Underwater OFDM communication system and communication method

Abstract

To provide an underwater OFDM communication system and a communication method that improve the performance of acoustic wireless communication by performing processing to remove inter-carrier interference on received signals in which the inter-carrier interference has occurred due to disturbance in orthogonality between subcarrier signals that make up an OFDM signal, during OFDM reception processing in reception at a mobile station in a multipath environment with different arrival directions (angles).SOLUTION: An underwater OFDM communication system and a communication method that perform OFDM signal processing in a multipath environment in which OFDM signals transmitted from a transmitting unit that transmits an acoustic signal propagates through two or more different transmission paths underwater and the signals from the different transmission paths are received by a receiving unit in a superimposed state. The OFDM communication system includes means (ADC) for converting the received superimposed signals from an analog signal to a digital numeric sequence at a predetermined sampling frequency Fs, and means (3rdFFT) for performing a discrete Fourier transform on the digital numeric sequence, converting the signal into a frequency domain signal, and analyzing the frequency domain signal.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to an acoustic wireless communication system and a communication method using an orthogonal frequency division multiplexing (OFDM) method as wireless communication in water.

Background Art

[0002] In conventional acoustic wireless communication using the OFDM method in water, when comparing the utilized acoustic bandwidth B (Hz) with the center frequency fc (Hz) of the utilized acoustic signal, B << fc, that is, the condition that fc is sufficiently larger than B does not hold. Under such circumstances, when the transmitter or the receiver, or both, move, a so-called Doppler phenomenon (Doppler frequency shift phenomenon) occurs, and simply performing phase correction on the receiver side cannot compensate for the influence of the Doppler phenomenon. Therefore, in order to compensate for this Doppler phenomenon, a method of performing stretching and shrinking processing on the received signal in the time direction is known.

[0003] However, in a multipath environment where an acoustic OFDM wireless communication using the OFDM method in water (hereinafter referred to as an "OFDM signal") propagates through a plurality of different paths and is received by a receiver, the OFDM signal received by the receiver arrives from different directions (angles), and there is a problem that the amount of variation due to Doppler shift is different. This is because the OFDM signal received by the receiver is a superposition of a plurality of OFDM signals arriving from different paths. Therefore, a method of detecting the Doppler shift of the OFDM signal received by the receiver based on the temporal change of a part of the subcarrier signals constituting the OFDM signal disclosed in Patent Document 1 and correcting the stretching and shrinking of the OFDM signal is considered as a solution.

[0004] However, if correction is made to the direct wave, which is the largest component among the superimposed received signals, the expansion / contraction coefficient cannot be appropriately corrected because the amount of expansion / contraction differs for multiple OFDM signals arriving from different directions (angles). In other words, in a multipath environment where an OFDM signal propagates through multiple different paths and arrives at a receiver from different directions (angles), the stretching and contraction processing of the OFDM signal disclosed in Patent Document 1 alone is insufficient, and as a result, there is a problem that inter-carrier interference occurs during demodulation processing of the OFDM signal, resulting in an increase in the bit error rate during demodulation processing. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP2017-195566A Summary of the Invention [Problem to be solved by the invention]

[0006] Therefore, an objective of the present invention is to provide an underwater OFDM communications system and communication method which, with regard to OFDM signal processing in a multipath environment in which an OFDM signal propagates through a number of different paths and arrives at a receiver from different directions (angles), detects Doppler shift from temporal changes in some of the subcarrier signals that make up the OFDM signal, and during demodulation processing of the OFDM signal after correcting for expansion and contraction, performs processing to remove inter-carrier interference on received signals in which inter-carrier interference has occurred due to disruption of orthogonality between the subcarrier signals that make up the OFDM signal, thereby improving the performance of acoustic wireless communications. [Means for solving the problem]

[0007] The underwater OFDM communication system according to the present invention comprises: The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication system that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A means for converting the superimposed received signal from an analog signal into a digital number sequence, which is a digital signal, at a predetermined sampling frequency; means for performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; means for analyzing said frequency domain signal; It consists of: The means for analyzing the frequency domain signal includes: For each propagation wave in the transmission path, four parameters, namely, a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, are estimated; Using the parameters, a transmission path transfer function including interference between complex symbols of the frequency domain signal is estimated for each of the transmission paths, and an inter-carrier interference cancellation coefficient is calculated; A function for removing inter-carrier interference based on the inter-carrier interference removal coefficient; having It is characterized by:

[0008] The underwater OFDM communication method according to the present invention comprises: The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication method that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A step of converting the superimposed received signal from an analog signal to a digital number sequence, which is a digital signal, at a predetermined sampling frequency; performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; analyzing the frequency domain signal; It consists of: The step of analyzing the frequency domain signal includes: For each propagation wave of the transmission line, four parameters, namely, the complex attenuation coefficient, the relative delay time, the Doppler frequency shift amount, and the sampling clock error, are estimated, Using the parameters, for each transmission line, a transmission line transfer function including interference between complex symbols of the frequency domain signal is estimated to calculate an inter-carrier interference cancellation coefficient, Based on the inter-carrier interference cancellation coefficient, inter-carrier interference is removed, which is characterized by the above.

Advantages of the Invention

[0009] According to the underwater OFDM communication system and communication method of the present invention, even when the acoustic bandwidth B (Hz) used and the center frequency fc (Hz) of the acoustic signal used are compared and the condition B << fc, that is, the condition that fc is sufficiently larger than B is not satisfied, inter-carrier interference can be removed. Even in a multipath environment where the OFDM signal propagates through a plurality of different paths and arrives at the receiver from different directions (angles), the bit error rate during the demodulation process of the OFDM signal on the receiver side can be reduced, and higher-performance acoustic wireless communication can be realized.

