This thesis provides an overview of the standard randomized benchmarking protocol, the extension to interleaved randomized benchmarking and the related unitarity randomized benchmarking, used to characterize noisy implementations of quantum gates from the Clifford group. It contains a brief overview of the textbook and literature preliminaries that are needed to understand the description and proof of these protocols. It outlines how previous work has reduced the number of random sequences needed in standard randomized benchmarking. Then it presents the new result of reduced number of sequences for unitarity randomized benchmarking and discusses the implication of these new results.