Sound field reconstruction with sparse sampling and the equivalent source method

Abstract

Abstract The equivalent source method is reformulated under the compressive sensing framework for tackling with the spatially extended sound source. The sound field to be reconstructed is first modeled using the equivalent source method according to the topology of sound source, and then the singular value decomposition is used to obtain a series of orthogonal basis of the sound field. Utilizing the sparse property of the basis and combining with the compressive sensing theory, the sound field is finally reconstructed from sparse sampling. Its performance is investigated by using simulations and an experiment, the results show that the sparsity of the solution is greatly enhanced by using the proposed method, and the sound field can be accurately reconstructed with sparse sampling.