Brief Description of the Drawings

[0010] [Figure 1] Schematic configuration diagram of an embodiment of an underwater OFDM communication system and communication method [Diagram 2] Block diagram showing the transmission and reception processing procedures in an underwater OFDM communication system and communication method [Diagram 3] Diagram representing a plurality of OFDM signals in the time direction in a two-dimensional plane with the horizontal axis being time and the vertical axis being frequency, and arranging the complex-modulated sub-carrier symbols with circles [Figure 4] Block diagram showing the reception-side processing procedures in an underwater OFDM communication system and communication method [Diagram 5] Diagram representing the received signal as a horizontally long block corresponding to time when the receiving transducer receives only one direct wave from the transmitting transducer while moving [Figure 6]A diagram showing the received signal when a receiving transducer moves and receives two waves, a direct wave and a delayed wave, from a transmitting transducer, with horizontal blocks corresponding to time. [Figure 7] A block diagram showing the procedure for generating inter-carrier interference cancellation coefficients using various parameters of each propagating wave (direct wave and delayed wave) revealed by the "Modified Delay and Doppler Profiler" block (44) used in the receiving transducer. [Figure 8] A block diagram showing the procedure for generating inter-carrier interference cancellation coefficients using various parameters of each propagating wave (direct wave and delayed wave) revealed by the conventional "Delay and Doppler Profiler" block. [Figure 9] Block diagram showing the receiving side processing of an OFDM communication system for underwater acoustic communication [Figure 10] A diagram showing multiple OFDM signals in the time direction of an OFDM communication system for underwater acoustic communication, expressed on a two-dimensional plane of time (horizontal axis) and frequency (vertical axis), with complex-modulated subcarrier symbols arranged as circles. [Figure 11] A diagram showing various parameters of each propagation wave (direct wave and delayed wave) estimated by the mDDP block used in the OFDM communication system for underwater acoustic communication. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] FIG. 1 is a schematic diagram showing an embodiment of an underwater OFDM communication system and communication method according to the present invention. A mother ship (11) equipped with a transmitting transducer (13) that serves as a transmitter for transmitting acoustic signals is located on the sea surface, and an AUV (Autonomous Unmanned Vehicle) equipped with a receiving transducer (14) that serves as a receiver for receiving the acoustic signal transmitted from the transmitting transducer (13) is located underwater.

[0012] The acoustic signal transmitted from the transmitting transducer (13), which is the transmitting section, is transmitted to the receiving transducer (14), which is the receiving section, via two paths: a straight path L1 and a path L2 that reflects off the sea surface. The acoustic signals transmitted through paths L1 and L2 arrive at the receiving transducer (14) at an angle θ different from each other. When the mother ship (11) moves with a velocity v, the acoustic signals transmitted to the receiving transducer (14) through path L1 and through path L2 have different Doppler displacements.

[0013] FIG. 2 is a block diagram showing a processing procedure for transmitting and receiving an acoustic signal in the underwater OFDM communication system and communication method according to the present invention.

[0014] The upper part of FIG. 2 shows the procedure for transmitting an acoustic signal transmitted from the transmitting transducer (13).

[0015] The acoustic signal transmitted from the transmitting transducer (13), shown as "Bit Data" in Figure 2, is digitally modulated to QPSK / 16QAM / 64QAM, etc., by the "Symbol map" block and converted into a complex symbol. In order to estimate the characteristics of the transmission path (channel) at the receiving side (direct wave and delayed wave (collectively called "propagation wave"), complex attenuation of each propagating wave, relative delay time, and Doppler frequency shift amount), a pilot signal composed of a predetermined complex value is inserted into the complex symbol in the "Pilot insert" block as necessary.

[0016] Then, the complex symbols are subjected to an inverse Fourier transform (IFFT) by an "IFFT" block for each fixed symbol size, and are then orthogonal frequency division multiplexed (OFDM) modulated to generate OFDM symbols. In addition, the “CP add” block performs a process of copying and adding a part of the rear part of the OFDM symbol to the front part (CP addition: cyclic prefix addition). The addition of this CP suppresses interference between OFDM symbols transmitted at different times.

[0017] The signal processing up to this point is so-called baseband signal processing, which corresponds to signals near 0 Hz in the frequency spectrum. In order to convert this acoustic signal to be transmitted into the desired frequency band, only the real part of the acoustic signal is extracted by up-conversion processing in the "up-cnv" block and transmitted from the transmitting section. The area indicated by a hollow triangle pointing downward (with only one apex on the lower side) on the upper right side of Figure 2 indicates this transmitter. The area indicated by a hollow triangle facing to the right (with only one apex on the right) in front of the transmitter unit indicates the transmitting power amplifier.

[0018] "Channel" in the middle of Figure 2 indicates the transmission path along which the signal transmitted from the transmitter propagates. An example of the transmission path is the path L1 and the path L2 shown in FIG.

[0019] The lower part of FIG. 2 shows the procedure for processing the transmission of an acoustic signal received by the receiving transducer (14). The area indicated by a hollow triangle pointing downward (with only one apex on the bottom left) in the lower left corner of Figure 2 indicates the receiving section. The signal received by the receiver is amplified to the required signal size by a low-noise amplifier (indicated by a hollow triangle facing to the right (with only one vertex on the right side)).

[0020] Then, the signal received at the receiver is converted to a complex baseband signal by down-conversion processing in the "down-cnv" block. Furthermore, the "CP remove" block performs processing to remove a portion of the front part of the OFDM symbol (CP removal), and the "Time-domain Pre-processing" block performs time-domain pre-processing. This time-domain pre-processing involves stretching and shrinking the complex baseband signal to compensate for the stretching and shrinking of the received signal introduced by the Doppler shift.

[0021] Then, the "FFT" block performs the Fourier transform (FFT), which is the main process of OFDM demodulation. The output of this FFT includes the complex symbols converted from the acoustic signal and the complex symbols inserted into the pilot signal during processing on the transmitting side. Of these, by using the complex symbol inserted in the pilot signal, the "Modified Delay and Doppler profiler" block can estimate parameters related to the acoustic signal propagating through path L1 (hereinafter referred to as the "direct wave") and the acoustic signal propagating through path L2 (hereinafter referred to as the "delayed wave") shown in Figure 1.

