The basic idea of the binary search algorithm is that when multiple tags enter the reader workspace, the reader sends a query command with restrictions, and the tag answers satisfying the restrictions. If a collision occurs, the restrictions are modified according to the bit where the error occurs, and the query command is sent again until a correct reply is found. Answer and complete the read and write operation of the tag. Repeat the above operation until all labels are read and written.