A multi-channel optical module and its module structure

Abstract

The present invention relates to the field of optical communication technology, and specifically, to a multi-channel optical module and its module structure. It includes a multi-channel state monitoring unit, a transmission capacity output unit, a scheduling allocation unit and a node adjustment unit. The present invention outputs the temperature in each channel through a temperature mapping model, and outputs the transmission capacity of each channel in combination with a capacity mapping model, which is conducive to load-balanced division of data streams into multiple sub-streams according to the transmission capacity of each channel, so that multiple sub-streams are allocated to multiple channels according to the transmission capacity, effectively balancing the load of each channel, improving the overall performance of the system, avoiding the situation where some channels are overloaded while other channels are idle, and improving the overall efficiency of the system; at the same time, after reaching the time window, the temperature in each channel is updated, and the channel allocation is dynamically adjusted to adapt to the ever-changing environment, and the channel status is regularly rechecked to timely discover and handle errors that occur during the transmission process, thereby improving communication reliability.

Description

Technical Field

[0001] The present invention relates to the field of optical communication technology, and in particular to a multi-channel optical module and a module structure thereof. Background Art

[0002] Optical modules are composed of optoelectronic devices, functional circuits and optical interfaces. Optoelectronic devices include transmitting and receiving parts. Simply put, the function of optical modules is photoelectric conversion. The transmitting end converts electrical signals into optical signals. After transmission through optical fibers, the receiving end converts the optical signals into electrical signals. Among them, multi-channel optical modules usually refer to optical modules that can transmit multiple data channels at the same time. They are mainly used in high-performance computing, data center interconnection and parallel computing. This type of optical module can realize the transmission of multiple independent data streams through one optical module at the same time, greatly improving the efficiency of data transmission.

[0003] However, temperature changes can affect the working performance of optical components, resulting in different transmission efficiencies in multiple channels of a multi-channel optical module, especially in high or low temperature environments. Due to differences in the positions of multiple channels and manufacturing differences, the heat distribution inside the optical module may be uneven. Even if the external ambient temperature is the same, the temperatures of various internal parts may be different, resulting in performance differences between different channels. For example, some channels may be closer to the heat source, causing local temperature increases, thereby affecting performance. When transmitting multiple data streams or larger data streams, it is not convenient to consider the impact of temperature on transmission efficiency. In view of this, we propose a multi-channel optical module and its module structure. Summary of the invention

[0004] The object of the present invention is to provide a multi-channel optical module and a module structure thereof to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, one of the objectives of the present invention is to provide a multi-channel optical module, including a multi-channel status monitoring unit, a transmission capacity output unit, a scheduling allocation unit and a node adjustment unit;

[0006] The multi-channel state monitoring unit is used to establish a temperature mapping model that associates multiple channels with temperatures, detect the ambient temperature and input it into the temperature mapping model, and output the temperature information in each channel;

[0007] The transmission capacity output unit is used to establish a capacity mapping model that associates temperature with transmission capacity, receives temperature data from the multi-channel status monitoring unit and inputs it into the capacity mapping model, and outputs the transmission capacity of each channel, wherein:

[0008] The scheduling and allocation unit is used to perform load balancing on the data stream according to the transmission capacity of each channel, split the data stream into multiple sub-streams, and allocate the sub-streams to each channel according to the transmission capacity;

[0009] The node adjustment unit is used to preset a time window. When the time window arrives, a recheck signal is transmitted to the multi-channel status monitoring unit, the temperature information of each channel is updated, and the untransmitted sub-flows are summarized and re-input into the scheduling allocation unit to reallocate the channels.

[0010] Preferably, the multi-channel state monitoring unit includes a data acquisition module and a model training module;

[0011] The data acquisition module is used to collect temperature data of each channel and the surrounding environment for multiple times, and pre-process the collected data to construct a temperature training set;

[0012] The model training module is used to receive the temperature training set provided by the data acquisition module, and construct a temperature mapping model through a linear regression model to establish a linear relationship between the temperature of each channel and the ambient temperature. The mathematical expression is:

[0013] T C =β0+β1T e +ε

[0014] Among them, T C Represents the temperature of each channel, T e Represents the ambient temperature, β0 is the intercept, which represents the temperature value of each channel when the ambient temperature is 0, β1 is the slope, ε is the error term, and the least squares method is used to solve the minimum value of the objective function, estimate the model parameters β0 and β1, obtain the ambient temperature data in real time, and input it into the trained linear regression model. The model predicts the temperature of each channel based on the trained parameters β0 and β1, and outputs the temperature value of each channel predicted by the model.