[0022] Using these parameters for the direct wave and delayed wave, it is possible to analyze the transmission characteristics, including the interference between the complex symbols converted from the OFDM-demodulated acoustic signal and the complex symbols inserted into the pilot signal, which are simultaneously transmitted by OFDM, and it is also possible to calculate the inter-carrier interference cancellation coefficient that removes this interference. The interference cancellation coefficient is calculated in the “ICI Cancel Coef. generation” block (Inter Carrier Interference cancellation coefficient generation) in the lower part of Figure 2.

[0023] Interference cancellation using this inter-carrier interference cancellation coefficient is performed in the "Multi Tap Equalizer" block, and the complex symbols converted from the acoustic signal are digitally modulated to QPSK / 16QAM / 64QAM, etc. Then, the "De-map" block converts it into received data (bit data), and digital bit data is transmitted from the transmitting side to the receiving side.

[0024] FIG. 3 shows multiple OFDM signals in the time direction, expressed on a two-dimensional plane with time on the horizontal axis and frequency on the vertical axis, with complex-modulated subcarrier symbols arranged as circles. (31) is an OFDM signal "OFDM symbol" (also referred to as "OFDM symbol" in this specification) represented as a horizontal block corresponding to time, and (32) indicates the guard interval (GI) portion which is a copy of the latter part of the OFDM signal and is attached to the beginning of the OFDM signal, and means the cyclic prefix (CP). One OFDM symbol contains a complex symbol (34) that is the digitally modulated acoustic signal shown in “Bit Data” in Figure 2, and complex symbols (35) and (36) inserted into the pilot signal.

[0025] The pilots indicated by black circles (35) are called Scattered Pilots (SP), and in FIG. 3, they are arranged horizontally (in the time direction) at odd numbers in the frequency direction index. The pilots indicated by light black circles (36) are called continuous pilots (CPs), and in FIG. 3, they are arranged horizontally (in the time direction) at even numbers in the frequency direction index. In this way, the complex symbol (34) digitally modulated from the acoustic signal shown in “Bit Data” in FIG. 2 and the complex symbols (35) and (36) inserted into the pilot signal can be represented on a two-dimensional plane as shown in FIG. 3.

[0026] FIG. 4 is a block diagram showing a processing procedure in which the receiving transducer 14 receives an acoustic signal in the underwater OFDM communication system and communication method according to the present invention. The acoustic signal transmitted to the receiving transducer (14) is converted into a weak electrical signal and the signal is amplified by a low-noise amplifier in the “PreAMP” block at the left end of the upper part of Figure 4. The "ADC" block then converts the analog signal into a digital signal (digital number sequence). "Fs" indicates the sampling frequency of the underwater OFDM communication system and communication method according to the present invention.

[0027] Then, the acoustic signal (received signal) converted into a digital signal is converted into a complex baseband signal by down-conversion processing in the "Down-cnv" block. (41) is the “Resample and De-rotate” block, which performs stretching and shrinking of the received signal. "Resample and De-rotate" is performed for multiple OFDM symbols, and in this embodiment, is performed every two OFDM symbols.

[0028] "β1" in the figure means the expansion / contraction ratio detected in the process immediately before the current process, and "β2" in the figure means the expansion / contraction ratio detected in the current process. Therefore, β2 is a more accurate expansion / contraction ratio than β1. In "Resample and De-rotate", the received signal is stretched or shrunk using β2, which is a more accurate stretch / shrunk ratio than β1.

[0029] In terms of signal processing, this expansion and contraction process is called resampling and derotation, and derotation processing means frequency correction. However, since the signal is represented as a complex number and vibration is mathematically represented as rotation, the term derotation is used to refer to the inverse rotation of rotation.

[0030] Then, the phase shift is adjusted in the "Phase Shift Compensation" block (42) for the output of the "Resample and De-rotate" block (41). The processing in blocks (41) and (42) corresponds to the time-domain pre-processing in the “Time-domain Pre-processing” block in FIG.

[0031] The "3rd FFT" (43) is a block that performs the Fourier transform (FFT), which is the main process of OFDM demodulation. The digital numeric sequence converted from the analog signal by the “ADC” block undergoes time-domain pre-processing in blocks (41) and (42) (the “Time-domain Pre-processing” block), and is then discrete Fourier transformed by the “3rd FFT” (43). The frequency domain signal after the discrete Fourier transform is then analyzed.

[0032] In the analysis, the scattered pilots, which are complex symbols included in the output signal (frequency domain signal) of the discrete Fourier transform, are used to measure the channel transfer function (CTF) of the scattered pilots for each transmission path. Using these measurements, the “Modified Delay and Doppler Profiler” block (44) estimates various parameters of each propagating wave (direct wave and delayed wave) for each transmission path that constitutes the acoustic signal (received wave) transmitted to the receiving transducer (14). The various parameters will be described in detail later.

[0033] By using detailed parameters of each estimated propagation wave (direct wave and delayed wave), it is possible to estimate the transmission path transfer function h(k, l, n) from the nth subcarrier of the transmitted signal to the lth subcarrier of the received signal at symbol number k, including the interference between the complex symbol converted from the acoustic signal simultaneously transmitted by OFDM and the complex symbol inserted in the pilot signal, based on Equation (7), Equation (8a), and Equation (8b) described below, and it also becomes possible to calculate the inter-carrier interference removal coefficient that removes this interference. The inter-carrier interference cancellation coefficient is calculated in the “ICI Cancel Coef. Generation” block (45) in the lower part of Figure 4. The inter-carrier interference cancellation coefficient can be calculated by equation (6) described later.

