Design of underwater acoustic communication system based on OFDM

The most difficult problem faced by shallow-water high-speed underwater acoustic communication is strong multipath and rapid time variation caused by ocean surface reflection, internal waves, etc. Among them, the amplitude fading of the received signal caused by multipath, the inter-symbol interference of the received signal caused by multipath, coupled with the ocean environmental noise, low carrier frequency, extremely limited bandwidth and the time-space-frequency variation characteristics of the transmission conditions, make the underwater acoustic channel the most difficult wireless communication channel to date [1-2]. The underwater acoustic channel has serious multipath delay. The general multi-carrier technology requires a good channel estimation and equalization technology at the receiving end to achieve a reply signal with very little distortion. The orthogonal frequency division multiplexing technology greatly improves the anti-multipath characteristics by adding a cyclic prefix to the signal at the transmitting end. The main idea of ​​this technology is to divide the available frequency band channel into multiple orthogonal sub-channels, and perform parallel transmission on each divided sub-channel, thereby reducing the signal transmission rate on the channel. The signal bandwidth is smaller than the coherent bandwidth of the channel, thereby greatly eliminating inter-symbol interference, and the carriers on the sub-channels have partial overlap, which improves the utilization of the frequency band. This technology has been widely used in underwater communications.

1 OFDM Principle

Weinstein proposed using DFT (Discrete Fourier Transform) to implement modulation and demodulation of OFDM (Orthogonal Frequency Division Multi-plexing) system [3]. When transmitting binary data at the transmitter, the data is first mapped into a set of complex sequences {d0, d1,-,dN-1} by modulating each subcarrier, where dn = an + jbn. If the above complex sequence is transformed by IDFT, a new complex sequence {S0, S1,-,SN-1} will be obtained, where:

At the receiving end, the received signal is sampled at a time interval of Δt, and then DFT transform is performed to restore the original complex sequence {d0, d1,-,dN-1}, and then after carrier deconvolution, the original binary data can be restored. For the calculation of IDFT/DFT transform, the more mature IFFT/FFT algorithm is usually used to implement it [4-5]. In this way, the use of fast Fourier algorithm can greatly reduce the amount of calculation to improve the system operation efficiency.

2 High-speed underwater acoustic communication based on OFDM

2.1 System Block Diagram

The implementation process of the OFDM underwater acoustic communication system is shown in Figure 1. At the transmitting end of the underwater acoustic communication system, in order to combat the errors caused by randomness and burstiness of the underwater acoustic channel, the input binary digital signal is firstly subjected to channel convolution coding and interleaving, and then the digital signal on each subcarrier is carrier mapped through serial/parallel conversion; then the pilot information for channel characteristic estimation is inserted; and then the OFDM modulated signal is formed through IFFT transformation; in order to better combat the multipath effect of the underwater acoustic channel, a cyclic prefix greater than the maximum channel delay is added after the formed OFDM symbol to ensure that the received signal is not subject to inter-code interference and the orthogonality between the subcarriers is ensured; finally, the digital signal is converted into an analog signal through D/A conversion, and the signal is transmitted in the underwater acoustic channel through ultrasound after RF modulation. At the receiving end, the opposite process to that of the transmitting end is carried out to finally restore the original data.

2.2 Cyclic Prefix

Since the channel will cause intersymbol interference (ISI) and interference between channels, the orthogonality between subcarriers will be destroyed, and the signals on each subcarrier cannot be separated by FFT at the receiving end. Although multi-carrier modulation can enhance the system's ability to resist ISI, the current symbol still overlaps with the previous symbol due to the delay, thus generating ISI. In order to reduce the impact of ISI on the signal, a guard interval (GI) should be added in front of each symbol. The guard interval will make the multipath interference signal generated by the previous symbol disappear before the current symbol reaches the receiver, thereby overcoming the impact of ISI. If the information in the guard interval is set to empty, the orthogonality between the subcarriers will be lost due to the influence of multipath propagation, resulting in intercarrier interference (ICI). In order to eliminate ICI caused by multipath propagation, the original OFDM symbol with a width of T is periodically extended. This guard interval using a periodic extended signal is called a cyclic prefix (CP). Figure 2 shows an OFDM symbol with a cyclic prefix, which copies the length Tg of the OFDM symbol period to the guard interval of length Tg. In this way, as long as the cyclic prefix length is greater than the channel delay, ISI will not be caused.

2.3 Modulation and demodulation methods

OFDM is a multi-carrier modulation method. Each subcarrier can select a different modulation method according to the channel conditions. When the reliability of transmission is given priority, the modulation method MPSK with a lower bit error rate is selected, such as BPSK and QPSK; and when the system transmission rate is considered, the modulation method MQAM with a higher spectrum utilization rate can be selected, such as 8QAM and 16QAM.