[0015] Preferably, the ambient temperature is acquired by using a digital temperature sensor, which is made of semiconductor material and measures the temperature by utilizing the property that the resistance value changes with temperature.

[0016] Preferably, the transmission capacity output unit includes a training set acquisition module and a model building module;

[0017] The training set acquisition module is used to collect the transmission capacity data of data stream A at different temperatures and construct a capacity training set;

[0018] The model building module is used to receive the capacity training set provided by the training set acquisition module, build a capacity mapping model through a neural network model to establish a mapping relationship between temperature and transmission capacity, and input real-time temperature data into the model to predict the transmission capacity of each channel.

[0019] Preferably, the scheduling allocation unit includes a data stream division module and a segment transmission module;

[0020] The data stream division module is used to quantify the transmission capacity of each channel output by the model building module, and divide the data stream into multiple sub-streams according to the ratio of the multi-channel transmission capacity;

[0021] The segmented transmission module is used to respectively receive the multiple sub-streams divided by the data stream division module, split each sub-stream into N parts, and send out a transmission signal, which is transmitted in turn through channels of corresponding proportions.

[0022] Preferably, the segmented transmission module splits each substream into N parts, which includes the following steps:

[0023] The technical parameters of each channel are collected, including the maximum bandwidth and the maximum packet rate. The maximum load capacity of each channel is calculated based on the channel parameters. The size of each sub-stream is calculated according to the sub-stream ratio and the total size of the data stream determined by the data stream division module. The transmission time of each sub-stream is calculated based on the maximum load capacity of each channel. The expression is:

[0024]

[0025] Preferably, the node adjustment unit includes a time window determination module, a communication interception module and an update triggering module;

[0026] The time window determination module is used to construct a temperature adjustment model according to the correlation between the channel use time and the temperature change node, receive the transmission time in the segmented transmission module and input the temperature adjustment model, and output the temperature change node;

[0027] The communication stop module is used to receive the transmission start time of the segment transmission module, and when the time point reaches the temperature change node, detect the number of untransmitted sub-streams, if there are untransmitted sub-streams, send a recheck signal, if there are no untransmitted sub-streams, send an end signal;

[0028] The update trigger module is used to trigger the data acquisition module to work cyclically after receiving the re-check signal until the segment transmission module sends out a transmission signal again.

[0029] The second object of the present invention is to provide a multi-channel optical module structure, comprising any one of the multi-channel optical modules described above, and further comprising a processor, wherein the processor is used to control the operation of the multi-channel status monitoring unit, the transmission capacity output unit, the scheduling allocation unit and the node adjustment unit;

[0030] The processor is connected at both ends with a multi-channel structure, and the two ends of the multi-channel structure are respectively provided with four channel interfaces for signal transmission and reception, and a communication channel is formed between two opposite channel interfaces.

[0031] Preferably, the processor adopts a microcontroller unit MCU, and a transmitting and receiving channel formed by multiple channel interfaces is driven by the same microcontroller unit MCU;

[0032] The output end of the micro control unit MCU is electrically connected to a power source POWER for providing energy.

[0033] Compared with the prior art, the present invention has the following beneficial effects: the temperature in each channel is output through the temperature mapping model, and the transmission capacity of each channel is output in combination with the capacity mapping model, which is conducive to load-balanced division of the data stream into multiple sub-streams according to the transmission capacity of each channel, so that the multiple sub-streams are distributed to multiple channels according to the transmission capacity, effectively balancing the load of each channel, improving the overall performance of the system, avoiding the situation where some channels are overloaded while other channels are idle, and improving the overall efficiency of the system;

