Technical method for multiplexing multiple paths of 10G Ethernet into 100G optical fiber link

Abstract

The invention provides a method for multiplexing 10 paths of Ethernet into a 100G optical fiber link. The input interface module receives 10 paths of Ethernet signals from different clients, and the multiplexing module integrates the signals into a composite signal stream according to a preset multiplexing rule. The transmission module transmits the composite signal stream to a destination through a 100G optical fiber link. And at a receiving end, the demultiplexing module separates the composite signal flow into original 10 paths of Ethernet signals according to the same multiplexing rule, and outputs the original 10 paths of Ethernet signals through the output interface module.

Description

Technical Field

[0001] The present invention relates to the field of optical communications technology, and in particular to a method for multiplexing multiple 10G Ethernet networks into a 100G fiber optic link. This method multiplexes 10 Ethernet signals onto a single 100G fiber optic link, improving fiber resource utilization and reducing network deployment costs. This method utilizes advanced multiplexing technology and efficient data packet management mechanisms to achieve efficient transmission of multiple Ethernet signals over a single fiber optic link. This method is suitable for use in scenarios such as data centers, enterprise networks, and telecommunications operators, effectively increasing network bandwidth to meet the growing demand for secure, real-time, and large-scale data transmission. Background Art

[0002] With the growing demand for data centers and high-speed network communications, the requirements for transmission bandwidth and efficiency are also increasing. Traditional Ethernet transmission methods are unable to cope with large-scale data transmission. Therefore, a simple, efficient, and reliable solution is needed to achieve high-bandwidth, low-latency data transmission. This invention aims to provide a method for multiplexing 10 Ethernet channels into 100G fiber links to meet the high bandwidth and high efficiency requirements of modern network communications. Summary of the Invention

[0003] In existing technology, relatively low-speed 10G Ethernet data from the client side is typically multiplexed into 100G optical port signals on the line side for transmission in data centers or network aggregation points using expensive and complex 100G switches or OTN optical transport network equipment. This approach presents challenges such as complex management, high installation costs, and difficult maintenance. The present invention multiplexes 10 channels of 10G Ethernet signals into 100G using a relatively simple multi-channel multiplexing technology, offering significant technical advantages and excellent economic benefits.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] The present invention proposes a method for multiplexing 10 Ethernet channels into a 100G optical fiber link. The working principle is as follows:

[0006] The input interface module receives 10 Ethernet signals from different clients. The multiplexing module combines these signals into a composite signal stream according to preset multiplexing rules. The transmission module transmits the composite signal stream to the destination via a 100G fiber link. At the receiving end, the demultiplexing module separates the composite signal stream into the original 10 Ethernet signals according to the same multiplexing rules and outputs them through the output interface module.

[0007] The present invention mainly includes the following technical features:

[0008] Sending direction:

[0009] Client signal input SFP+: The client 10GBASE-R Ethernet signal is converted into an electrical signal through the SFP+ optical module;

[0010] Clock and Data Recovery (CDR): This technology is used to recover the client-side 10G Ethernet clock and data, where the data is sent to the multi-channel block synchronization module. The clock can be selected to output the 10G Ethernet signal to the client or to drive a network timing architecture, such as the SyncE network timing architecture.

[0011] Block sync: used to recover 66B data blocks of 10GBase-R Ethernet. The processing process complies with IEEEstd 802.3-2008.

[0012] Descramble: Descramble the previously recovered block data. If the data is erroneous, insert an error indication or generate a test pattern.

[0013] Buffer queue FIFO: buffers data and synchronizes all asynchronous client data to the same clock domain for subsequent processing;

[0014] Align marker insertion: Alignment markers are inserted into the data block to identify the start and end positions of the data block;

[0015] Muxplexing: Using time-division multiplexing technology, 10 channels of customer data are multiplexed byte-by-byte onto four virtual 25.78125G links.

[0016] Data output 100G PMA: The previously multiplexed data is distributed bit-wise across four virtual links, using the 100GBASE-R PMA standard for data output, encompassing four 25.78125Gbps virtual links.

[0017] QSFP28 optical module transmission: Converts 100GBASE-R electrical signals into optical signals, which are transmitted on a 4-core optical fiber.

[0018] Receiving direction:

[0019] The processing in the receiving direction is opposite to that in the sending direction.

