Network node and buffer device, and scheduling method

Abstract

The invention discloses a buffering device which comprises at least one internal conversion unit and at least one basic buffering device. The internal conversion unit at least comprises two input ends and two output ends. The internal conversion unit is connected with the basic buffering device in close way at intervals through one input end of the two input ends and one output end of the two output ends. The other input end of the internal conversion unit receives light waves; the internal conversion unit outputs the light waves according to a first control signal and the basic buffering device is used for buffering the light waves from the internal conversion unit. Meanwhile, the invention further provides a network node and a regulating method. By employing the technical proposal of the invention, the scale of the network node is small, thus being easy for practice, reducing loss rate of data packets as well as avoiding blocking of head of line at the same time.

Description

technical field [0001] The invention relates to the technical field of optical networks, in particular to a network node, a cache device and a scheduling method in an optical packet switching network. Background technique [0002] With the rapid growth of packet services, future communication network design should focus on optimizing packet support. Optical network packet switching technology can broaden the existing network bandwidth, and network nodes have high information throughput, suitable for supporting burst Packet services with strong characteristics and frequent service changes, especially IP services. [0003] Now, the optical packet switching network is generally composed of edge nodes and core nodes, which are connected by dense wavelength division multiplexing DWDM fiber links, and data packets are transmitted from edge nodes to core nodes through DWDM fiber links, and then exchanged The modules are exchanged and finally sent to the destination edge node. [...

AI Technical Summary

Problems solved by technology

The disadvantage of this feed-forward connection scheme is: a dedicated multi-level FDL is configured for each wavelength, and the number of FDLs required is large, resulting in a large scale of core nodes, and a single-plane switch matrix structure, that is, a PM×PM scale switch matrix structure , the switching matrix is ​​composed of switches, if represented by 1×2 basic optical switches, the number of basic optical switches needs to be 2×PM(PM-1), so the scale of the core nodes is larger and it is difficult to integrate and realize; Because the cache adopts a linear connection structure, the number of FDLs is limited, and there is an upper limit of the cache time. When the cache time required by the data packet exceeds the upper limit, the data packet can only be discarded; the optical cache used is in the switching plane In front of the front, there is the problem of head-of-line blocking, that is, two burst data of the same wavelength sequentially coming from the same ingress port will arrive at different egress ports, if the previous burst data is delayed in the optical buffer The time overlap with the subsequent burst data out of the optical buffer will inevitably cause the subsequent burst data to be discarded when the switching plane is idle, and cannot be sent to the corresponding outbound port

Images