Why does Kademlia structure its routing table how it does?
问题 I understand that the Kademlia routing table is made up of 160 buckets. Nodes are put into buckets 0-159, depending on their prefix length (which is the number of leading unset bits in the XOR of the local node key and the node). Why is this so, is there any performance benefits involved (other than the fact that iterating through 160*20 nodes to find the closest is infeasible)?. 回答1: Kademlia uses the XOR of 2 node IDs as a measure of their distance apart. The idea with the routing table