[0020] Optical module QSFP28 receiving: realizes the conversion of 100GBASE-R optical signal to electrical signal;

[0021] Data output 100G PMA: Combines the signals from four 25.78125Gbps virtual links into one 100G link based on the bit rate.

[0022] Demultiplexing: demultiplex 10 channels of customer data;

[0023] Alignment lock: Searches for and locks alignment markers in the data to determine the boundaries of client data blocks;

[0024] Buffer queue FIFO: buffers data and transfers data from the line-side clock domain to the client-side clock domain through the buffer;

[0025] Idle signal insertion / deletion: According to the 10GBASE-R standard, idle signals are inserted or deleted when needed, thereby mapping dead 10GBASE-R signals to the clock domain required for output.

[0026] Scramble: scrambles data according to the 10GBASE-R standard.

[0027] Customer signal output SFP+: The electrical signal is converted into an optical signal through a 10G SFP+ module and sent to the customer.

[0028] Beneficial effects

[0029] Improved transmission bandwidth: Through multi-channel multiplexing technology, 10 Ethernet signals are integrated into a single 100G optical fiber link for transmission, significantly improving transmission bandwidth.

[0030] Reduced transmission costs: Compared with the traditional transmission method of multiple independent Ethernet links, this invention reduces the use of optical fibers and equipment, and reduces transmission costs.

[0031] Enhanced system stability: Intelligent routing and traffic management technologies as well as fault detection and recovery mechanisms ensure efficient and stable data transmission. DETAILED DESCRIPTION

[0032] In a preferred embodiment, the method and system of the present invention can be applied to internal network connections in a data center. The system implemented using the present invention first connects ten 10G Ethernet signals through input interface modules. The multiplexing module uses time division multiplexing (TDM) technology to combine these signals into a single 100G signal stream. The transmission module transmits the signal stream to the data center at the other end via a single 100G optical fiber.

[0033] At the receiving end, the demultiplexing module separates the 100G signal stream into the original 10-channel Ethernet signals and connects them to the corresponding server or network equipment through the output interface module.

[0034] The multiplexing / demultiplexing part can be implemented using a high-end FPGA with 25Gbps SERDES, such as Xilinx's Vertex 7 and above series; it can also be implemented using a dedicated ASIC;

[0035] The network management part can be realized by using ordinary industrial-grade CPU with peripheral memory and flash memory. The management software based on Linux system is used to manage the system.

[0036] The power module, considering the high speed of the optical module and the total current of 3.3V voltage close to 20A, brings about large power consumption. It is necessary to use power chips and modules with higher power and conversion efficiency, such as LTM4628 chip and EPS-65S-12 power module.

[0037] Conventional methods can be used for the heat dissipation module and status indicator lights. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 In the sending direction: Schematic diagram of the principle of multiplexing 10G signals into 100G signals.

[0039] Figure 2 In the receiving direction: Schematic diagram of the principle of demultiplexing 100G signals into 10G signals.

[0040] Figure 3 Multiplex and demultiplex schematic diagrams for signals.

Claims

1. A method for multiplexing 10 Ethernet channels into a 100G optical fiber link, comprising an input interface module, a multiplexing module, a transmission module, a demultiplexing module, and an output interface module, characterized in that: The multiplexing module uses time division multiplexing technology to integrate Ethernet signals into a 100G signal stream, the transmission module transmits the signal stream through a 100G optical fiber link, and the demultiplexing module separates the received signal stream into the original 10-channel Ethernet signals.

2. The method according to claim 1, characterized in that The multiplexing module and the demultiplexing module adopt the same multiplexing rule to ensure correct separation of signals.

3. The method according to claim 1, characterized in that The method further comprises a signal quality monitoring module for monitoring signal quality and performance indicators during the transmission process.

4. The method according to claim 1, wherein An FPGA or dedicated chip with high-speed SERDES is used to achieve clock and data recovery, block synchronization, and data scrambling and descrambling.

5. The method according to claim 1, characterized in that A first-in-first-out buffer queue FIFO is used to realize data transmission across clock domains.

6. The method according to claim 1, characterized in that The method adopts the method of inserting and searching alignment marks to judge the start and end positions of the data blocks.

7. The method according to claim 1, characterized in that The method includes using a QSFP28 optical module to achieve optical-electrical signal conversion at a 100G line rate.

Images