the differences between the unstructured and structured p2p network

the unstructured is relatively easy to maintain, such as node join and leave, and supports complex queries . but it cannot guarantee the total recall of an search. the main problems in unstructured research area is to increase the recall within an acceptable time, reduce the use of bandwidth, and load balance..

oppositely ,the structured is very difficult to maintain ,as it uses structured overlay networks and DHT to retrieve data, it is just like a Internet-scale hash table, or a large distributed database. each data fragment is assigned to a specified node according to one protocal, that's why it's hard to maintain, following a node join or leave ,there is always a data reassignment between nodes. also, each data is identified by it's key, and this key does not support part query, that's the shortcoming derived from hash table. so , the main problem in structured research area is to find a new protocal of data placement( good balance,easy search,easy maintain), to find a way to support the part query(new hash algorithm?)