[0034] In addition, a time window is preset, and a recheck signal is transmitted to the multi-channel status monitoring unit after the time window is reached. The temperature in each channel is updated, and the sub-streams that have not been transmitted are summarized and re-input into the scheduling allocation unit to allocate channels. The channel allocation can be dynamically adjusted according to the dynamic changes of the channel temperature to adapt to the changing environment. Regular rechecking of the channel status can timely discover and handle errors that occur during the transmission process, thereby improving the reliability of data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is the overall principle diagram of the present invention;

[0036] Figure 2 It is the arrangement principle circuit diagram of the present invention;

[0037] Figure 3 It is a flowchart of the multi-channel optical module of the present invention;

[0038] Figure 4 Schematic diagram of the L-shaped multi-channel structure of the present invention.

[0039] The meaning of each number in the figure is:

[0040] 100. Multi-channel status monitoring unit; 110. Data acquisition module; 120. Model training module;

[0041] 200, transmission capacity output unit; 210, training set acquisition module; 220, model building module;

[0042] 300, scheduling allocation unit; 310, data flow division module; 320, segment transmission module;

[0043] 400, node adjustment unit; 410, time window determination module; 420, communication stop module; 430, update trigger module;

[0044] 500, processor; 600, multi-channel structure; 610, channel interface. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Optical modules are composed of optoelectronic devices, functional circuits and optical interfaces. Optoelectronic devices include transmitting and receiving parts. Simply put, the function of optical modules is photoelectric conversion. The transmitting end converts electrical signals into optical signals. After transmitting through optical fibers, the receiving end converts the optical signals into electrical signals. Among them, multi-channel optical modules usually refer to optical modules that can transmit multiple data channels at the same time. They are mainly used in high-performance computing, data center interconnection and parallel computing. This type of optical module can realize the transmission of multiple independent data streams through one optical module at the same time, greatly improving the efficiency of data transmission. Therefore, Figure 1-Figure 4 As shown, one of the purposes of the present invention is to provide a multi-channel optical module, including a multi-channel status monitoring unit 100, a transmission capacity output unit 200, a scheduling allocation unit 300 and a node adjustment unit 400;

[0047] The multi-channel state monitoring unit 100 is used to establish a temperature mapping model that associates multiple channels with temperatures, detect the ambient temperature and input it into the temperature mapping model, and output the temperature information in each channel. Since the heat distribution inside the optical module may be uneven, even if the external ambient temperature is the same, the temperature of each internal part may be different, resulting in performance differences between different channels. The peripheral channel may be closer to the heat source, resulting in a local temperature increase, thereby affecting the performance. The temperature mapping model can effectively reflect the actual working state of the channel, so that after detecting the ambient temperature, the temperature of each channel can be accurately monitored;

[0048] Specifically, the multi-channel state monitoring unit 100 includes a data acquisition module 110 and a model training module 120;

[0049] The data acquisition module 110 is used to collect temperature data of each channel and the surrounding environment for multiple times, and pre-process the collected data to construct a temperature training set, such as removing abnormal data, data smoothing, etc. When obtaining the training set data, the temperature of each channel can be measured in the initial stage, and the surrounding temperature can be measured at the same time, such as by using sensors, thermometers and other equipment to obtain data, which provides a basis for subsequent data flow transmission judgment;

[0050] The model training module 120 is used to receive the temperature training set provided by the data acquisition module 110, and construct a temperature mapping model through a linear regression model to establish a linear relationship between the temperature of each channel and the ambient temperature. The mathematical expression is:

[0051] T C =β0+β1T e +ε

[0052] Among them, T C Represents the temperature of each channel, T e Represents the ambient temperature, β0 is the intercept, which represents the temperature value of each channel when the ambient temperature is 0, β1 is the slope, and ε is the error term. The least squares method is used to solve the minimum value of the objective function, estimate the model parameters β0 and β1, obtain the ambient temperature data in real time, and input it into the trained linear regression model. The model predicts the temperature of each channel according to the trained parameters β0 and β1, and outputs the temperature value of each channel predicted by the model. The expression of the linear regression model can be used to predict the temperature of each channel through the linear relationship between the independent variable and the dependent variable, and the parameters estimated by the least squares method. The error term ε reflects the difference between the model predicted value and the actual value. It is usually assumed to obey a normal distribution with a mean of 0. The existence of the error term is due to the simplification of the model and the randomness of the data itself. The larger the β1 value, the greater the impact of the ambient temperature change on the temperature of each channel.

