In active sound control, a domain is supposed to be protected from noise generated outside via constructing secondary sound sources also known as controls. The controls are applied on the boundary of the domain to be shielded. Apart from the noise, a desired sound generated by interior sources is also allowed to be present inside the shielded domain. If the desired sound is present, it is required to remain it unaffected by the control. This problem becomes much more complicated since the secondary sources have a reverse effect on the input data. To take into account the reverse effect, a novel algorithm based on nonlocal control is realised in the frequency and time domains for the first time. The technique is based on a projection property of the CalderoÂ´n surface potentials. As a result, the noise component can be separated automatically from the total field. In previous research, the theoretical framework for the algorithm has been provided. In this thesis, the applicability of the algorithm is demonstrated via a series of numerical experiments. A real-time practical realization of the algorithm presumes a preliminary tuning to the real surrounding conditions. A number of test cases is considered including broadband noise and monochromatic desired sound to be retained. A sensitivity analysis is carried out with respect to some key design parameters such as the density of sensors and controls as well as their respective geometrical displacement from each other determined by the Hausdorff distance. In addition, the characteristics of the noise sources and shielded domain are also discussed. It is demonstrated that the nonlocal control is able to provide the level of noise attenuation not sensitive to the presence of desired sound.