The reason is to reduce the size of the routing table and meet the needs of different scale networks. IP addresses are classified into five categories by prefix. Level A network: The prefix is 0. high 8 bits are used as net-id and low 24 bits as host-id. B network: The prefix is 10. high 16 bits are used as net-id and low 16 bits as host-id. C network: The prefix is 110. high 24 bits are used as net-id and low 8 bits as host-id. Category D network: The prefix is 1110. they are used as multicast addresses. Category E network: The prefix is 1111. they are used as tests.