[0034] The process of removing inter-carrier interference using this inter-carrier interference removal coefficient is executed in a "Multi Tap Equalizer" block (46), and the complex symbol from which interference has been removed is output from "EQ_OUT2." The "Multi Tap Equalizer" block (46) is a processing means in which multiple "complex multipliers" represented by "x" operate in parallel, as shown in FIG. Specifically, the output signal Z -1 For each of the numerical subsequences shifted by the “shift register” indicated by “”, the inter-carrier interference is cancelled by multiplying them by multiple “complex multipliers” based on the inter-carrier interference cancellation coefficients generated by the “ICI Cancel Coef. Generation” block (45). For example, "Z -1 Assuming that the width of the numerical subsequence shifted by the "shift register" indicated by " " is an odd integer called "NumCOL", and the subcarrier in the center of the "complex multiplier" is the l-th subcarrier, the two sides of it correspond to subcarriers from "l-NumCOL / 2" to "l+NumCOL / 2". Therefore, inter-carrier interference is eliminated by performing a process to subtract (remove) the interference component with adjacent subcarriers for the l-th subcarrier.

[0035] Figure 5 shows the received OFDM signal (OFDM symbol) as a horizontal block corresponding to time ("Main Path #1" in the figure means "direct wave") when the receiving transducer (14) moves and receives only one direct wave (an acoustic signal transmitted through path L1 shown in Figure 1) from the transmitting transducer (13). In the figure, each of the OFDM signals "k-2" and "k" indicates a signal in which scattered pilots are arranged.

[0036] The top row shows three OFDM signals when the receiving transducer (14) is not moving through the water between the time the acoustic signal is transmitted from the transmitting transducer (13) and the time it receives the acoustic signal. The middle and bottom rows show three OFDM signals when the receiving transducer (14) moves through the water between the time the acoustic signal is transmitted from the transmitting transducer (13) and the time it receives the acoustic signal. In other words, the middle and bottom rows show a situation in which the received signal (OFDM signal) is stretched, i.e., the propagation distance of the acoustic signal between the receiving transducer (14) and the transmitting transducer (13) is expanded as the receiving transducer (14) moves through the water between the time the acoustic signal is transmitted from the transmitting transducer (13) and the time it receives the acoustic signal.

[0037] The bottom row shows a case where the following expansion / contraction correction means is applied to the OFDM signal in the middle row. 1. Resample and derotate with low precision stretch ratio β1 2. High-precision detection of the stretch ratio β2 using the resampled output signal 3. Resample and derotate using high-precision stretch ratio β2

[0038] When a means for correcting expansion and contraction is applied, as shown in the lower part, the receiving transducer (14) moves through the water between the time the acoustic signal is transmitted from the transmitting transducer (13) and the time the receiving transducer receives the acoustic signal, and the expansion and contraction of the received signal introduced by the Doppler displacement is compensated for by the correction, resulting in a received waveform similar to that shown in the upper part, which shows the case when there is no movement.

[0039] Figure 6 shows the received OFDM signal (OFDM symbol) represented by horizontal blocks corresponding to time when the receiving transducer (14) moves and receives from the transmitting transducer (13) two superimposed waves: a direct wave (acoustic signal transmitting through path L1 shown in Figure 1) and a delayed wave (acoustic signal transmitting through path L2 shown in Figure 1) arriving from a direction shifted by an angle θ from the direct wave. In other words, the horizontally long block in Figure 6 is the "Main Path #1" in Figure 5 (hereinafter referred to as the "direct wave" in the explanation of Figure 6) to which "Delayed Path #2" ("Delayed Path #2" means "delayed wave", and hereinafter referred to as the "delayed wave" in the explanation of Figure 6) has been added.

[0040] The top row shows three OFDM signals when the receiving transducer (14) is not moving through the water between the time the acoustic signal is transmitted from the transmitting transducer (13) and the time it receives the acoustic signal. The middle and bottom rows show three OFDM signals when the receiving transducer (14) moves through the water between the time the acoustic signal is transmitted from the transmitting transducer (13) and the time it receives the acoustic signal. In other words, the middle and bottom rows show a situation in which the received signal (OFDM signal) is stretched, i.e., the propagation distance of the acoustic signal between the receiving transducer (14) and the transmitting transducer (13) is expanded as the receiving transducer (14) moves through the water between the time the acoustic signal is transmitted from the transmitting transducer (13) and the time it receives the acoustic signal.

[0041] Since the direct wave and the delayed wave are received as a mixed signal, the stretching and contraction correction for the delayed wave is performed in the same way as for the direct wave. In addition, when the difference in the arrival angle between the direct wave and the delayed wave is 0 degrees, i.e., there is no difference at all and they arrive from the same direction, the expansion and contraction of the signal introduced by the Doppler shift relative to the delayed wave is corrected by the same resampling and derotation processing as for the direct wave.

[0042] However, the direction in which a delayed wave arrives usually does not coincide with the direction in which a direct wave arrives, and the expansion / contraction ratio is different. Therefore, assuming that the difference in the arrival angle between the direct wave and the delayed wave is θ degrees, the speed of the direct wave is v, and the speed of the delayed wave is vcosθ, if a means of correcting the expansion and contraction is applied to the delayed wave in the same way as for the direct wave, it will become too contracted, as shown in the "Delayed Path#2" block at the bottom. In addition, “v·(1-cosθ)” in the figure indicates the expansion / contraction ratio.

[0043] In other words, in OFDM signal processing in a multipath environment where the OFDM signal propagates through multiple different paths and arrives at the receiver from different directions (angles), even if the above-mentioned resampling and derotation processing is applied, the expansion and contraction of the signal introduced by the Doppler shift for the delayed wave is not completely corrected, causing a disturbance in the orthogonality between carriers, resulting in inter-carrier interference during demodulation processing of the OFDM signal.

