Self-adaptive rank selection method

The invention provides a self-adaptive rank selection method. The self-adaptive rank selection method comprises the following steps of: A) acquiring the instantaneous rank information of a current frame; B) judging whether the instantaneous rank information of the current frame is the same with the actual rank information of the current frame or not, if so, keeping the count value of a rank change accumulator unchanged, otherwise, adding 1 to the count value of the rank change accumulator; C) judging whether the instantaneous rank information within a feedback period is accumulated or not, if so, implementing the step D), otherwise, returning to the step A); and D) calculating the ratio of the count value of the rank change accumulator to the number of frames in the feedback period, judging whether the ratio is larger than a first threshold value which is preset or not, if so, updating the rank information in the next transmission period, otherwise, keeping the rank information in the next transmission period