OFDM underwater acoustic communication method based on pseudo-random noise sequence

Abstract

The application provides an OFDM underwater acoustic communication method based on a pseudo-random noise sequence, compared with existing TDS-OFDM and DPN-OFDM, due to the existence of CP in the PN-CPOFDM frame structure design, the influence of inter-symbol interference (ICI) on OFDM symbols is greatly weakened, therefore, the bit error rate performance of the PN-CPOFDM is superior to that of TDS / DPN-OFDM; the application simulates an actual underwater acoustic channel by using BELLHOP underwater acoustic channel simulation software, under the condition of a lower signal-to-noise ratio, the bit error rate performance of DPN-OFDM is equivalent to that of TDS-OFDM, when the signal-to-noise ratio is larger, the bit error rate performance of DPN-OFDM is superior to that of TDS-OFDM, and the bit error rate performance of the modulation scheme is improved with the increase of the signal-to-noise ratio, and the bit error rate performance of the modulation scheme is always superior to that of the former two modulation technologies, which further verifies the performance of the technical scheme.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of underwater acoustic communication, and particularly relates to an OFDM underwater acoustic communication method based on pseudo-random noise sequences. BACKGROUND

[0002] Due to the double selectivity of the propagation environment, underwater acoustic channels bring great technical challenges to reliable communication. Severe multipath delay causes inter-symbol interference, and large Doppler spread causes time-varying channel impulse response. When considering how to mitigate the impact of inter-symbol interference, the time-domain processing is of high computational complexity to combat multipath, especially when the multipath delay spread is long and involves matrix inversion. In contrast, the Orthogonal Frequency-Division Multiplexing (OFDM) modulation technology, as a multi-carrier method, has an advantage over time-domain methods in reducing computational overhead because it performs frequency-domain calculations on all parallel sub-channels.

[0003] At present, a widely used modulation technology in underwater acoustic communication is single-carrier modulation technology. Single-carrier modulation technology first groups the to-be-transmitted sequence at the transmitting end, and maps each group to a transmitting symbol in a certain constellation structure. The number of symbols carried is different when different modulation orders are used. When the modulation order increases, a single transmitting symbol will carry more information, and the communication system will obtain a higher data rate. However, as the modulation order increases, the distance between the symbol points in the constellation diagram becomes smaller, and the reliability of the communication system also decreases.

[0004] Another widely used modulation technology in underwater acoustic communication is multi-carrier modulation technology. Compared with single-carrier modulation technology, multi-carrier modulation technology can convert high-speed serial data into low-speed parallel data, thereby realizing high-rate transmission. At present, the multi-carrier modulation technology widely used in underwater acoustic channels is OFDM modulation technology. However, the traditional Cyclic Prefix (CP) or Zero Padding (ZP) OFDM modulation technology has low data efficiency because the block transmission of OFDM frames requires a large guard interval, and pilot subcarriers also need to be set in the frame structure for estimating the underwater acoustic channel. In addition, when the channel state changes rapidly due to large Doppler spread, empty subcarriers also need to be set to reduce inter-carrier interference, resulting in even lower spectral efficiency. Therefore, in order to improve the spectral efficiency, some researchers have proposed a Time Domain Synchronous OFDM (TDS-OFDM) modulation technology for underwater acoustic communication. The TDS-OFDM modulation and demodulation block diagram is shown in Figure 1 , and the corresponding frame structure is shown in Figure 2The TDS-OFDM frame structure is shown in Fig. 1. It shares a similar frame structure with the Chinese national standard of Digital Terrestrial Television Broadcasting (DTTB). A Pseudo Noise (PN) sequence is used to replace the CP and ZP guard interval before each OFDM block. The PN sequence can be used as a guard interval and also for channel estimation, synchronization and Doppler estimation compensation. No additional pilot subcarriers and null subcarriers are needed in the active subcarriers, thus improving the spectral efficiency. Based on the TDS-OFDM, a Dual Pseudo Noise TDS-OFDM (DPN-TDS-OFDM) modulation technique is also applied to underwater acoustic communications, and the corresponding frame structure is shown in Fig. 2. That is, the same two PN sequences are used to fill the guard interval before each OFDM block. The first block is used to combat inter-block interference, and the other block is used for channel estimation and signal reconstruction. Figure 3

[0005] Although the TDS-OFDM improves the spectral efficiency, its bit error rate performance is low due to the severe inter-symbol interference in the underwater acoustic channel. The DPN-TDS-OFDM improves the bit error rate performance, but there is a certain loss of spectral resources due to the use of two identical PN sequences to fill the guard interval, as shown in Table 1.

