Resource And Implement Of Resource Directory Server

There is about 2 billion nodes in the Internet and it will grow up to 7 billion as the expert predicts and this will be a disaster for DNS:the exploding node will bring great pressure for DNS and decrease the security at the same time,in addition,it will take a long time to redesign the architecture of DNS.On the other hand,it is typical M2M communication for nodes in constrained network which could not be satisfied by DNS.So,referring the draft of IETF,this paper come up with a resolution for the discovery of resource in constrained network which named the resource directory server.The paper also resolve the communication problem of the sleeping nodes.Taking the constrained network as the background,the paper firstly gives a detailed layout design on the base of the analysis the difference of resource discovery between constrained network and traditional network and the constrained situation provided by OMA.The paper depart the system into three part:the CoAP module,the resource discovery module and the support of sleeping nodes module.The CoAP module provides communication with CoAP,the resource discovery module provides the mechanism of discovery of resource in constrained network,the support of sleeping nodes module realize the gain of resource when the nodes are sleeping.After that the paper gives detailed design of the system.The CoAP module defines most of data structure and the interface,it also provide reliable transform,congestion control,the match of request and response and the process of the request.The discovery module provides the discovery of the resource directory server,the registration of the resource,the update and delete of the resource and the lookup of the resource.In the support of the sleeping node module,the paper firstly divide the resource into the representable resource and the un-representable resource.For the representable resource the nodes can register their state to the resource directory server so that the client can also access the resource when the nodes are asleep.For the un-representable resource,the system cache their request and response to solve the problem,the system can also take different action according to the memory resource.The paper also present a memory replacement algorithm based on statistic to increase the performance of the system.At last the paper deploy the system and test them in a Linux environment,the paper first test the validity and feasibility of the resource discovery project and optimize the system based the result of the test.Then the paper compare the statistic data between using the project and without the project and analysis the result.