“…In the proposed method, the parameter p is selected so that the information obtained by a single observation is maximized, which can be achieved when the number of zeros in the binary observation, such as y B 1 and y B 2 , is the same as that of ones. Note that a similar approach has been originally proposed for the pool size control in adaptive group testing [7,8]. The probability that a single binary observation y B becomes zero can be approximated by the probability that all bits sent by L tags are zero, i.e.,…”