A method for integrating communication between analog signal intercom and digital signal intercom

Abstract

The invention discloses a method for integrated communication of an analog signal intercom and a digital signal intercom, and relates to the field of integrated communication. First, a integrated communication gateway is constructed, and then the analog signal intercom is connected to the integrated communication gateway, and then the sound signal is input detected and processed, and then the sound signal is encoded, decoded and integrated communication is performed on the sound signal to realize voice intercom from the analog signal intercom to the digital signal intercom, and then the voice intercom from the digital signal intercom to the analog signal intercom, and the function is expanded. The invention realizes signal conversion and network communication by constructing a integrated communication gateway, and at the same time realizes communication protocol fusion and intercommunication between the digital signal intercom and the analog signal intercom, and also realizes integrated communication of newly added equipment through a function expansion module, and has high practicality and scalability.

Description

Technical Field

[0001] The present invention relates to the field of information interaction analysis, and more specifically to a method for integrated communication of an analog signal intercom and a digital signal intercom. Background Art

[0002] Traditional analog signal intercoms, also called private network intercoms, require the construction of base stations and the use of self-built signal channels (such as 350M, 800M, etc.) for communication. Due to limited distance, long-distance communication requires the construction of a large number of relay stations, which is costly. With the maturity of 5G technology and the popularization of smart terminals, more and more manufacturers, including the three major operators, are deploying public network intercoms (digital signal intercoms). Public network intercoms use VoIP technology to implement half-duplex voice intercom services based on digital network communication. With the help of the operator's existing mobile data network, it has a wide coverage range, is not restricted by distance, and has low cost. However, in remote areas such as mountainous areas, the operator network cannot cover and voice intercom cannot be realized. Therefore, the fusion communication of analog signal intercoms and digital signal intercoms came into being.

[0003] However, the existing integrated communication of analog signal walkie-talkies and digital signal walkie-talkies has poor interoperability due to the different communication protocols and encoding methods used by analog signal walkie-talkies and digital signal walkie-talkies. When analog signal walkie-talkies and digital signal walkie-talkies transmit through the same channel, signal interference will occur, affecting the communication quality; analog signal walkie-talkies have no security guarantees and are easily eavesdropped and attacked; due to the different functions and characteristics of analog signal walkie-talkies and digital signal walkie-talkies themselves, integrated communication can often only achieve basic voice transmission.

[0004] Therefore, the present invention discloses a method for integrated communication between analog signal walkie-talkies and digital signal walkie-talkies, which integrates the communication between analog signal walkie-talkies and digital signal walkie-talkies, and performs communication protocol integration and intercommunication between digital signal walkie-talkies and analog signal walkie-talkies. It also realizes integrated communication of newly added equipment through a function expansion module, and has high practicality and scalability. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention discloses a method for integrated communication of an analog signal walkie-talkie and a digital signal walkie-talkie, which can realize the integrated communication of the analog signal walkie-talkie and the digital signal walkie-talkie; the analog signal walkie-talkie and the digital signal walkie-talkie are integrated into the same communication system through an aviation head interface, which solves the compatibility problem of two different signals, and adopts a standardized interface to facilitate the expansion of other devices; the Android operating system is used as a processing platform to realize real-time detection and processing of sound signals, and the real-time performance and compatibility of the system with different devices are improved; the error correction code LDPC is used to encode and decode the sound signal, which improves the reliability of signal transmission and reduces the possibility of signal loss and interference during transmission; the MESDM integrated communication model is used to forward the signal after noise reduction processing to the digital signal walkie-talkie, which realizes the fusion of communication protocols, reduces communication costs and improves communication quality; the digital sound signal is forwarded by connecting the aviation head interface through an audio signal line, which realizes the function of forwarding the sound of the digital walkie-talkie to the analog walkie-talkie, and solves the transmission problem between different signal devices; the plug-in architecture is used to divide the device into independent plug-ins, which improves the flexibility of function expansion, and the integrated communication of the newly added equipment becomes easier; the automation and intelligence levels are high.

