At the first the infection of the same separation degree to different network's transmission was discussed, then the infection of different separation degree to same network's transmission was discussed by giving an example of small world network; Internet separation degree based on hops was defined. Based on the data that were supported by 30 monitors of global Internet from CAIDA, Internet hops was analyzed by statistic, Internet separation degree was computed. Small world character of Internet was illustrated more amply than the other international research result in this field; Aim at hops and RTT, 10 kinds regression analysis were done and a conclusion was draw that it more accords with power function between hops and RTT, the key parameters were found, which proved Internet complex network transmission is power function. Internet separation degree time sensitivity model (ISDTSM) was created; The model of time series stochastic process of Internet separation degree (MTSSPISD) was created, and then the time evolvement rule of Internet separation degree was discussed; At last, the Internet separation and average transmission time of IP level data during 2008 Beijing Olympic Game were forecast by using model ISDTSM and model MTSSPISD.