[0053] Specifically, when communicating and transmitting data streams, a digital temperature sensor is used to obtain the ambient temperature. The temperature sensor is made of semiconductor materials and uses the characteristic that its resistance value changes with temperature to measure the temperature. Taking DS18B20 as an example, it contains a precision digital temperature sensor, a non-volatile memory and a single-wire interface. Its working principle is as follows:

[0054] The internal sensor senses temperature changes by measuring its own resistance value. The internal circuit converts the temperature information into a digital signal and stores the digital temperature in the internal memory. The digital temperature information stored in the internal memory is read out through a single-wire interface. This not only facilitates timely acquisition of the ambient temperature, but also provides a basis for subsequent prediction of the temperature of each channel. In addition, by providing real-time feedback of the ambient temperature, it avoids the failure to timely discover the damage of the optical module when extreme temperatures occur, which is conducive to the effect of convenient reading of early warnings.

[0055] The transmission capacity output unit 200 is used to establish a capacity mapping model that associates temperature with transmission capacity, receives temperature information from the multi-channel status monitoring unit 100 and inputs it into the capacity mapping model, and outputs the transmission capacity of each channel. The capacity mapping model can effectively reflect the actual transmission capacity of the channel, thereby achieving accurate evaluation;

[0056] Then, the transmission capability output unit 200 includes a training set acquisition module 210 and a model building module 220;

[0057] The training set acquisition module 210 is used to collect the transmission capability data of the A data stream at different temperatures and construct a capability training set;

[0058] The model building module 220 is used to receive the capacity training set provided by the training set acquisition module 210, build a capacity mapping model through a neural network model to establish a mapping relationship between temperature and transmission capacity, and input real-time temperature data into the model to predict the transmission capacity of each channel, so that there is a correlation between temperature and transmission capacity, and use data mining technology to establish a mapping relationship between temperature and transmission capacity, that is, to build a capacity mapping model, wherein the neural network model usually uses a multi-layer perceptron (MLP) model to set the number of network layers and the number of neurons: according to the complexity of the data and the accuracy requirements of the model, the number of network layers and the number of neurons in each layer are reasonably set, and the appropriate activation is selected. Functions, such as ReLU, sigmoid or tanh, are used to improve the nonlinear fitting ability of the model. Appropriate loss functions, such as mean square error (MSE) or cross entropy, are selected to evaluate the difference between the model prediction value and the true value. Appropriate optimizers, such as gradient descent algorithm, Adam algorithm, etc., are selected to update the model parameters. The training set data is input into the neural network model, and the model parameters are updated using the optimizer to minimize the loss function. The training process is continuously iterated until the model converges, which is conducive to learning the complex relationship between the data, realizing accurate prediction of the transmission capacity of each channel, and making corresponding adjustments based on the prediction results to improve network performance.

[0059] The scheduling and allocation unit 300 is used to load balance the data stream according to the transmission capacity of each channel, split the data stream into multiple sub-streams, and allocate the sub-streams to each channel according to the transmission capacity to effectively balance the load of each channel, improve the overall performance of the system, avoid the situation where some channels are overloaded while other channels are idle, and improve the overall efficiency of the system;

[0060] It is worth noting that the scheduling allocation unit 300 includes a data stream partitioning module 310 and a segment transmission module 320;

[0061] The data stream division module 310 is used to quantify the transmission capacity of each channel output by the model building module 220, and divide the data stream into multiple sub-streams according to the ratio of the multi-channel transmission capacity;

[0062] The segment transmission module 320 is used to receive the multiple sub-streams divided by the data stream division module 310, divide each sub-stream into N parts, and send out a transmission signal to transmit through channels of corresponding proportions in sequence;