[0006] The present invention adopts the following technical solutions:

[0007] A method for integrated communication between an analog signal intercom and a digital signal intercom, comprising the following steps:

[0008] Step 1: Build a converged communication gateway, which is used for network integration, signal conversion and network communication of digital signal intercom and analog signal intercom. The converged communication gateway includes a network integration module, an Android operating system, an IP communication network, a signal encoding and decoding module and a function extension module. The network integration module, the Android operating system, the IP communication network, the signal encoding and decoding module and the function extension module are bidirectionally connected.

[0009] Step 2: Connect the analog signal intercom to the fusion communication gateway. The network integration module integrates the analog signal intercom and the digital signal intercom into the same communication system through the aviation head interface. The audio interface of the analog signal intercom is connected to the aviation head interface through an audio signal line, and the sound signal is input to the fusion communication gateway for processing.

[0010] Step 3: Real-time input detection and processing of sound signals, the fusion communication gateway performs real-time detection and processing of sound signals through the Android operating system;

[0011] Step 4: Encoding and decoding of sound signals and fusion communication. The signal encoding and decoding module performs encoding and decoding operations on the sound signals through the error correction code LDPC to improve the reliability of signal transmission. The IP communication network is connected to the server through the operator's SIM card. The IP communication network forwards the noise-reduced signal to the digital signal walkie-talkie through the MESDM fusion communication model to achieve communication protocol fusion, and encrypts the signal communication through the IPsec encryption protocol;

[0012] Step 5: The digital signal intercom conducts voice intercom to the analog signal intercom. When the digital signal intercom receives the voice intercom, the server forwards the digital sound signal to the fusion communication gateway through the IP communication network. The fusion communication gateway is connected to the analog PTT button switch through the aviation head interface to simulate the PTT button operation. The aviation head interface uses the digital sound signal as the analog intercom microphone input signal through the audio signal line to realize the forwarding of the sound of the digital intercom to the analog intercom.

[0013] Step six, function expansion, the converged communication gateway performs converged communication of newly added devices through the function expansion module, the function expansion module integrates different types of devices through standardized interfaces to achieve function expansion, and adopts a plug-in architecture to divide the devices into independent plug-ins to improve the flexibility of function expansion.

[0014] As a further technical solution of the present invention, the aviation head interface shields physical obstacles by setting a shielding cover inside or outside the device, and uses filters and isolation transformers to suppress radio frequency interference and electromagnetic noise to reduce mutual interference between devices and the impact on the surrounding environment. The aviation head interface automatically switches the signal access path through the signal detection circuit and the switch control circuit to achieve the fusion access of analog sound signals and digital sound signals. The signal detection circuit detects the type of access sound signal through the frequency range, amplitude, level or waveform characteristics of the sound signal, and the switch control circuit realizes the switching of the input path through the field effect tube FET or the transistor MOS tube.

[0015] As a further technical solution of the present invention, the audio signal line uses multiple strands of thin copper wires as signal transmission conductors and is coated with polytetrafluoroethylene for outer insulation to reduce the impact of interference in the external environment on signal transmission.

[0016] As a further technical solution of the present invention, the Android operating system includes a sound signal detection module, a format conversion module, an anti-interference module, a quality management module and a signal transmission module. The sound signal detection module performs input detection and recognition on the sound signal through a CNN-LSTM neural network recognition model, the format conversion module converts the analog sound signal into a digital sound signal through an audio format converter, the anti-interference module performs noise reduction processing through a frequency selective filter, the quality management module performs quality evaluation on the sound signal after noise reduction processing through a quality management DQM framework, and the signal transmission module forwards the processed sound signal to the server through a 5G signal antenna. The output end of the sound signal detection module is connected to the input end of the format conversion module, the output end of the format conversion module is connected to the input end of the anti-interference module, the anti-interference module is bidirectionally connected to the quality management module, and the output end of the quality management module is connected to the input end of the signal transmission module.

[0017] As a further technical solution of the present invention, the CNN-LSTM neural network recognition model realizes sound signal input detection and recognition by combining a convolutional neural network CNN and a long short-term memory module LSTM. The convolution layer of the convolutional neural network performs a convolution operation on the input sound signal with a convolution kernel, which is expressed as:

[0018]

[0019] In formula (1), Z o represents the convolutional layer output of the convolutional neural network, o represents the output, x i Represents the input of the convolution kernel of the recognition model, i represents the input, k i represents the input sound signal, ∑ i x i *k i Represents the convolution set of sound signal input, * represents the convolution operation, represents the bias of the convolutional layer of the recognition model. Formula (1) can be used to integrate the input sound signal into the convolutional neural network for convolution operation and extract the initial features of the input sound signal.

