Deblending is the dominant way to deal with simultaneous-source data. Each deblending method requires some differences between signal and interference (e.g., coherency difference, dip difference). Those differences are controlled by survey parameters. Except for the encoding scheme, there are still many parameters in simultaneous-source survey controlling the coherency difference. Inappropriate survey parameters lead to unsatisfactorily deblending results. Optimizing survey parameters to obtain the best deblending performance requires understanding the influence of survey parameters on deblending performance. To address the influence of simultaneous-source survey parameters, we simulate the blended datasets with different survey parameters (shot interval, shot number, dithering range and source distance). Inversion methods with two well-known constraints (sparsity and rank-reduction constraints) are implemented to separate the blended data. From the deblended results, some preliminary conclusions can be drawn. Large shot interval reduces the coherency of signal, which is bad for deblending. In common receiver gathers, small shot number shortens the coherency difference between signal and interference, making deblending difficult. A large dithering range can improve the interference incoherence, which is easy to be suppressed by coherency-pass constraint. Large source distance reduces the degree of overlap, which can enhance separation. These conclusions can give the guidance of simultaneous-source survey design, which is beneficial for the successful deblending.