The Remote Transparent Transport Platform Based On Socket Communication Mechanism

In existing Internet uses, the corporation network is realized through the virtual LAN on the basis of Internet. The corporation network is mainly used to contact and communicate each other in the subsidiary companies all around the world. Therefore, a remote transport platform is wanted. Through the platform, the subsidiary companies of the whole corporation can expediently and reliably exchange the information and share the data. So-called "transparent" in the article means that the transport of the information and data in the lower layer is invisible to the users. The major problem in the paper is how to transfer the information and data expediently and reliably to the destination. The problem is divided into three parts. Firstly, the paper discusses the familiar star-shaped network structure, analyze the reliability of the network structure, and then design the multi-link redundancy network structure so that the corporation network transmits data packets to the next node through the redundancy links. Secondly, the system settles the problem how to rapidly find out failures and round the failures, that need dynamically detect the links' state, and unify the whole network's link state through the valid synchronization. Thirdly, the platform transmits data to next node through the best link according to the route-choosing algorithm. Among the numerous methods transmitting messages and communicating data, SOCKET communication mechanism is in common use. SOCKET communication mechanism is used to design the major messages in soft routing. These messages are encapsulated by different data format, some use IP data packet, some use UDP data packet. The various formats are chosen according to the demand on the efficiency. At the same time, for the sake of controlling network congestion, the article use a weighted round-robin queuing method, which joints Custom Queuing (CO) and Weighted Fair Queuing (WFQ). The fashion can make the interactive, affairs-oriented applications not be delayed, as well as the block of transportation not be starved. In the paper the corresponding treatment methods are put forward on measuring and treat with that breakdown, too. And then, the whole network can quickly measure through setting some timers where the failures come forth, what troubles are offered, and may recover effectively when these troubles disappear. The route-choosing algorithm adopts Dijkstra shortest-path algorithm that can choose effectively the valid path according to dynamic link-detecting results,...