[0006] In summary, the existing conventional CP / ZP-OFDM receiver processing method has a low data rate and requires pilot subcarriers for channel estimation and null subcarriers for Doppler compensation, resulting in low spectral efficiency. The TDS-OFDM and DPN-TDS-OFDM meet the demand for improving spectral efficiency, but due to the different frame structures and the presence of inter-symbol interference, the receiver processing for applying the equalization method related to CP / ZP-OFDM requires signal reconstruction, resulting in high computational complexity of the receiver. SUMMARY

[0007] To solve the problems in the prior art, the present application provides an OFDM underwater acoustic communication method based on a pseudo-random noise sequence, which improves the spectral efficiency while also achieving good bit error rate performance. The OFDM receiver processing method proposed in the present application has a spectral efficiency comparable to that of DPN-OFDM and a significantly improved bit error rate performance. Moreover, the effectiveness of the method is verified by computer simulation in the embodiments.

[0008] The technical solution of the present application is an OFDM underwater acoustic communication method based on a pseudo-random noise sequence, comprising a transmitter step and a receiver step:

[0009] The transmitter step comprises the following steps:

[0010] ​Step 1: a series of random bit stream a(t) is generated at the transmitting end, channel coding is performed by using LDPC coding mode and scrambling is performed to obtain b(t);

[0011]

[0012] wherein G represents a generation matrix in LDPC coding;

[0013] Step 2: after the signal b(t) obtained in Step 1 is modulated by using QPSK mapping, inverse Fourier transform is performed to obtain OFDM modulated signal x(t):

[0014]

[0015]

[0016] wherein T is an OFDM symbol period, K is a subcarrier number, Δf is a subcarrier interval, d(t) is an information symbol after QPSK mapping, is a signal after Hilbert transform of b(t); d k is an information symbol d(t) after QPSK mapping on the kth subcarrier to be modulated, and x(t) represents an OFDM baseband signal in an OFDM block period:

[0017] Step 3: the effective data of a tail part of the OFDM modulated signal is copied and moved to the front end of the signal to obtain a CP segment signal, then a signal p(t) obtained by IFFT transform of a PN sequence is added to the most front end of the OFDM signal to constitute a complete OFDM signal; the constructed frame signal is up-converted and transmitted after passing through a channel to complete the work of the transmitting end;

[0018] The receiving end step includes the following steps:

[0019] Step 4: after the signal is received by the receiving end, the channel is estimated by using a PN sequence in the received signal y k ; assuming that the received PN sequence is p'(n) and the transmitted PN sequence after IFFT transform of the transmitting end is q(n), the channel estimation value is:

[0020]

[0021] Step 5: MMSE equalization is performed by using the channel estimation value obtained in Step 4:

[0022]

[0023] wherein σ 2 represents noise power; finally, QPSK demapping and LDPC decoding are performed on the equalized signal to obtain an output bit stream.

[0024] Further, in step 3, the modulated OFDM block signal is represented as s ofdm = [s0, s1, …, s K-1 ] T The CP segment signal is copied from the tail signal segment of the OFDM block signal in which it is located and moved to the front end of the signal, and the PN sequence is represented as N p is the PN sequence length, and p i,k represents the i-th signal in the PN sequence, then the actual transmitted PN-CP OFDM block signal is The complete symbol length is N p +K g +K.

[0025] Further, in step 3,

[0026] Further, the received signal received at the receiving end adopts the following model:

[0027] The signal model at the receiving end is established:

[0028] r = Hs + n

[0029] H represents the channel matrix, n represents the channel noise, and the expression of the channel matrix is:

[0030]

[0031] h = [h0, h1, …, h L ] represents the channel vector, L is the channel length, and K is the number of carriers;

[0032] The received signal is down-converted to a baseband signal y(t) at the receiving end, and the demodulated block form of the receiving end baseband signal is:

[0033]

[0034] F k is a KxK-dimensional discrete Fourier transform unitary matrix, which satisfies is a frequency domain channel matrix, b is a transmitted data bit stream, and w is a noise signal contained in the OFDM demodulation block;

[0035] The frequency domain channel matrix is expressed as:

[0036]

[0037] denoted as channel frequency domain response coefficients; thereby further describing the received signal model as:

[0038]

[0039] Inventive Effects

[0040] The technical effects of the present application are:

