The Overlay Networks in Peer-to-Peer Architecture
P2P architecture began to proliferate on the internet relatively recently owing to the need to share data. The main characteristic of P2P networks is that they do not rely directly on the services of an always on server. The end systems usually communicate directly, and peers are intermittently connected and exchange IP addresses. In hybrid forms of P2P architecture, a server may be incorporated to handle some tasks.
In terms of physical architecture, it would be expected that the peer is linked to the closest physical neighbour via IP address. However, in P2P architecture, small portions of files are distributed among clients. Thus, a higher speed channel may allow a client to be more productive in the network due to the accumulation of larger file portions. This leads to the formation of another pattern within the network. The network IDs are issued from a field distribution table, and thus, the keys are assigned depending on the number of files being held. This new priority leads to the formation of a network based on accumulated content, also referred to as an overlay network.
In terms of performance, file distribution time is significantly better with the P2P network when compared with traditional client-server networks. This is because in the client-server network, the time to distribute or retrieve a file increase as the number of requests increases. In P2P architecture, this time gradually reaches an optimum level where the time taken is significantly reduced owing to the presence of a large number of users with portions of the file.