[0044] FIG. 7 is a diagram showing a procedure for generating an inter-carrier interference cancellation coefficient for canceling interference between a complex symbol converted from an acoustic signal simultaneously transmitted by OFDM and a complex symbol inserted into a pilot signal, using various parameters of each propagating wave (direct wave and delayed wave) constituting the acoustic signal (received wave) transmitted to the receiving transducer (14) as revealed by the “Modified Delay and Doppler Profiler” block (44) used in the “Processing procedure for receiving an acoustic signal by the receiving transducer (14) in the underwater OFDM communication system and communication method” in FIG. 4.

[0045] FIG. 8 shows a case where the "Modified Delay and Doppler Profiler" block (44) in FIG. 7 is used in a conventional resampling and derotation process ("Delay and Doppler Profiler" block (44)) utilized in a reception system and communication method using radio waves or the like.

[0046] In the conventional method of FIG. 8, using the scattered pilot of the OFDM symbol, for each delayed wave i constituting the multipath, the complex attenuation coefficient r i , the relative delay time τ i , and the Doppler frequency shift amount α i of each parameter were obtained. However, in radio wave communication, when comparing the normally used frequency bandwidth B (Hz) with the center frequency fc (Hz) of the signal being used, B << fc, that is, the condition that fc is sufficiently larger than B holds. Therefore, for the complex attenuation coefficient r i , the relative delay time τ i , and the Doppler frequency shift amount α i , although approximate values can be calculated by analyzing the delayed wave using a total of three parameters, in underwater acoustic communication, usually the condition that fc is sufficiently larger than B does not hold. Therefore, on the receiving side of the signal, a process of stretching and shrinking the received signal is required.

[0047] However, when not only the direct wave but also a delayed wave exists and the delayed wave arrives from a direction shifted by an angle θ from the direct wave, when performing the stretching and shrinking process of the received signal consisting of the following three processes shown within the square frame in the center of FIG. 6, as shown in "Delayed Path#2" (delayed wave) at the bottom of FIG. 6, the correction is not made and it appears as if the stretching and shrinking ratio has changed. 1. Rough Resampled and De-rotate by β1 Resampling and derotation with a low-precision stretching and shrinking ratio β1 2. Detect β2 = β1*(1+(de-ds) / (GI+FFTsize) / 4) Detecting a highly accurate stretching / compressing ratio β2 using the resampled output signal 3. Fine Resampled and De-rotate by β2 Resampling and derotation using a highly accurate stretching / compressing ratio β2

[0048] That is, there is a problem that inter-carrier interference occurs during the demodulation process of the OFDM signal for the delayed wave. Correcting the sampling clock error corresponds to stretching / compressing the received signal as a result. However, stretching / compressing the received signal is a process in the time domain and is the process of the "Resampled and De-rotate" block (41) that performs stretching / compressing processing on the received signal before the "3rd FFT" block (43) in FIG. 4. Therefore, after the "3rd FFT" block (43), the received signal becomes a signal in the frequency domain, and the received signal cannot be stretched / compressed.

[0049] In the present invention, it is decided to cancel (remove) the inter-carrier interference for the effects that could not be compensated by the stretching / compressing process in the time domain such as the "Delayed Path#2" (delayed wave). In order to remove the inter-carrier interference, when B << fc does not hold, as shown in FIG. 7, for the delayed waves constituting the multipath, the sampling clock error β p needs to be added. Therefore, as shown in FIG. 7, for each delayed wave p constituting the multipath, the complex attenuation coefficient r p , the relative delay time τ p , the Doppler frequency shift amount α p , and in addition, the sampling clock error β p are obtained to obtain the inter-carrier interference coefficient, and the inter-carrier interference is removed by performing the inverse process.

[0050] Note that the sampling clock error is represented by the same symbol "β" as the stretching / compressing ratio because the correction of the sampling clock error corresponds to stretching / compressing the received signal. The received signal itself is expanded or contracted by processing in the time domain, i.e., processing before the “3rdFFT” block (43) in FIG. 4, and the signal after processing by the “3rdFFT” block is a signal in the frequency domain. In the following formula, "β( p )" is expressed as a sampling clock error rather than using the term "stretch / contract ratio" in order to accommodate signal processing calculations in the frequency domain.

[0051] The sampling clock error β used in the calculation of the “ICI Cancel Coef. generation” block (Inter Carrier Interference cancellation coefficient generation) in the lower part of Figure 2 and Figure 4 (45) p can be derived by the following formula:

[0052]

number

[0053] Equation (1) is the baseband signal of the transmission signal. As shown in equation (2), the symbol length Ts, which is the sum of the OFDM symbol length and the guard interval length, is given by equation (2).

[0054]

number

[0055]

number

[0056]

number

[0057] Equation (4) shows the baseband signal received via Np transmission paths, and the complex attenuation coefficient r for the delayed wave p that constitutes the multipath isp , relative delay time τ p , Doppler frequency shift Δf p This can be shown using:

[0058]

number

[0059] Equation (5) is a transformation equation for discretizing the sampling time of the received signal of equation (4). This converts the analog signal from the PreAMP into a discrete signal sequence using the ADC (analog-to-digital converter) in Figure 4. The sampling clock error of this ADC sampling frequency Fs is denoted as γ. The sampling clock error γ is a small value of about 1 / 10,000. This sampling clock error γ is a sum of the sampling clock error β p The two differ in that, after being processed by the “Resample De-rotate” block (41), the pth wave of multiple propagating waves propagating in the channel is detected by the “Modified Delay and Doppler Profiler” block (44), after being discrete Fourier transformed into a frequency domain signal by the “3rd FFT” (43). Moreover, the sampling clock error γ is different from β1 and β2 of the “Resample De-rotate” block (41).

[0060] By applying the discrete Fourier transform to the discretized received signal, equation (6) is obtained.