[0063] For example, there are four channels a, b, c, and d for receiving / transmitting, and it is detected that the transmission capacity ratio of the four channels is 1:1.2:1.1:0.9, then the data stream is divided into four sub-streams according to the transmission capacity ratio, and the ratio of the four sub-streams is 1:1.2:1.1:0.9, and they are prepared for transmission corresponding to the four channels a, b, c, and d respectively. For example, in channel a, the sub-stream with a ratio of 1 is split into N parts through the segmented transmission module 320, so that the data stream is transmitted from the channel transmitting end to the receiving end in sequence, which is not only conducive to allocating channels according to the transmission capacity of the channels, avoiding the slow transmission of some channels affecting the overall communication efficiency, but also allows the sub-streams to be transmitted in sequence after being split, avoiding a large influx of data streams at one time causing a large channel load, thereby improving the service life.

[0064] Each channel has a maximum load capacity. Exceeding this capacity will cause the transmission speed to drop or even data loss. Therefore, the segmented transmission module 320 splits each substream into N parts, which includes the following steps:

[0065] The technical parameters of each channel are collected, including the maximum bandwidth and the maximum data packet rate, and the maximum load capacity of each channel is calculated based on the channel parameters. The size of each sub-stream is calculated according to the sub-stream ratio and the total size of the data stream determined by the data stream division module 310, and the transmission time of each sub-stream is calculated based on the maximum load capacity of each channel. The expression is:

[0066]

[0067] Among them, the maximum bandwidth is the maximum amount of data that the channel can transmit in a unit of time, usually in Mbps or Gbps; the maximum packet rate is the maximum number of packets that the channel can transmit in a unit of time, usually in packets / second; the packet size is the size of each packet, usually in bytes, and the protocol overhead is different. Different protocols will introduce additional overhead, such as the header information of the TCP protocol. If the packet size and protocol overhead are known, the maximum load capacity can be calculated more accurately. For example, assuming the packet size is 1500 bytes and the protocol overhead is 40 bytes, the maximum load capacity can be calculated as:

[0068] Maximum load capacity packets / second = maximum bandwidth Mbps*1024*1024 / (8*(packet size+protocol overhead)).

[0069] The node adjustment unit 400 is used to preset a time window. When the time window arrives, a recheck signal is transmitted to the multi-channel status monitoring unit 100, the temperature information of each channel is updated, the sub-streams that have not been transmitted are summarized and re-input into the scheduling allocation unit 300 to reallocate the channels. The channel allocation can be dynamically adjusted according to the dynamic changes of the channel temperature to adapt to the ever-changing environment, and the channel status is rechecked regularly, so that errors occurring during the transmission process can be discovered and processed in a timely manner, thereby improving the reliability of data transmission.

[0070] Furthermore, the node adjustment unit 400 includes a time window determination module 410, a communication interception module 420, and an update triggering module 430;

[0071] The time window determination module 410 is used to build a temperature adjustment model according to the correlation between the channel use time and the temperature change node, receive the transmission time in the segmented transmission module 320 and input the temperature adjustment model, and output the temperature change node. As the channel use time increases, the channel use will generate heat. This temperature change node indicates the time point when the temperature change reaches the time threshold. When establishing the temperature adjustment model, historical data is collected: data on channel use time and temperature change are collected, such as recording the temperature change of each channel under different use times. Machine learning algorithms, such as linear regression, support vector machines, etc., can be used to train the collected historical data, establish an association model between channel use time and temperature change, and use a validation set to evaluate the performance of the model to ensure that the model can accurately predict channel temperature changes. As for the time threshold, it can be set according to the actual application scenario and the tolerance for channel temperature changes. For example, if the channel temperature changes by more than 10 degrees, a re-inspection is required. The time threshold can be set to 10 minutes, indicating that within 10 minutes, if the channel temperature changes by more than 10 degrees, a re-inspection is required;

[0072] The communication stop module 420 is used to receive the transmission start time of the segment transmission module 320, and when the time point reaches the temperature change node, detect the number of untransmitted sub-streams, and send a recheck signal if there are untransmitted sub-streams, and send an end signal if there are no untransmitted sub-streams;

[0073] The update trigger module 430 is used to trigger the data acquisition module 110 to work in a loop after receiving the re-check signal until the segment transmission module 320 sends a transmission signal again. The program code can be implemented in programming languages ​​such as C / C++ and Python according to the above functions.