[0041] Compared with the existing TDS-OFDM and DPN-OFDM, the PN-CPOFDM frame structure design greatly weakens the influence of inter-symbol interference (ICI) on OFDM symbols due to the presence of CP, so the bit error rate performance of PN-CPOFDM is superior to that of TDS / DPN-OFDM. The bit error rate performances of the three modulation schemes are simulated and compared under two channel conditions, and the simulation parameters are shown in Table 2, and the simulation results are shown in Figure 6 、 Figure 7 It is found from the simulation results that the bit error rate of the proposed modulation scheme is superior to that of the first two modulation schemes. In addition, the BELLHOP underwater acoustic channel simulation software is used to simulate the actual underwater acoustic channel, and the water environment configuration parameters are shown in Table 3, the depth of the seabed is 20 m, and the depth of the sound source and the receiving hydrophone is 10 m. Figure 8 is the normalized impulse response of the underwater acoustic channel. The bit error rate performance comparison of the three modulation technologies in the above simulation underwater acoustic channel is shown in Figure 9 In the simulation, the number of carriers is 1024, the length of the guard interval is 256, the length of the local PN sequence is 256, the mapping mode uses QPSK mapping, and the receiving end uses MMSE linear equalization. Figure 9 The simulation results show that under the condition of low signal-to-noise ratio, the performance of DPN-OFDM is equivalent to that of TDS-OFDM, when the signal-to-noise ratio is large, the bit error rate performance of DPN-OFDM is superior to that of TDS-OFDM, and the bit error rate performance of the proposed modulation scheme improves with the increase of the signal-to-noise ratio, and the bit error rate performance is always superior to that of the first two modulation technologies, which further verifies the performance of the proposed technology scheme.

[0042] Table 2 Simulation parameters

[0043] Parameter Value range Modulation mode QPSK Sampling frequency 48 kHz Bandwidth 13 k-19 kHz Subcarrier number 1024 PN sequence length 256 OFDM symbol period 170.7 ms Guard interval period 42.7 ms

[0044] Table 3 Water environment configuration parameters

[0045] Environmental parameter Setting value Sea surface roughness 2 Sea water density 1021 Sea water depth 20 Horizontal distance between sound source and receiver 1000 BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 Block diagram of TDS-OFDM modulation and demodulation system

[0047] Figure 2TDS-OFDM signal frame structure schematic diagram

[0048] Figure 3 DPN-TDS OFDM signal frame structure schematic diagram

[0049] Figure 4 PN-CP OFDM modulation and demodulation system block diagram

[0050] Figure 5 PN-CP OFDM signal frame structure schematic diagram

[0051] Figure 6 Comparison of performances of three modulation schemes under Gaussian channel

[0052] Figure 7 Comparison of performances of three modulation schemes under multipath channel

[0053] Figure 8 Channel normalized impulse response

[0054] Figure 9 Comparison of bit error rate performances of three modulation technologies in sound channel axis simulation channel DETAILED DESCRIPTION

[0055] The existing conventional CP / ZP-OFDM receiving end processing mode has low data rate and needs pilot subcarriers for estimating a channel and empty subcarriers for Doppler compensation, thereby leading to low spectral efficiency; TDS-OFDM and DPN-TDS OFDM meet the demand of improving spectral efficiency, but due to different frame structures and inter-symbol interference, the receiving end processing for applying the equalization method related to CP / ZP OFDM needs to reconstruct the received signal, thereby leading to high receiving end calculation complexity. In view of these problems, the present application proposes an OFDM modulation technology based on a PN sequence, which improves spectral efficiency and also obtains good bit error rate performance. The overall communication system block diagram is shown in Figure 4 The corresponding signal frame structure is shown in Figure 5 The specific method comprises two parts of a transmitting end and a receiving end:

[0056] The transmitting end step comprises the following contents:

[0057] A random bit stream a(t) is generated at the transmitting end, then a channel encoding is performed by using a low density parity-check code (LDPC) encoding mode and scrambling is performed to obtain b(t);

[0058]

[0059] G represents a generation matrix in the LDPC encoding.

[0060] The signal b(t) after scrambling is mapped and modulated by QPSK to obtain a signal, and then inverse fast Fourier transform (IFFT) is performed to obtain an OFDM modulated signal x(t), wherein T is an OFDM symbol period, K is a subcarrier number, and Δf is a subcarrier spacing:

[0061]

[0062]

[0063] d(t) is an information symbol after QPSK mapping, is a signal after Hilbert transform of b(t). For convenience of representation, the information symbol d(t) after QPSK mapping on the kth subcarrier to be modulated is represented as d k , and x(t) represents an OFDM baseband signal in an OFDM block period.