[0061]

number

[0062] d(k,n) is the transmission symbol, and as shown in Figure 3, multiple OFDM signals in the time direction are represented on a two-dimensional plane with time on the horizontal axis and frequency on the vertical axis, and are complex values ​​equivalent to the complex-modulated subcarrier symbols arranged with circles. k is a time direction index and corresponds to the column number in FIG. n is the frequency index and corresponds to the row number in FIG. l is an index in the frequency direction. After receiving the data, 3rdFFT is executed, and the output corresponds to the row number when arranged as in Figure 3. Therefore, the left side of equation (6) corresponds to the output of the 3rd FFT and corresponds to the received symbol on the receiving side.

[0063] The first term on the right side of equation (6) indicates that transmission from the transmission symbol d(k,l) at the frequency index is performed with the transfer coefficient h(k,l,l), which is the principal component when there is no inter-carrier interference. The two terms on the right side of equation (6) are components from transmission carriers different from the reception carrier index, and are components that indicate inter-carrier interference.

[0064] In particular, h(k,l,n) is a coefficient indicating this contribution, which denotes the interference coefficient from transmitting carrier index n to receiving carrier index l. This coefficient has a large value when l=n and decreases as the difference between l and n becomes larger. The three terms w(k,l) on the right side of equation (6) are noise components. Therefore, if we consider the range of n from l-1 to l+1 and ignore the noise w(k,l), we can express it as follows: By multiplying the inverse matrix of JPEG2025018926000008.jpg181343x3 matrix on the left, it can be transformed as follows. JPEG2025018926000009.jpg37137And if you take out a part of this, JPEG2025018926000010.jpg1859, and these three complex coefficients d, e, and f are the inter-carrier interference cancellation coefficients. Furthermore, the accuracy can be improved by expanding this 3X3 matrix calculation.

[0065]

number

[0066] The transmission path transfer function h of the delayed wave for each delayed wave index p shown in equations (8a) and (8b) is p By summing up (k, l, n), we obtain h(k, l, n) in equation (7).

[0067]

number

[0068]

number

[0069] As shown in FIG. 7, the channel transfer function h(k, l, l) of the arrangement portion of the scattered pilots embedded at indexes k and k−2 can be measured on the receiving side. Therefore, consider the cost function of equation (9). x(k,l) and x(k-2,l) are symbol values ​​obtained by performing a discrete Fourier transform on the received signal, and d(k,l) and d(k-2,l) are transmitted symbol values. However, the sum (sigma) shown in equation (9) is calculated only for the scattered pilot arrangement portion. All transmission symbol values ​​are predetermined values.

[0070] Where: JPEG2025018926000014.jpg1518 is the transmission channel transfer function h(k,l,l) at symbol number k, which is superimposed with interference noise from other carriers measured by the scattered pilot. JPEG2025018926000015.jpg1429 is a channel transfer function h(k-2, l, l) for symbol number k-2 on which interference noise from other carriers is superimposed and measured by scattered pilots.

[0071] Here, the transmission channel transfer function h(k,l,n) indicates the transmission channel transfer function from the nth subcarrier to the lth subcarrier of the transmitted signal at symbol number k, and the transmission channel transfer function h(k,l,l) measured by the scattered pilot indicates the transmission channel transfer function of the lth subcarrier where the transmitted signal and received signal are the same.

[0072]

number

[0073] By minimizing this cost function, the parameters of each delayed wave shown in FIG. 7 can be obtained. The procedure for determining the parameters is as follows: for each delayed wave, four corresponding parameters are determined from the elements of delayed waves with the greatest power contained in the received signal.

[0074]

number

[0075]

number

[0076] Equations (10) and (11) are obtained by modeling the first wave using the four parameters r1, τ'1, α1, and β1 of the first wave and substituting them into equation (9). These four parameters are obtained by minimizing equation (11). Here, τ'1 has a changed unit and indicates the delay time in units of the number of sampling points, and α1 is converted to a dimensionless value by dividing the Doppler frequency shift by the OFDM subcarrier spacing frequency f0. Various parameters are found by minimizing E1(k), so by complex differentiating E1(k) with respect to r1 and setting it to 0, equations (12a) and (12b) can be obtained.

[0077]

number

[0078]

number

[0079] Moreover, by substituting r1 into equation (11), we obtain equation (13a).

[0080]

number

[0081] The denominator of the fraction in equation (13a) is not the part that changes value very much. In other words, by maximizing equation (13b), E1(k) can be minimized.

[0082]

number

[0083]

number

[0084] By differentiating F(α1) in equation (13c) with respect to α1 and setting it to 0, we obtain equation (14).

[0085]

number

[0086] To find τ'1, it is possible to substitute various values, roughly find a value close to the true value, and then find the value using the Newton method of convergence. The added β1 parameter can also be roughly calculated as τ'1.

[0087] Incidentally, the means for estimating the channel characteristics of the transmission path in the present invention is a “Modified Delay and Doppler Profiler” (hereinafter also referred to as “mDDP”) block (44) that receives the transmission path coefficients of pilot symbols as input. However, in mobile communications where the receiving side moves, the orthogonality between the symbols of each subcarrier constituting the OFDM is disturbed due to the effect of Doppler shift. Specifically, the value of a certain subcarrier number l interferes with other subcarriers with nearby subcarrier numbers. That is, in mobile reception, there is a problem in that the subcarrier channel coefficient h(k,l,l) has interference components multiplexed due to interference between carriers, reducing the accuracy of the value.

[0088] In addition, since the transmission line coefficients h(k,l,l) with reduced accuracy are the input to the mDDP, the accuracy of the parameters used to estimate the channel characteristics of the transmission line is degraded, and the accuracy of removing distortion and interference generated in the transmission line from the received signal after propagation through the transmission line, which is performed based on these parameters, is also degraded, resulting in an increase in the error rate of the received data on the receiving side.

[0089] In order to solve these problems, it is necessary to obtain accurate channel characteristics of the transmission path. As a means of obtaining the channel characteristics of the transmission path, by repeatedly applying mDDP as shown in the following embodiment, the estimation accuracy of the channel parameters indicating the channel characteristics of the transmission path can be improved, the bit error rate during the demodulation process of the OFDM signal on the receiving side can be reduced, and higher performance wireless communication can be realized.

