The growth of the electric demand and the extensive integration of renewable energy sources pose a problem concerning the control and stability of the grid. To solve this issue several methods and algorithms such as Demand Response, Direct Load Control and, more recently, Dynamic Demand Control (DDC), where the power of individual devices adapt to the global system load, have been proposed. These techniques applied to the demand-side of the grid form what it is known as Demand-Side Management. In this talk, in addition to global information about the system (provided by the electric frequency), we consider local interactions based on the communication among smart appliances belonging to the same household in DDC protocols. The methodology is explained and a comparative study with the case without interactions is presented. A considerable reduction of both frequency fluctuations and the total number of pending tasks is observed, overall improving the performance of the power grid.