The probability is very small that taking a vector randomly from the permutation set S makes it with a known vector is zero correlation. For example, when n=7, the probability of one success is about 0.0365; when n=8, the probability is about 0.0232; the larger the n, the smaller the probability. Not to mention zero correlation (or orthogonality) to multiple vectors at the same time. If there is no exchange algorithm, OLHD can not be constructed. So, the exchange algorithm is the core of the technical core of the construction of the orthogonal hypercube. If there are no ghosts in her mind, as a doctoral dissertation, how can she just use the three letters "get" to omit the most basic introduction and discussion for the exchange algorithm principle and operation method Including the most basic derivation step (5.1) that reflects the exchange behavior formula (6.1).