[0090] FIG. 9 is a block diagram showing an embodiment of a receiving side process in underwater acoustic communication. The OFDM signal in underwater acoustic communication propagated through water is received by a transducer. The signal is then amplified in the "PreAMP" block in Figure 9, converted from an analog signal to a digital signal (digital number sequence) in the "ADC" block, and converted into a discrete signal sequence. "Fs" stands for the sampling frequency.

[0091] Then, the signal received by the receiving section is converted into a complex baseband signal by down-conversion processing in the "Down-Conv." block. As the main demodulation of the OFDM complex modulation, a discrete Fourier transform is performed in an FFT block (111).

[0092] The processing by this FFT block (111) is a high-speed processing algorithm for discrete Fourier transform, which breaks down the input time domain signal into the frequency signals that it is made up of. Therefore, the signal after this FFT processing is called the "frequency domain signal" and is distinguished from the "time domain signal" before the processing by the FFT block (111).

[0093] This “frequency domain signal” will be explained using Figure 10, which shows multiple OFDM signals in the time direction of an OFDM communication system in underwater acoustic communication, represented on a two-dimensional plane of time (horizontal axis) and frequency (vertical axis), with complex-modulated subcarrier symbols arranged as circles.

[0094] Each circle is a complex symbol, with the unfilled circles indicating data carrying symbols and the shaded circles indicating discrete pilot signals, each of predetermined values. The filled circles indicate continuous pilots, which are pilot signals inserted at specific carrier numbers on the frequency axis and are continuous in the time direction, and therefore can be used to measure changes in the time direction.

[0095] The “CTF at SP” block (112) in FIG. 9, which is a CTF generation block that determines the channel transfer function (CTF), determines the channel coefficient h(k,l,l) from the value of “a certain location” (time index k, frequency index l) of the discrete pilot signal. Here, the symbols on the transmitting side of each circle in Fig. 10 are d(k,l), and the symbols on the receiving side are If we use JPEG2025018926000025.jpg818, the relationship in equation (15) can be shown.

[0096]

number

[0097] The first term on the right side of equation (15) indicates that the frequency index l (el) on the transmitting side and receiving side is the same, so transmission is on subcarriers of the same frequency, which represents the main transmission component. The two terms on the right side of equation (15) represent the sum when the frequency index n on the transmitting side does not match the frequency index l on the receiving side, and represent interference from different carriers. The three terms on the right side of equation (15) correspond to noise. Therefore, h(k,l,n) denotes a coefficient indicating the inter-carrier interference.

[0098] The details of the first "1st Coarse Modified Delay and Doppler Profiler" block (113) and the second "2nd Fine Modified Delay and Doppler Profiler" block (116) in FIG. 9 will be described with reference to FIG. 11 which shows various parameters of each propagation wave (direct wave and delayed wave) estimated by the mDDP block.

[0099] In this embodiment, the transmission line coefficient h(k,l,l) at a certain position of the SP and the time direction index are inputted with the value of the transmission line coefficient h(k-2,l,l) two positions before, and the coefficient of the transmission line channel is detected according to the table in FIG. The frequency index l indicates the portion where the SP is located.

[0100] Details of the "Modified Delay and Doppler Profiler" are as described above in

[0032] ,

[0044] to

[0050] .

[0101] The first input of the “Modified Delay and Doppler Profiler” (113) can be expressed as in equations (16a) and (16b) by ignoring the second and third terms on the right-hand side of equation (15).

[0102]

number

[0103]

number

[0104] If the effect of the Doppler shift is large and the two terms in equation (15) cannot be ignored, there is a problem in that the accuracy of the transmission channel coefficient h(k,l,l) and the transmission channel coefficient two channels before it, h(k-2,l,l), decreases. In that case, improvement in accuracy can be expected by ignoring only the third term in equation (15) instead of ignoring the second term in equation (15). In this way, if only the three terms in equation (15) are ignored, it can be expressed as equations (17a) and (17b).

[0105]

number

[0106]

number

[0107] To calculate equation (17b), it is necessary to find h(k,l,n) and h(k-2,l,n). To do this, if the complex attenuation coefficient, relative delay time, Doppler frequency shift amount, and sampling clock error for each propagation path Path number p ("Path No. p" in FIG. 11) constituting the transmission line shown in the table of FIG. 11 are known, the calculations can be made as shown in

[0065] to

[0068] above.

[0108] The right-hand sides of equations (17a) and (17b) subtract the sum of the interference components, and this calculation is performed by the “ICI Remove” block (114) in FIG. Then, the left side of equations (17a) and (17b) is calculated in the “Denoised CTF at SP” block (115), which calculates the noise-reduced CTF, and a similar calculation is performed a second time in the “2nd Fine Modified Delay and Doppler Profiler” block (116). In other words, we can expect an improvement in input accuracy, and an improvement in the accuracy of the estimated parameters in the table on the right of Figure 11, which is the output. The subsequent processing is the same as that described above in

[0032] to

[0034] , and an estimate of the transmission symbol d(k,l) with reduced inter-carrier interference is calculated in the multi-tap equalizer (118).

[0109] As described above, by applying the "Modified Delay and Doppler Profiler" at least twice, the estimation accuracy of the transmitted symbol value can be improved.