[0064] The tail part of the OFDM modulated signal is copied and moved to the front end of the signal to obtain a CP segment signal, then a signal p(t) obtained by IFFT transform of the PN sequence is added to the front end of the OFDM signal, thereby forming a complete OFDM signal.

[0065] Specifically, the modulated OFDM block signal is represented as s ofdm = [s0, s1, …, s K-1 ] T , and the CP segment signal is copied and moved from the tail signal segment of the OFDM block signal in which it is located, and here The PN sequence is represented as N p is the PN sequence length, and p i,k represents the ith signal in the PN sequence, and then the PN-CP OFDM actual transmission signal block is The complete symbol length is N p +K g +K. The frame structure is shown in Figure 5 . Then the constructed frame signal is up-converted and transmitted through the channel to complete the work of the transmitting end.

[0066]

[0067] K, N P , K grespectively, S[k] represents the kth baseband modulated data symbol, q(t) represents a pulse shaping filter, which is equal to 1 in a symbol period T, and is used to control the spectrum of the transmitted signal.

[0068] The receiving end steps include the following:

[0069] The receiving end does not need to reconstruct the signal, removes the cyclic prefix part of the receiving end, uses the PN sequence to do channel estimation and equalization, and then outputs the bit stream.

[0070] The receiving end signal model is

[0071] r = Hs + n (5)

[0072] H represents a channel matrix, and n represents channel noise. Its channel matrix still satisfies the cyclic characteristic, and is specifically represented as follows:

[0073]

[0074] h = [h0, h1, …, h L ] represents a channel vector, L is the channel length, and K is the number of carriers. The receiving end first converts the received signal into a baseband signal y(t) through frequency down-conversion. The receiving end baseband signal demodulation block form is:

[0075]

[0076] F k is a K×K-dimensional discrete Fourier transform unitary matrix, which satisfies is a frequency domain channel matrix, b is a transmitted data bit stream, and w is a noise signal contained in the OFDM demodulation block.

[0077] The frequency domain channel matrix is represented as follows:

[0078]

[0079] is represented as a channel frequency domain response coefficient. Based on this, the receiving signal model can be further described as:

[0080]

[0081] Generally, when a signal passes through an unknown channel, the receiving end needs to estimate the channel and use equalization to eliminate inter-symbol interference before demodulation. First, the PN sequence in the received signal y k is used to estimate the channel. Assuming that the received PN sequence is p'(n) and the transmitted PN sequence after IFFT transformation at the transmitting end is q(n), the channel estimation value is:

[0082]

[0083] MMSE equalization is performed using the estimated channel values.

[0084]

[0085] where σ 2 represents the noise power. Finally, QPSK demapping and LDPC decoding are performed on the equalized signal to obtain an output bit stream.

[0086] The technical scheme of the present application is further explained and described below with reference to a specific embodiment:

[0087] The system block diagram corresponding to the modulation technique of the present embodiment is shown in Figure 4 The method involved includes the following steps:

[0088] Step 1: A modulation scheme using QPSK mapping is adopted considering the requirements of data rate and spectral efficiency. To enhance the anti-interference capability, LDPC encoding is adopted with an encoding efficiency of 1 / 2. The number of carriers K is selected to be 1024, the length of PN sequence Np is selected to be 256, and the length of cyclic prefix is 1 / 4 of the number of carriers. First, an LDPC encoding check matrix E and a generating matrix G are generated according to the length N of the locally randomly generated bit stream, the row weight row, and the column weight col, wherein the check matrix is a sparse matrix, and an example can be represented as follows, the row weight of the check matrix is 3, and the column weight is 2.

[0089]

[0090] Then, the generating matrix G is obtained according to a specific encoding criterion, and the encoded code word b

[0091]

[0092] represents XOR.

[0093] The spectral efficiency can be obtained as shown in Table 1 using the modulation order, encoding efficiency, and the number of carriers.

[0094] Table 1 Comparison of spectral efficiency of three modulation schemes

[0095] Modulation mode DPN-OFDM TDS-OFDM PN-CPOFDM Spectrum efficiency η 66.64％ 79.98％ 66.64％

[0096] Step 2: QPSK mapping and OFDM modulation are performed on the encoded information.

