modulators 20A and 20B have the same configuration as the modulator 20 shown in FIG. 1. The transmitters 2A and 2B have the same configuration as the transmitter 2 shown in FIG. 1, and transmit LoRa signals or FSF-LoRa signals in an AWGN (Additive White Gaussian Noise) channel in an environment where signals interfere with each other. Here, the SIR (Signal to Interference power Ratio) as seen from the receiver is 0.0 dB. Under the above conditions, a computer simulation was performed to compare the results of demodulation using three different combinations of transmission signals transmitted by transmitters 2A and 2B.

[0056] In the first combination, LoRa signals with the same spreading factor s are used. That is, both transmitters 2A and 2B transmit LoRa signals with a spreading factor s of 7. Since this LoRa signal is not an FSF-LoRa signal, the fractional coefficient ρ is 0 and the oversampling coefficient q is 1.

[0057] The second combination uses LoRa signals with different spreading factors s: transmitter 2A transmits a LoRa signal with spreading factor s of 7, and transmitter 2B transmits a LoRa signal with spreading factor s of 8. Since neither of these LoRa signals is an FSF-LoRa signal, the fractional coefficient ρ is 0 and the oversampling coefficient q is 1.

[0058] The third combination uses a LoRa signal and an FSF-LoRa signal with the same spreading factor s. That is, transmitter 2A transmits a LoRa signal with a spreading factor s of 7, and transmitter 2B transmits an FSF-LoRa signal with a spreading factor s of 7, a fractional coefficient ρ of 0.25, and an oversampling coefficient q of 2. Note that because the LoRa signal transmitted by transmitter 2A is not an FSF-LoRa signal, the fractional coefficient ρ is 0 and the oversampling coefficient q is 1.

[0059] FIG. 13 is a graph comparing the average SNR and average SER of the demodulation results using the three combinations of transmitted signals. The graph in FIG. 13 includes five graphs G31, G32, G33, G34, and G35. In all five graphs G31, G32, G33, G34, and G35, the horizontal axis represents the average SNR and the vertical axis represents the average SER. The first graph G31 corresponds to the LoRa signal in the first combination. In the first combination, transmitters 2A and 2B transmit the same LoRa signal, so the SER characteristics resulting from the demodulation are the same. The second graph G32 corresponds to the LoRa signal transmitted from transmitter 2A in the second combination. The third graph G33 corresponds to the LoRa signal transmitted from transmitter 2B in the second combination. The fourth graph G34 corresponds to the LoRa signal transmitted from transmitter 2A in the third combination. The fifth graph G35 corresponds to the FSF-LoRa signal transmitted from the transmitter 2B in the third combination.

[0060] As can be seen from Figure 13, the average SER for the first combination is approximately 4.10 -1 In the second combination, the average SNR did not exceed approximately -4.8 dB. In the third combination, the average SER was approximately 9.10 -4 The average SER and SNR were both lower than the average SER and SNR, and the average SNR was higher than about -3.8. Thus, transmitting the FSF-LoRa signal simultaneously with the LoRa signal provides superior performance in terms of both average SER and average SNR compared to transmitting two LoRa signals simultaneously.

[0061] Fig. 14 is a graph comparing the throughput of communication system 1B among the demodulation results using the above three combinations of transmission signals. The graph in Fig. 14 includes a total of three graphs G41, G42, and G43. In all three graphs G41, G42, and G43, the horizontal axis represents the average SNR and the vertical axis represents the throughput of communication system 1B. The first graph G41 corresponds to the first combination, the second graph G42 corresponds to the second combination, and the third graph G43 corresponds to the third combination.

[0062] As can be seen from Figure 14, with the first combination, the throughput did not exceed approximately 5,000 bps. With the second combination, the throughput exceeded that of the third combination when the average SNR was approximately -10 dB or less, but it did not exceed approximately 10,800 bps. With the third combination, the throughput exceeded that of the second combination when the average SNR was approximately -10 dB or more, and in the range where the average SNR was approximately -6 dB or more, the throughput was approximately 14.5% higher than that of the second combination, demonstrating excellent performance.

[0063] As described above, the demodulator 30, demodulation method, and demodulation program according to one embodiment can suppress the demodulation error rate while using FSF-LoRa modulation, which uses a dispersion spreading factor for LoRa modulation.