Claims

1. The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication system that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A means for converting the superimposed received signal from an analog signal into a digital number sequence, which is a digital signal, at a predetermined sampling frequency; means for performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; means for analyzing said frequency domain signal; It consists of: The means for analyzing the frequency domain signal includes: A measurement function for measuring a transmission path transfer function of a portion where the scattered pilots are arranged for each transmission path; An estimation function for estimating four parameters, i.e., a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave in the transmission path; a calculation function for estimating a transmission path transfer function including interference between complex symbols of a frequency domain signal for each of the transmission paths using the parameters, and calculating an inter-carrier interference cancellation coefficient; A cancellation function for canceling inter-carrier interference based on an inter-carrier interference cancellation coefficient; having An underwater OFDM communication system comprising:

2. The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication system that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A means for converting the superimposed received signal from an analog signal into a digital number sequence, which is a digital signal, at a predetermined sampling frequency; means for performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; means for analyzing said frequency domain signal; It consists of: The means for analyzing the frequency domain signal includes: A measurement function for measuring a transmission path transfer function of a portion where the scattered pilots are arranged by using the scattered pilots of the complex symbols inserted into the pilot signal among the complex symbols included in the frequency domain signal for each transmission path; An estimation function for estimating four parameters, i.e., a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave in the transmission path; a calculation function for estimating a transmission path transfer function including interference between a complex symbol included in a frequency domain signal and a complex symbol inserted in a pilot signal for each transmission path using the parameters, and calculating an inter-carrier interference cancellation coefficient; A cancellation function for canceling inter-carrier interference based on an inter-carrier interference cancellation coefficient; having An underwater OFDM communication system comprising:

3. The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication system that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A means for converting the superimposed received signal from an analog signal into a digital number sequence, which is a digital signal, at a predetermined sampling frequency; means for performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; means for analyzing said frequency domain signal; It consists of: The means for analyzing the frequency domain signal includes: A measurement function for measuring a transmission path transfer function of a portion where the scattered pilots are arranged by using the scattered pilots of the complex symbols inserted into the pilot signal among the complex symbols included in the frequency domain signal for each transmission path; An estimation function for estimating four parameters, i.e., a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave in the transmission path; a calculation function for estimating a transmission path transfer function including interference between a complex symbol included in a frequency domain signal and a complex symbol inserted in a pilot signal for each transmission path using the parameters, and calculating an inter-carrier interference cancellation coefficient; a cancellation function for canceling inter-carrier interference by canceling interference components from the output signal that has been subjected to the discrete Fourier transform based on the inter-carrier interference cancellation coefficient; having An underwater OFDM communication system comprising:

4. The means for analyzing the frequency domain signal includes: said computing functionality; The removal function; In between, The parameters estimated by the estimation function are generating a scattered pilot by removing an interference component from the scattered pilot; a second estimation function for estimating four parameters, including a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave of the transmission path; having 4. The underwater OFDM communication system according to claim 1, wherein the underwater OFDM communication system is a cascade of amplified signals.

5. The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication method that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A step of converting the superimposed received signal from an analog signal to a digital number sequence, which is a digital signal, at a predetermined sampling frequency; performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; analysing said frequency domain signal; It consists of: The step of analyzing the frequency domain signal includes: A measurement function for measuring a transmission path transfer function of a portion where the scattered pilots are arranged for each transmission path; An estimation function for estimating four parameters, i.e., a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave in the transmission path; a calculation function for estimating a transmission path transfer function including interference between complex symbols of a frequency domain signal for each of the transmission paths using the parameters, and calculating an inter-carrier interference cancellation coefficient; A cancellation function for canceling inter-carrier interference based on an inter-carrier interference cancellation coefficient; Execute 2. An underwater OFDM communication method comprising:

6. The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication method that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A step of converting the superimposed received signal from an analog signal to a digital number sequence, which is a digital signal, at a predetermined sampling frequency; performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; analysing said frequency domain signal; It consists of: The step of analyzing the frequency domain signal includes: A measurement function for measuring a transmission path transfer function of a portion where the scattered pilots are arranged by using the scattered pilots of the complex symbols inserted into the pilot signal among the complex symbols included in the frequency domain signal for each transmission path; An estimation function for estimating four parameters, i.e., a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave in the transmission path; a measurement function for estimating a transmission path transfer function including interference between a complex symbol included in a frequency domain signal and a complex symbol inserted in a pilot signal for each transmission path using the parameters, and calculating an inter-carrier interference cancellation coefficient; A cancellation function for canceling inter-carrier interference based on an inter-carrier interference cancellation coefficient; Execute 2. An underwater OFDM communication method comprising:

7. The OFDM signal transmitted from the transmitting unit that transmits the acoustic signal is An underwater OFDM communication method that performs OFDM signal processing in a multipath environment in which signals are propagated through two or more different transmission paths underwater and received by a receiving unit in a superimposed state from the different transmission paths, A step of converting the superimposed received signal from an analog signal to a digital number sequence, which is a digital signal, at a predetermined sampling frequency; performing a discrete Fourier transform on the sequence of digital numbers to convert the signal into a frequency domain signal; analysing said frequency domain signal; It consists of: The step of analyzing the frequency domain signal includes: A measurement function for measuring a transmission path transfer function of a portion where the scattered pilots are arranged by using the scattered pilots of the complex symbols inserted into the pilot signal among the complex symbols included in the frequency domain signal for each transmission path; An estimation function for estimating four parameters, i.e., a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave in the transmission path; a calculation function for estimating a transmission path transfer function including interference between a complex symbol included in a frequency domain signal and a complex symbol inserted in a pilot signal for each transmission path using the parameters, and calculating an inter-carrier interference cancellation coefficient; a cancellation function for canceling inter-carrier interference by canceling interference components from the output signal that has been subjected to the discrete Fourier transform based on the inter-carrier interference cancellation coefficient; Execute 2. An underwater OFDM communication method comprising:

8. The step of analyzing the frequency domain signal includes: said computing functionality; The removal function; In between, The parameters estimated by the estimation function are generating a scattered pilot by removing an interference component from the scattered pilot; a second estimation function for estimating four parameters, including a complex attenuation coefficient, a relative delay time, a Doppler frequency shift amount, and a sampling clock error, for each propagation wave of the transmission path; Execute 8. An underwater OFDM communication method according to claim 5, wherein the underwater OFDM communication method is a digital underwater communication method.

Images