Synchronization of Peers in Peer-to-Peer Publish/Subscribe Systems

In distributed systems, a group of multiple processes are cooperating with one another by exchanging messages in networks. In this paper, we consider a peer-to-peer (P2P) model of a topic-based publish/subscribe (P2PPS) system composed of peer processes (peers). Each peer pi can both subscribe a subscription p_i. S and publish an event message e with a publication e.P, which are specified in terms topics, i.e. Keywords. An event message e is notified to a target peer p_i whose subscription p_i.S includes a common keyword with the publication e.P. Each event message e carries a vector e.V = (V₁, , V_m) of keywords k₁, , k_m. An event message e₁ causally precedes an event message e₂ with respect to a subscription S_i iff e₁ causally precedes e₂ with respect to the send-receive relation and e₁.V_h <; e₂.V_h for every keyword k_h in the publications e₁.P and e₂.P and the subscription p_i.S. If a pair of event messages e₁ and e₂ are ordered so that e₁ precedes e₂ even if e₁ does not causally precede e₂, the event messages e₁ and e₂ are unnecessarily ordered. In this paper, every pair of messages are causally delivered to every common target peer by using the linear clock and keyword vector to reduce pairs of messages unnecessarily ordered. We evaluate the keyword vectors in terms of number of messages unnecessarily ordered.