[0064] The invention made by the inventor has been specifically described above based on the embodiments, but it goes without saying that the present invention is not limited to the embodiments and can be modified in various ways without departing from the spirit of the invention. Furthermore, the features described in the embodiments can be freely combined within the scope of technical compatibility. [Explanation of symbols]

[0065] 1, 1B communication system 2, 2A, 2B transmitter 20, 20A, 20B Modulators 3 Receiver 30 Demodulator 301 Input section 302 Sampling section 303 Basic chirp signal generator 304 Despreading Unit 305 DFT section 306 Correction Unit 307 Estimation Department 308 Output section 31 Bus 32 Arithmetic unit 321 Receiving Unit 322 Demodulation Unit 323 Output Section 33 Storage device 330 Recording Media 331 Program Memory Unit 332 Data storage unit 34 Communication equipment 35 Input / Output Devices 40 Demodulator 406 Correction Unit d[n] m Absolute amplitude of the despread DFT signal G11, G12, G13, G21, G22, G31, G32, G33, G34, G35, G41, G42, G43 graphs P1, P2, P3 local peaks T c Tip Length T s Symbol Length W Bandwidth

Claims

1. a sampling unit that generates an equivalent low-pass signal by oversampling a received signal that has been subjected to FSF (Fractional Spreading Factor)-LoRa (Long Range) modulation using a predetermined spreading factor and a predetermined fractional coefficient, at a sampling interval obtained by dividing the chip length used in the FSF-LoRa modulation by a predetermined oversampling coefficient; a correction unit that performs a correction based on the spreading factor, the fractional coefficient, and the oversampling coefficient on a despread DFT signal obtained by performing a despreading process and a DFT (Discrete Fourier Transform) process on the equivalent low-pass signal to generate a corrected signal; an estimation unit that estimates a frequency index having a maximum amplitude from among a plurality of frequency indexes included in the correction signal as an index of a desired signal; an output unit that outputs the desired signal demodulated based on the estimated index to an external device; Equipped with Demodulator.

2. 2. The demodulator of claim 1, the correction unit calculates, as a correction value of the amplitude corresponding to each of the frequency indexes, a sum of a first absolute value of an amplitude corresponding to each of the frequency indexes included in the despread DFT signal and a second absolute value of an amplitude corresponding to each of the frequency indexes obtained by adding a predetermined relative distance to each of the frequency indexes. Demodulator.

3. 3. The demodulator of claim 2, The correction unit calculates the relative distance based on the spreading factor, the fractional coefficient, and the oversampling coefficient. Demodulator.

4. 2. The demodulator of claim 1, a despreading unit that performs despreading processing on the equivalent low-pass signal based on a fundamental down-chirp signal having the chip length to generate a despread signal; a DFT unit that performs DFT processing on the despread signal to generate the despread DFT signal; Further equipped Demodulator.

5. 2. The demodulator of claim 1, the fractional coefficient is greater than 0 and less than 1; Demodulator.

6. 6. The demodulator of claim 5, the fractional coefficients are rational numbers; Demodulator.

7. 2. The demodulator of claim 1, the oversampling factor is an integer greater than or equal to 2; Demodulator.

8. generating an equivalent low-pass signal by oversampling a received signal that has been subjected to FSF-LoRa modulation using a predetermined spreading factor and a predetermined fractional coefficient, using a sampling interval obtained by dividing the chip length used in the FSF-LoRa modulation by a predetermined oversampling coefficient; generating a corrected signal by performing correction based on the spreading factor, the fractional coefficient, and the oversampling coefficient on a despread DFT signal obtained by performing despreading processing and DFT processing on the equivalent low-pass signal; estimating a frequency index having a maximum amplitude from among a plurality of frequency indexes included in the correction signal as an index of a desired signal; outputting the desired signal demodulated based on the estimated index to an external device; Including, Demodulation method.

9. A demodulation program for realizing a predetermined process by being executed by a computing device, The process comprises: generating an equivalent low-pass signal by oversampling a received signal that has been subjected to FSF-LoRa modulation using a predetermined spreading factor and a predetermined fractional coefficient, using a sampling interval obtained by dividing the chip length used in the FSF-LoRa modulation by a predetermined oversampling coefficient; generating a corrected signal by performing correction based on the spreading factor, the fractional coefficient, and the oversampling coefficient on a despread DFT signal obtained by performing despreading processing and DFT processing on the equivalent low-pass signal; estimating a frequency index having a maximum amplitude from among a plurality of frequency indexes included in the correction signal as an index of a desired signal; outputting the desired signal demodulated based on the estimated index to an external device; Including, Recovery program.

Images