[0020] After the convolution operation is completed, the convolutional neural network deeply mines the features of the input sound signal through batch normalization and activation function, which is expressed as:

[0021]

[0022] In formula (1), y represents the batch normalization result of the convolutional neural network. Indicates batch normalization of the convolution results, relu indicates the activation function of the convolutional neural network, Indicates activation of the convolution result, and deeply mines the characteristics of the input sound signal through formula (2);

[0023] After the convolutional neural network extracts the feature output, the long short-term memory module LSTM is integrated into the convolutional neural network. The long short-term memory module LSTM determines whether the input sound signal meets the requirements through the memory gate and filters it. The memory gate function model of the long short-term memory module LSTM is expressed as:

[0024]

[0025] In formula (3), tanh represents the activation function of the memory gate, W i , W c represents the weight matrix in the long short-term memory module LSTM, b t , b c represents the offset of the long short-term memory module LSTM, (yh) 2 ,h indicates that it does not conform to the characteristics of the input sound signal. h represents the characteristics of the input sound signal, A represents the discarded input sound signal, and C represents the retained input sound signal.

[0026] As a further technical solution of the present invention, the quality management DQM framework obtains the sound signal after noise reduction processing through the native interface API and performs quality detection on the sound signal through the rule engine. The rule engine evaluates the clarity, stability and noise suppression effect of the sound signal according to preset rules and thresholds. When there are errors in the evaluation indicators of the sound signal, the quality management DQM framework adopts a quality monitoring mechanism and an error handling mechanism to perform secondary processing on the sound signal.

[0027] As a further technical solution of the present invention, the operator SIM card is used as a connection medium to connect the server and the IP communication network. The converged communication gateway is used as a vehicle-mounted device and a temporary command center relay station through the mobile SIM card slot, thereby improving the flexibility of use of the converged communication gateway.

[0028] As a further technical solution of the present invention, the MESDM fusion communication model performs frequency modulation operation according to the signal type during the transmission of the sound signal to integrate the communication protocol. The frequency modulation operation expression is:

[0029]

[0030] In formula (4), E represents the matching probability between the sound signal and the communication protocol, m represents the amount of signal transmitted in the IP communication network, r represents the channel radius, s represents the amount of analog sound signal transmitted in the channel, q represents the degree of change of the communication protocol, and N aRepresents a random frequency hopping sequence; in the process of matching the sound signal and the communication protocol, the bit error rate of different channels is different. In order to reduce the bit error rate, the average value is calculated:

[0031]

[0032] In formula (5), M 0 represents the initial channel state, P represents the mean bit error rate of the communication transmission channel, L represents the length of the established transmission channel, and P G Indicates the bit error rate before frequency modulation, P B It represents the bit error rate after the improvement of the ME-SDM communication algorithm model. The bit error data of the transmission channel is suppressed by the ME-SDM communication algorithm model, and the transmission channel is matched with the sound signal and the communication protocol layer by layer in the form of data transmission. The matching transmission amount is:

[0033] R[M 0 ]=1-P B [T(1-aL)]+P B [T(1-a(L+L f ))] (6)

[0034] In formula (6), R represents the transmission amount of the sound signal, T represents the improvement of the ME-SDM communication algorithm model after the communication protocol is integrated, a represents the change amount after the communication protocol is integrated, and L f Indicates the channel widening amount after the communication protocol is integrated.