2.4 Channel Coding Technology

OFDM technology can overcome inter-code interference and frequency attenuation caused by multipath delay, but it cannot solve the flat fading of amplitude. In addition, in underwater acoustic channels, the influence of the noise environment will cause bit errors in the transmitted signal, which will reduce the reliability of communication. In order to improve the quality of communication, channel coding should be performed at the front end of the system. Convolutional codes are the first choice for this solution because of their good error correction performance. At present, they are used in many communication systems. (2,1,7) code is the preferred standard convolutional code using Viterbi decoding, which has the function of minimizing the bit error rate of the relevant communication system and overcoming phase errors.

2.5 Pilot-based channel estimation

Since many channels cannot directly transmit baseband signals, in order to better adapt to the channel, most actual communication systems must use modulation technology. Different modulation methods have different corresponding demodulation methods. The main demodulation methods are incoherent method, coherent method and differential coherent method commonly used in differential coding. The use of differential coherent method and incoherent method can avoid channel estimation, but for multi-level high-speed underwater acoustic communication, coherent demodulation requires carrier information with the same frequency and phase as the transmitter, otherwise it cannot be demodulated correctly. Therefore, channel estimation must be performed, and channel estimation can correct useful data. This paper adopts block pilot insertion method to estimate the channel.

2.6 OFDM Parameter Selection

In an OFDM system, the following parameters need to be determined: guard interval, symbol period, and number of subcarriers. These parameters depend on the required channel bandwidth, delay spread, and required information transmission rate. The various parameters of an OFDM system are generally determined in the following way [7]:

(1) Determination of the protection interval: The protection interval must be greater than the maximum delay spread of the information.

(2) Selecting the symbol period: Generally, the symbol period length (excluding the guard interval length) is selected to be 4 times the guard interval length.

(3) Number of subcarriers: The number of subcarriers can be calculated by dividing the -3 dB bandwidth by the subcarrier spacing (i.e., the inverse of the symbol period after removing the guard interval). The number of subcarriers can also be determined by dividing the required bit rate by the bit rate in each subchannel. The bit rate transmitted in each subchannel is determined by the modulation type, the coding rate, and the symbol rate.

The OFDM parameters used in this paper are shown in Table 1.

3 Communication Simulation Experiment

In order to verify the performance of the underwater acoustic OFDM communication system, this paper uses Matlab 7.1 software to simulate the algorithm. The simulation parameters are shown in Table 1. The signal transmitted by the OFDM system in this paper is random binary data 0 and 1 generated by Matlab. First, the signal is encoded and interleaved, and then each subcarrier is baseband modulated, i.e. mapped. As shown in the constellation diagram in Figure 3, the data is allocated at a fixed position in the constellation space, which is consistent with the theoretical value.

After FFT transformation, the constellation diagram has completely changed, and the data cannot be located at the exact position in the constellation diagram. Therefore, channel estimation is required for the data. Figure 5 is the constellation diagram of the signal after channel estimation. Figure 5(a) and Figure 5(b) are the constellation diagrams after channel estimation for BPSK and 8QAM, respectively.

Figure 6 is a comparison of the bit error rate curve of the OFDM system under 8QAM baseband modulation and the bit error rate curve of the single carrier under 8QAM modulation. It can be seen from the figure that the bit error rate is very high under single carrier modulation, and the error rate of the received signal is high after the signal passes through the multipath interference of the underwater acoustic channel, while the bit error rate is significantly reduced under the OFDM communication system. This shows that the OFDM system has obvious anti-multipath interference performance. The bit error rate will drop significantly after increasing the signal-to-noise ratio, while the single carrier system will not reduce the bit error rate when the signal-to-noise ratio is increased. It can be seen that the OFDM system has obvious advantages in resisting multipath interference than the single carrier system.

After the subcarrier is demodulated, the deinterleaving decoding is performed to restore the original data. The communication bit error rate is shown in Figure 7. Figure 7 (a) shows the bit error rate under the BPSK modulation mode, and Figure 7 (b) shows the bit error rate curve under the 8QAM modulation mode. It can be seen that the bit error rate under the BPSK mode is lower than that under the 8QAM modulation mode. This verifies that when reliable transmission performance is required, the BPSK modulation mode is selected, and when a high rate is required, the 8QAM modulation mode with higher spectrum utilization is selected.

4 Conclusion

OFDM is a high-speed communication technology suitable for multipath fading and limited bandwidth channels. It has been widely used in the radio field, but has been rarely used in underwater acoustic communication. This paper applies OFDM communication technology to underwater acoustic communication, designs an OFDM-based underwater acoustic communication system, and verifies through analysis and simulation that OFDM-based underwater acoustic communication has a strong ability to resist multipath interference.