Implementation method for many-to-one data transmission

Abstract

The invention relates to an implementation method for many-to-one data transmission. According to the method, data output by each node on the network is received in parallel in a point-to-point manner based on the Ethernet for a specific transmission network with fixed transmission purpose, fixed data output frequency and fixed transmission bandwidth requirement in data transmission; and many-to-one data transmission is completed in a mode of re-framing and packaging into a shared transmission frame according to a preset frame rate, frame length and load distribution. The data transmission realized by the mode has the characteristics of data transmission synchronization, transmission delay determination and the like. The defects of uncertain transmission delay, low many-to-one transmission efficiency and the like caused by the adoption of a traditional Ethernet switch are avoided.

Description

Technical Field

[0001] The present invention belongs to the field of data communication, and specifically provides an implementation method for one-to-many data transmission in a specific transmission network with fixed transmission purposes, fixed node data output frequencies, and fixed transmission bandwidth requirements. Background Art

[0002] Due to the widespread application of Ethernet technology, the vast majority of data transmission networks use Ethernet to complete data transmission. Therefore, Ethernet has gradually become the mainstream architecture of data transmission networks. Building a data transmission network using Ethernet has the advantages of low cost, easy implementation, and wide availability of related components. However, for application scenarios where multiple data output devices simultaneously output data to a system recording device, that is, a one-to-many data transmission network, the use of traditional Ethernet results in uncertain transmission delays, low efficiency, and a risk of data loss. Such application scenarios are very common, such as the specific data transmission networks involved in data recording in data acquisition systems, sensing information aggregation in intelligent equipment, and intelligence collection in command and control systems.

[0003] For the above application scenarios, currently, a star network is basically formed through an Ethernet switch to connect multiple data output nodes and a transmission target node (such as a recording node, an information processing node, an intelligence analysis node, etc.). The data output nodes directly send the data to the target node through the switch, or the data output nodes output the data in a broadcast manner, and then the target node receives this data. In such an environment, it is often necessary to face the situation where multiple output nodes send data to the target node at the same moment, which will result in transmission path competition, leading to a significant decrease in transmission efficiency. When the total data output of the underlying devices is large, data loss may even occur.

[0004] The present invention realizes parallel transmission from multiple nodes to the same node by sharing transmission frames, avoiding competition, and thus realizing efficient exchange and transmission of one-to-many data. Summary of the Invention

[0005] As shown in the appendix Figure 1As shown in the figure, the present invention adds a framing module to the switching node of the data transmission network and designs a shared transmission frame to converge the data output by each data output node in the same time period. The framing module uses the shared transmission frame to pack the data output by each data output node in the same time period into the same transmission frame, and then sends it to the target node for reception. In this way, parallel transmission from all data output nodes to the same target node is achieved, avoiding competition for the transmission path generated by many-to-one transmission, thereby improving the data exchange efficiency and avoiding the risk of data loss. At the same time, since the above operations are independent of the underlying communication, any underlying communication protocol including Ethernet can be used for the underlying communication.

[0006] The above-mentioned shared transmission frame consists of a frame header, a payload, and a frame tail. Among them, the frame header includes three items: a frame identifier, a time tag, and a frame length. The payload contains the data output by all data output nodes in the current same time period, and the frame tail is the check code of this frame. Since the outputs of all data output nodes of the transmission network targeted by the present invention have specific transmission networks with determined transmission paths, determined data transmission frequencies, and determined data output amounts per unit time, the relevant parameters of the transmission frame can be set before the specific application runs, including the space occupied by each node's output in the transmission frame, the frame length of the transmission frame, the frame rate, etc.

