A new method called Linear Construction (LC) has been developed to quickly calculate the covariance matrix for a two-point galaxy correlation function. By using small random catalogs and combining them in a linear way, the method can estimate the covariance for any catalog size. This approach has been tested with simulations and found to be unbiased. With this method, the speed of covariance estimation can be increased by a factor of 14, which can improve the accuracy of parameter estimation in scientific studies.