An ML estimation based robust Chinese remainder theorem for reals

In this paper, we consider the CRT problem for real numbers with noisy remainders that follow wrapped Gaussian distributions. We propose the maximum likelihood (ML) estimation based CRT when the remainder noises may not necessarily have the same variances. The proposed algorithm only needs to search for the solution among L elements, where L is the number of remainders. We compare the performances of the newly proposed algorithm and the existing algorithm in term of numerical simulations. The results demonstrate that the proposed algorithm not only has a better performance when the remainders have different error levels/variances, but also has a much lower computational complexity.