Study On Self-Stabilizing One Hop DHT Based Data Storage

As more and more services are introduced to telecom core network, the user service profiles are increasing rapidly. However, Due to the centralized design of HSS, the capacity of HSS is constrained by the equipment's single point processing ability which is very difficult to increase.To solve the problem, we suggest to introduce distributed technologies to optimize data storage equipment. Existing distributed technology are often unable to meet telecom core network's demands for high reliability and high availability, therefore we make some optimization and improvement to the existing One Hop DHT protocol so as to introduce a new Self-Stabilizing One Hop DHT protocol. This protocol is based on DHT mechanism, but different from most of the DHT protocol, the protocol requires every node to save the route information of all the nodes in the network, in this way, a lookup can be routed within only one hop, rather than O(n) hops in other DHT protocols. The protocol is also highly reliable, it can stabilize into a steady-state itself even the network is churning(Nodes leave and packet lost).In order to prove the feasibility of the algorithm, we have established a routing maintenance model, then we make a theoretical analysis of the node routing latency, the update delay of the node's routing table, as well as the node's routing accuracy and so on. We also model the election process in the protocol and analyzed the stability of the election process using the self-stabilization theory. Through the analysis above, we come to the conclusion that the protocol is able to route the message within one hop, and self-stabilized. Finally, we implemented a simulation system with Planetsim, we take data set from real network measurement for input of the simulation system, and the simulation result verifies our conclusion.