[0035] Positive beneficial effects:

[0036] The invention discloses a method for integrated communication of an analog signal walkie-talkie and a digital signal walkie-talkie, which can realize the integrated communication of the analog signal walkie-talkie and the digital signal walkie-talkie; the analog signal walkie-talkie and the digital signal walkie-talkie are integrated into the same communication system through an aviation head interface, so as to solve the compatibility problem of two different signals, and adopt a standardized interface to facilitate the expansion of other devices; the Android operating system is used as a processing platform to realize real-time detection and processing of sound signals, and the real-time performance and compatibility with different devices of the system are improved; the error correction code LDPC is used to encode and decode the sound signal, so as to improve the reliability of signal transmission and reduce the possibility of signal loss and interference during transmission; the MESDM integrated communication model is used to forward the signal after noise reduction processing to the digital signal walkie-talkie, so as to realize the communication protocol integration, reduce the communication cost and improve the communication quality; the digital sound signal is forwarded by connecting the aviation head interface through an audio signal line, so as to realize the function of forwarding the sound of the digital walkie-talkie to the analog walkie-talkie, and solve the transmission problem between different signal devices; the plug-in architecture is used to divide the device into independent plug-ins, so as to improve the flexibility of function expansion, and make the integrated communication of the newly added device easier; the automation and intelligence degree are high. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the overall process of a method for integrating communication between an analog signal intercom and a digital signal intercom according to the present invention;

[0038] Figure 2 This is a module architecture diagram of a method for integrating communication between an analog signal intercom and a digital signal intercom in the present invention;

[0039] Figure 3 A model architecture diagram of an Android operating system in a method for integrating communication between an analog signal intercom and a digital signal intercom according to the present invention;

[0040] Figure 4 A model architecture diagram of a CNN-LSTM neural network recognition model in a method for integrating communication between an analog signal intercom and a digital signal intercom according to the present invention;

[0041] Figure 5 The internal architecture diagram of the aviation head interface in the method for integrated communication of an analog signal walkie-talkie and a digital signal walkie-talkie of the present invention. DETAILED DESCRIPTION

[0042] 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.

[0043] A method for integrated communication between an analog signal intercom and a digital signal intercom, comprising the following steps:

[0044] Step 1: Build a converged communication gateway, which is used for network integration, signal conversion and network communication of digital signal intercom and analog signal intercom. The converged communication gateway includes a network integration module, an Android operating system, an IP communication network, a signal encoding and decoding module and a function extension module. The network integration module, the Android operating system, the IP communication network, the signal encoding and decoding module and the function extension module are bidirectionally connected.

[0045] Step 2: Connect the analog signal intercom to the fusion communication gateway. The network integration module integrates the analog signal intercom and the digital signal intercom into the same communication system through the aviation head interface. The audio interface of the analog signal intercom is connected to the aviation head interface through an audio signal line, and the sound signal is input to the fusion communication gateway for processing.

[0046] Step 3: Real-time input detection and processing of sound signals, the fusion communication gateway performs real-time detection and processing of sound signals through the Android operating system;

[0047] Step 4: Encoding and decoding of sound signals and fusion communication. The signal encoding and decoding module performs encoding and decoding operations on the sound signals through the error correction code LDPC to improve the reliability of signal transmission. The IP communication network is connected to the server through the operator's SIM card. The IP communication network forwards the noise-reduced signal to the digital signal walkie-talkie through the MESDM fusion communication model to achieve communication protocol fusion, and encrypts the signal communication through the IPsec encryption protocol;

[0048] Step 5: The digital signal intercom conducts voice intercom to the analog signal intercom. When the digital signal intercom receives the voice intercom, the server forwards the digital sound signal to the fusion communication gateway through the IP communication network. The fusion communication gateway is connected to the analog PTT button switch through the aviation head interface to simulate the PTT button operation. The aviation head interface uses the digital sound signal as the analog intercom microphone input signal through the audio signal line to realize the forwarding of the sound of the digital intercom to the analog intercom.

[0049] Step six, function expansion, the converged communication gateway performs converged communication of newly added devices through the function expansion module, the function expansion module integrates different types of devices through standardized interfaces to achieve function expansion, and adopts a plug-in architecture to divide the devices into independent plug-ins to improve the flexibility of function expansion.

[0050] In a specific embodiment, when an analog signal intercom is talking to a digital signal intercom, a customized audio signal line is used to connect the analog intercom audio interface to the gateway aviation head; the analog intercom receives the voice intercom of other analog intercoms, and the sound is forwarded to the gateway through the customized audio line. The gateway detects in real time whether there is a sound signal input. When sound input is detected, the sound is converted into a digital signal and noise reduction is performed, forwarded to the server, and then forwarded to each digital signal intercom by the server.