[0074] The second object of the present invention is to provide a multi-channel optical module structure, including any one of the multi-channel optical modules described above, such as Figure 1 As shown, it includes a processor 500, which is used to control the operation of the multi-channel status monitoring unit 100, the transmission capacity output unit 200, the scheduling allocation unit 300 and the node adjustment unit 400;

[0075] The processor 500 is connected at both ends with a multi-channel structure 600, and the two ends of the multi-channel structure 600 are respectively provided with four channel interfaces 610 for signal transmission and reception, and a communication channel is formed between two opposite channel interfaces 610;

[0076] Therefore, the optical module is a four-transmitter and four-receiver multi-channel optical module. One transmitter and one receiver can be regarded as a single transceiver integrated dual-fiber module, and the multi-channel optical module can be regarded as multiple transceiver integrated dual-fiber modules.

[0077] like Figure 2 As shown, the processor 500 adopts a microcontroller unit MCU, and a transmitting and receiving channel formed by multiple channel interfaces 610 is driven by the same microcontroller unit MCU, including arithmetic logic components, register components, and control components, etc., and has functions such as processing instructions, executing operations, controlling time, and processing data. On the chip of the microcontroller unit MCU, transistors are connected by micro cables, and conduction between transistors is achieved through the micro cables. Other performance indicators such as optical power, extinction ratio, sensitivity, etc. are controlled separately, and each channel is independent of each other. Because the multiple channels of the optical module work independently of each other, the optical module can be customized according to the application scenario, which is suitable for more communication scenarios and personalized communication needs;

[0078] Wherein: channel interface 610 includes DAC-EA-1, DAC-EA-2, DAC-Vtec-1, DAC-Vtec-2, DAC-APD-1, DAC-APD-2, DAC-EN-1, DAC-EN-2, DAC-EA-3, DAC-EA-4, DAC-Vtec-3, DAC-Vtec-4, DAC-APD-3, DAC-APD-4, DAC-EA-3 and DAC-EA-4, through and Indicates that a communication channel is formed between the two.

[0079] The output end of the microcontroller unit MCU is electrically connected to a power source POWER for providing energy, which is the basis for ensuring the execution of the function of the multi-channel optical module.

[0080] Example 2

[0081] Since the optical module in the prior art can only enter and exit from the same side, the multiple input and output ports formed by the multiple channel interfaces 610 can each be connected independently, allowing the user to choose different entry and exit methods, which can meet the requirements of entry and exit from the same side or entry from one side and exit from the other side, making installation easier and reducing wiring difficulty. The input and output ports of the optical module can be adjusted according to relevant wiring requirements, further improving wiring efficiency;

[0082] Furthermore, the multi-channel structure 600 can be manufactured as follows according to installation requirements: Figure 4 The "L" shape shown can also be made as Figure 1 The "one" shape shown is suitable for different scenarios and increases the scope of use.

[0083] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A multi-channel optical module, Features: It comprises a multi-channel status monitoring unit (100), a transmission capacity output unit (200), a scheduling allocation unit (300) and a node adjustment unit (400); The multi-channel state monitoring unit (100) is used to establish a temperature mapping model that associates multiple channels with temperatures, detect the ambient temperature and input it into the temperature mapping model, and output temperature information in each channel; The transmission capacity output unit (200) is used to establish a capacity mapping model that associates temperature with transmission capacity, receives temperature information from the multi-channel status monitoring unit (100) and inputs it into the capacity mapping model, and outputs the transmission capacity of each channel, wherein: The scheduling and allocation unit (300) is used to load balance the data stream according to the transmission capacity of each channel, split the data stream into multiple sub-streams, and allocate the sub-streams to each channel according to the transmission capacity; The node adjustment unit (400) is used to preset a time window, and when the time window arrives, transmit a recheck signal to the multi-channel state monitoring unit (100), update the temperature information of each channel, summarize the untransmitted sub-flows and re-input them into the scheduling allocation unit (300) to reallocate the channels; The transmission capability output unit (200) comprises a training set acquisition module (210) and a model building module (220); The training set acquisition module (210) is used to collect the transmission capability data of the A data stream at different temperatures and construct a capability training set; The model building module (220) is used to receive the capacity training set provided by the training set acquisition module (210), build a capacity mapping model through a neural network model to establish a mapping relationship between temperature and transmission capacity, and input real-time temperature data into the model to predict the transmission capacity of each channel.

