Research On The Key Technology Of OFDM Underwater Acoustic Communication For The Small-size AUV Platform

With the rapidly increasing demands for ocean information perception and acquisition,autonomous underwater vehicle(AUV)has been widely used in marine resource exploration,marine hydrological environment investigation,seabed topographic mapping and many other related fields.Compared to the remoted operated vehicle(ROV)that need a cable for teleoperation,the AUV features higher flexibility for underwater operation and deployment due to its autonomous decision capability and cable-less teleoperation.As an important branch of AUV family,small-size AUV(generally within 3 m in length)is drawing more and more attention because of its significant advantages of low cost,deployment-convenience and easy-grouping.However,while the current commercial underwater acoustic modems are too large or too expensive for small-size AUV,there is a lack of underwater acoustic communication payload suitable for this scenario,thus posing a significant challenge for the research community.The Orthogonal frequency division multiplexing(OFDM)has become a hot research topic in underwater acoustic communication due to its superior capability of high spectrum efficiency,high data transmission rate and strong multipath resistance.However,underwater acoustic OFDM communication system is extremely sensitive to Doppler effect,which becomes much more challenging as the small-size AUV tend to experience more complicated moving state.Moreover,the computational capability onboard small-size AUV is limited,thus excluding the adoption of Doppler estimation and compensation method associated with huge computational complexity.In order to improve the information transmission robustness of underwater acoustic OFDM communication system on the small-size AUV platform,in this thesis the Doppler estimation and compensation algorithms of underwater acoustic OFDM were investigated to achieve a higher data transmission rate while taking into account the computational complexity.The main contributions of this thesis are as follows:1.In view of the large computational overhead in the implementation of classical cross-correlation Doppler estimation algorithm,a quasi-gradient cross-correlation Doppler estimation algorithm is proposed,which transforms the Doppler estimation problem based on cross-correlation function into a stochastic quasi-gradient optimization problem.The smoothed quasi-gradient and variable Doppler search interval are adopted to accelerate the convergence rate of optimization process and avoid the cost function from converging to the local optimum effectively.2.To address the mismatch between the Doppler estimation at signal level and the demodulation results at bit level in dynamic or scattered multipath channels,a bit-level Doppler estimation algorithm based on the minimum bit error rate(BER)criterion is proposed,namely BER direct search Doppler estimation scheme.By searching the minimum demodulation BER with respect to Doppler grids,the proposed approach can transform the Doppler estimation at signal level into BER search at bit level.Moreover,a BER gradient descent search Doppler estimation approach is designed to simplify the grid-by-grid Doppler search via gradient iterative search process.Meanwhile,smoothed gradient and variable Doppler increment are adopted to accelerate the convergence rate of the iterative process and avoid the solution process from converging to the local optimal.3.To overcome the high implementation complexity of the conventional Doppler compensation scheme,a low-complexity Doppler compensation scheme based on variable clock frequency is proposed.In this scheme,the sampling rate at the process of analog to digital converter(ADC)is controlled by directly adjusting the clock frequency of the external clock source of Inter-IC Sound(I2S)bus in the integrated circuit.In this way,the Doppler compensation can be simultaneously performed during the process of ADC,thus avoiding the computation needed in the classic resampling method.The effectiveness of the approaches proposed in this thesis have been verified through numerical simulation and sea trial experiments.The results show that,the above-mentioned Doppler estimation and compensation algorithms can significantly improve the performance of underwater acoustic OFDM communication under acceptable computational overhead.