[0051] When a digital signal intercom conducts voice intercom to an analog signal intercom, a customized audio signal line is also used to connect the analog intercom audio interface to the gateway aviation head; when other digital signal intercoms initiate voice intercom, the server also forwards the sound signal to the gateway. The gateway forwards the sound signal to the analog intercom through the customized audio line, and simulates the PTT button at the same time, using the sound signal as the analog intercom microphone input signal, thus forwarding the sound of the digital intercom to the analog intercom.

[0052] In the above embodiment, the aviation head interface shields physical obstacles by setting a shielding cover inside or outside the device, and uses filters and isolation transformers to suppress radio frequency interference and electromagnetic noise to reduce mutual interference between devices and the impact on the surrounding environment. The aviation head interface automatically switches the signal access path through the signal detection circuit and the switch control circuit to achieve the fusion access of analog sound signals and digital sound signals. The signal detection circuit detects the type of access sound signal through the frequency range, amplitude, level or waveform characteristics of the sound signal, and the switch control circuit realizes the switching of the input path through the field effect tube FET or the transistor MOS tube.

[0053] In a specific embodiment, the aviation head interface is a communication interface used in a device, and its main feature is that it can support both analog signal and digital signal input and output. In order to realize analog signal input, an analog signal input port can be added to the interface design, and the analog signal can be converted into a digital signal for processing with an ADC (analog-to-digital converter). The ADC can divide the continuously changing analog value into a limited number of discrete values ​​according to a certain accuracy, and encode it into a binary number for easy storage and processing. In order to support digital signal input, a serial or parallel transmission method can be adopted in the interface design to transmit the digital signal to the target device through the data bus. In addition, the impact of quantization error on system performance also needs to be considered when performing digital signal processing. In order to facilitate user use, an automatic switching function is usually added to the aviation head interface design. When an analog signal is detected, the analog input path is automatically selected; when a digital signal is detected, the digital input path is automatically selected. Appropriate anti-interference measures need to be taken in the interface design to ensure that the aviation head interface can work stably. For example, a shielding layer is added to the interface line, a differential transmission method is adopted, etc.

[0054] In the above embodiment, the audio signal line uses multiple strands of thin copper wires as signal transmission conductors and is coated with polytetrafluoroethylene for outer insulation to reduce the impact of interference in the external environment on signal transmission.

[0055] In a specific embodiment, a customized audio signal line is a signal transmission line specifically used to connect the analog walkie-talkie audio interface and the gateway aviation head. Its technical essence is that when transmitting audio signals, by adopting high-quality materials, reasonable wiring and other technical means, signal interference and distortion are minimized, thereby ensuring high-fidelity transmission of audio signals. Specifically, customized audio signal lines usually use multiple strands of fine copper wire as conductors and are coated with special insulating materials. At the same time, in terms of wiring, different arrangements will be made according to actual conditions to minimize electromagnetic interference and crosstalk. In addition, high-quality materials will also be used in the connector, and professional processing and testing will be carried out to ensure a stable and reliable connection between the connector and the analog walkie-talkie audio interface and the gateway aviation head.

[0056] In short, customized audio signal cables improve the quality and reliability of audio signal transmission through optimized design and fine manufacturing, and provide necessary support for the interconnection between analog walkie-talkies and converged communication gateways.

[0057] In the above embodiment, the Android operating system includes a sound signal detection module, a format conversion module, an anti-interference module, a quality management module and a signal transmission module. The sound signal detection module performs input detection and identification on the sound signal through the CNN-LSTM neural network recognition model, the format conversion module converts the analog sound signal into a digital sound signal through an audio format converter, the anti-interference module performs noise reduction processing through a frequency selective filter, the quality management module performs quality evaluation on the sound signal after noise reduction processing through the quality management DQM framework, and the signal transmission module forwards the processed sound signal to the server through a 5G signal antenna. The output end of the sound signal detection module is connected to the input end of the format conversion module, the output end of the format conversion module is connected to the input end of the anti-interference module, the anti-interference module is bidirectionally connected to the quality management module, and the output end of the quality management module is connected to the input end of the signal transmission module.