[0097] It is assumed that the interval between adjacent carriers is Δf. In the baseband system, the frequency of the kth subcarrier is represented as

[0098] f k= kAf, k = 0, 1, 2,..., K - 1 (14)

[0099] d k denotes the information symbol after symbol mapping to be modulated on the kth subcarrier. K symbols transmitted in parallel are called an OFDM block. T denotes the period of an OFDM block, and the OFDM baseband signal in one period is denoted as

[0100]

[0101] Step 3: locally generate the PN sequence and perform IFFT transformation.

[0102]

[0103] Step 4: copy the data of 1 / 4 length after the effective data part to the data header to form the CP-OFDM frame structure, and then add the PN sequence to the front end to form the PN-CP-OFDM frame structure, as shown in Figure 5 The signal after up-conversion at the transmitting end is denoted as:

[0104]

[0105] Step 5: the receiving end extracts the PN sequence for channel estimation and equalization. The received signal model is:

[0106] r = Hs + n (18)

[0107] H represents the channel matrix, and n represents the channel noise. The channel matrix still satisfies the cyclic property, and is specifically denoted as:

[0108]

[0109] The receiving end first converts the received signal into a baseband signal y(t) through down-conversion and other operations. The demodulation block form of the baseband signal at the receiving end is:

[0110]

[0111] is the frequency domain channel matrix, and w is the noise contained in the block.

[0112] The frequency domain channel matrix is denoted as:

[0113]

[0114] is denoted as the channel frequency domain response coefficient. Based on this, the received signal model can be further described as:

[0115]

[0116] The PN sequence is extracted from the received signal for channel estimation. Assuming that the received PN sequence is p'(n) and the transmitted PN sequence after IFFT transformation is q(n), the channel estimation value is:

[0117]

[0118] MMSE equalization is performed using the estimated channel value.

[0119]

[0120] Step 6: QPSK demapping and LDPC decoding are performed on the equalized data, a bit stream is output, and the bit error rate is calculated by comparing with the initial bit stream.

Claims

1. A method of OFDM underwater acoustic communication based on pseudo-random noise sequences, characterized in that, The method comprises a transmitting end step and a receiving end step: The transmitting end step comprises the following steps: Step 1: generating a random bit stream at the transmitting end , using LDPC encoding mode for channel coding and scrambling, obtaining ; Wherein G represents a generation matrix in LDPC coding; Step 2: The signal obtained in Step 1 is mapped using QPSK modulation After QPSK mapping modulation, inverse Fourier transform is performed to obtain the OFDM modulated signal : wherein is an OFDM symbol period, is a number of subcarriers, is a subcarrier spacing, is an information symbol after QPSK mapping, is a signal after Hilbert transform on ; is an information symbol after QPSK mapping on the kth subcarrier to be modulated , denotes an OFDM baseband signal within an OFDM block period: Step 3: copying and moving the effective data of the OFDM modulated signal tail to the front end of the signal to obtain a CP segment signal, and then performing IFFT transformation on the PN sequence to obtain a signal adding to the front end of the OFDM signal to form a complete OFDM signal; and performing frequency up-conversion on the constructed frame signal and transmitting the frame signal through a channel to complete the work of the transmitting end. The modulated OFDM block signal is represented as , CP segment signal is copied from the tail signal segment of the OFDM block signal in which it is located and moved to the front end of the signal, and the PN sequence is represented as , is the PN sequence length, represents the signal in the PN sequence, then the actual PN-CP OFDM transmitted signal block is , and the complete symbol length is , ; The receiving end step comprises the following steps: Step 4: After receiving the signal, the receiving end estimates the channel by using the PN sequence in the received signal Assuming the received PN sequence is , and the transmitted PN sequence after IFFT transformation at the transmitting end is , the channel estimation value is: Step 5: MMSE equalization is performed by using the channel estimation value obtained in step 4: wherein represents the noise power; and finally, the equalized signal is QPSK demapped and LDPC decoded to obtain an output bit stream.

2. The pseudo-random noise sequence based OFDM underwater acoustic communication method of claim 1, wherein, The received signal received by the receiving end adopts the following model: The signal model of the receiving end is established: representative channel matrix, representative channel noise, the expression of the channel matrix is represents the channel vector, L is the channel length, and K is the number of carriers. The receiving end converts the received signal into a baseband signal through down-conversion The demodulation of the baseband signal of the receiving end is in block form is a DFT matrix of dimension is a DFT matrix of dimension , is a frequency domain channel matrix, is a stream of transmitted data bits, is a noise signal contained within the OFDM demodulation block; Frequency domain channel matrix The expression is: denoted as channel frequency domain response coefficients; thereby further describing the received signal model as: 。

Images