2. The multi-channel optical module according to claim 1, Features: The multi-channel state monitoring unit (100) comprises a data acquisition module (110) and a model training module (120); The data acquisition module (110) is used to collect temperature data of each channel and the surrounding environment for multiple times, and pre-process the collected data to construct a temperature training set; The model training module (120) is used to receive the temperature training set of the data acquisition module (110), and to establish a temperature mapping model through a linear regression model to establish a linear relationship between the temperature of each channel and the ambient temperature, the mathematical expression of which is: in, represents the temperature of each channel, Indicates the ambient temperature, is the intercept, which indicates the temperature value of each channel when the ambient temperature is 0. is the slope, is the error term, and the least squares method is used to solve the minimum value of the objective function and estimate the model parameters and , obtain ambient temperature data in real time and input it into the trained linear regression model. The model uses the parameters obtained by training and Predict the temperature of each channel and output the temperature value of each channel predicted by the model.

3. The multi-channel optical module according to claim 2, Features: The ambient temperature is acquired by using a digital temperature sensor, which is made of semiconductor material and measures the temperature by using the property that the resistance value changes with the temperature.

4. The multi-channel optical module according to claim 3, Features: The scheduling allocation unit (300) comprises a data stream division module (310) and a segment transmission module (320); The data stream division module (310) is used to quantify the transmission capacity of each channel output by the model building module (220), and divide the data stream into multiple sub-streams according to the ratio of the transmission capacity of multiple channels; The segmented transmission module (320) is used to respectively receive the multiple sub-streams divided by the data stream division module (310), split each sub-stream into N parts, and send out a transmission signal to be transmitted in sequence through channels of corresponding proportions.

5. The multi-channel optical module according to claim 4, Features: The segmented transmission module (320) splits each substream into N parts, which includes the following steps: The technical parameters of each channel are collected, the technical parameters including the maximum bandwidth and the maximum data packet rate, and the maximum load capacity of each channel is calculated based on the channel parameters. The size of each sub-stream is calculated according to the sub-stream ratio and the total size of the data stream determined by the data stream division module (310), and the transmission time of each sub-stream is calculated based on the maximum load capacity of each channel. The expression is: N= 。 6. The multi-channel optical module according to claim 1, Features: The node adjustment unit (400) comprises a time window determination module (410), a communication interception module (420) and an update triggering module (430); The time window determination module (410) is used to construct a temperature adjustment model according to the correlation between the channel use time and the temperature change node, receive the transmission time in the segment transmission module (320) and input the temperature adjustment model, and output the temperature change node; The communication stop module (420) is used to receive the transmission start time of the segment transmission module (320), and when the time point reaches the temperature change node, detect the number of untransmitted sub-streams, and if there are untransmitted sub-streams, send a recheck signal, and if there are no untransmitted sub-streams, send an end signal; The update trigger module (430) is used to trigger the data acquisition module (110) to work cyclically after receiving the re-check signal until the segment transmission module (320) sends a transmission signal again.

7. A multi-channel optical module structure, applied to the multi-channel optical module according to any one of claims 1 to 6, Features: Further comprising a processor (500), the processor (500) being used to control the operation of the multi-channel status monitoring unit (100), the transmission capacity output unit (200), the scheduling allocation unit (300) and the node adjustment unit (400); The processor (500) is connected at both ends with a multi-channel structure (600), and four channel interfaces (610) for signal transmission and reception are respectively provided at both ends of the multi-channel structure (600), and a communication channel is formed between two opposite channel interfaces (610).

8. The multi-channel optical module structure according to claim 7, Features: The processor (500) adopts a micro control unit MCU, and a transmitting and receiving channel formed by a plurality of channel interfaces (610) is driven by the same micro control unit MCU; The output end of the micro control unit MCU is electrically connected to a power source POWER for providing energy.

Images