[0007] The determination of the frame rate of the shared transmission frame needs to meet the transmission delay index of the specific application. First, the framing time of the transmission frame cannot be greater than the transmission delay index, that is, the frame interval (transmission interval) must be less than the transmission delay index. For example, if the transmission delay index of a data acquisition system is 10 milliseconds, then the frame interval cannot exceed 10 milliseconds. Considering other additional overheads, the frame interval can be set to 5 milliseconds to ensure that the transmission delay is less than 10 milliseconds. After the transmission frame is framed, it is immediately sent out. In this way, the frame rate of the transmission frame is 200 frames per second. Since the effective bandwidth of Ethernet is generally about 600M, the frame length of the shared transmission frame of this system can be set to 3Mb.

[0008] The following is an example to illustrate the specific implementation steps of the present invention.

[0009] Taking a simplified data acquisition system consisting of 4 data acquisition devices, an eight-port switch, and a recording device as an example, all data acquisitions use 16-bit analog-to-digital conversion. Among them, the acquisition parameters of data acquisition device 1 are 40, the sampling rate is 10KHz, the acquisition parameters of data acquisition device 2 are 2, the sampling rate is 1MHz, the acquisition parameters of data acquisition device 3 are 130, the sampling rate is 1Hz, and the acquisition parameters of data acquisition device 4 are all 100, and the sampling rate is 50Hz. The switch uses the frame assembly module described in the present invention and introduces a shared transmission frame. Assuming that the system transmission delay index is not greater than 20 milliseconds, The specific implementation steps are as follows: 1) First, determine the relevant parameters of the shared transmission frame. According to the transmission delay index, the frame rate can be selected as 100 frames / second. The content to be transmitted in each frame includes 10 samplings of 40 parameters of acquisition device 1, 1000 samplings of 2 parameters of acquisition device 2, 1 sampling of 130 parameters of acquisition device 3, and 1 sampling of 100 parameters of acquisition device 4. That is, the data length of the shared transmission frame is 10520 (bytes). Considering additional overhead such as frame headers and tails, the frame length is set to 10600 bytes; 2) Improve the switch of the system using the present invention, and add a frame assembly module to realize the frame assembly and frame transmission of the shared transmission frame. The frame assembly module includes a setting module, 8 1Gb caches (1 dedicated cache for each port), a data packing module, and a data sending module; 3) Before the data acquisition system performs data acquisition, set the length, frame rate of the shared transmission frame, and the position and length of the data sent by each acquisition device within a single transmission cycle in the transmission frame on the switch; 4) The entire system starts to periodically acquire data, and the acquired data is immediately output to the network switching node through the Ethernet interface; 5) The network switching node simultaneously receives the data sent by each acquisition device and temporarily stores the received data in the corresponding cache; 6) The frame assembly module on the network switching node sequentially reads the received data from the corresponding caches of each acquisition device on the switching node according to the allocated length and writes the data into the transmission frame; 7) The frame assembly module of the network switching node packs the transmission frame according to the Ethernet communication protocol and sends it to the connection port of the recording device; 8) The recording device receives the shared transmission frame, unpacks it, and records it in the traditional Ethernet manner.

[0010] In the above implementation process, the data transmission of each node is parallel and there is no competition for the transmission path, so the entire data transmission is synchronous. Moreover, without considering the line overhead, the data transmission delay is determined. Description of the Drawings

[0011] Appendix Figure 1 It is a schematic diagram for realizing multi-to-one data transmission.

Claims

1. A method for implementing multi-to-one data transmission, characterized in that: Taking advantage of the characteristics of a specific data transmission network, such as fixed data transmission bandwidth requirements, fixed data output frequency for each node, and fixed transmission destinations, the data output by all network nodes in the system is transmitted in parallel to the target node by adopting the shared transmission frame method.

2. The shared transmission frame method according to claim 1, wherein The data output by all data output nodes on the transmission network in the same time period is packed into a shared transmission frame and sent to the target node, thus realizing many-to-one parallel transmission.

3. The shared transmission frame according to claim 1, wherein The frame rate, frame length of the shared transmission frame and the load distribution for the data output of each node are set in advance to ensure the data transmission delay index required by specific applications.

Images