[0058] In a specific embodiment, in this solution, the Android operating system includes a sound signal detection module, a format conversion module, an anti-interference module, a quality management module and a signal transmission module. These modules together constitute a complete voice signal processing system for converting analog sound signals into digital sound signals, and transmitting them to the server after noise reduction and quality evaluation.

[0059] Specifically, the sound signal detection module performs input detection and recognition of the sound signal through the CNN-LSTM neural network recognition model. The neural network can learn different types of voice signal features and can automatically extract key information, thereby realizing accurate recognition of the input voice signal. The format conversion module converts the analog sound signal into a digital sound signal through an audio format converter. Since different types of devices may use different voice coding formats, it is necessary to convert them into a unified digital coding format before transmission to ensure the compatibility and stability of data transmission. The interference module performs noise reduction processing through a frequency selective filter. In practical applications, voice signals are often affected by various interference factors such as environmental noise, so it is necessary to use an effective noise reduction algorithm to process them to improve the quality and reliability of voice signals. The quality management module evaluates the quality of the sound signal after noise reduction processing through the quality management DQM framework. The framework can automatically detect various problems in the voice signal and provide corresponding repair measures to ensure the integrity and accuracy of the transmitted data. The signal transmission module forwards the processed sound signal to the server through the 5G signal antenna. 5G technology has the advantages of high speed, low latency, and large capacity, which can meet the needs of real-time transmission of voice data. At the same time, the module also supports wired network access to meet data transmission needs in different scenarios.

[0060] In summary, this solution adopts a series of advanced technical means and algorithms to realize voice communication between analog walkie-talkies and digital walkie-talkies, and uniformly manages and controls various functional modules through the Android operating system, thus realizing an efficient, stable and reliable voice communication system.

[0061] In the above embodiment, the CNN-LSTM neural network recognition model realizes sound signal input detection and recognition by combining a convolutional neural network CNN and a long short-term memory module LSTM. The convolution layer of the convolutional neural network performs a convolution operation on the input sound signal with a convolution kernel, which is expressed as:

[0062]

[0063] In formula (1), Z o represents the convolutional layer output of the convolutional neural network, o represents the output, x i Represents the input of the convolution kernel of the recognition model, i represents the input, k i represents the input sound signal, ∑ i x i *k i Represents the convolution set of sound signal input, * represents the convolution operation, represents the bias of the convolutional layer of the recognition model. Formula (1) can be used to integrate the input sound signal into the convolutional neural network for convolution operation and extract the initial features of the input sound signal.

[0064] After the convolution operation is completed, the convolutional neural network deeply mines the features of the input sound signal through batch normalization and activation function, which is expressed as:

[0065]

[0066] In formula (1), y represents the batch normalization result of the convolutional neural network. Indicates batch normalization of the convolution results, relu indicates the activation function of the convolutional neural network, Indicates activation of the convolution result, and deeply mines the characteristics of the input sound signal through formula (2);

[0067] After the convolutional neural network extracts the feature output, the long short-term memory module LSTM is integrated into the convolutional neural network. The long short-term memory module LSTM determines whether the input sound signal meets the requirements through the memory gate and filters it. The memory gate function model of the long short-term memory module LSTM is expressed as:

[0068]

[0069] In formula (3), tanh represents the activation function of the memory gate, W i , W c represents the weight matrix in the long short-term memory module LSTM, b t , b c represents the offset of the long short-term memory module LSTM, (yh) 2 ,h indicates that it does not conform to the characteristics of the input sound signal. h represents the characteristics of the input sound signal, A represents the discarded input sound signal, and C represents the retained input sound signal.

[0070] In a specific embodiment, the CNN-LSTM neural network recognition model is a sound signal processing algorithm based on deep learning, which realizes the detection and recognition of sound signal input by combining the convolutional neural network CNN and the long short-term memory module LSTM.

[0071] Specifically, the algorithm first uses a convolutional neural network (CNN) to extract features from the input sound signal. The convolutional neural network (CNN) can automatically learn and extract useful features from the signal, such as frequency, energy, time domain, and other information. Through multi-layer convolution and pooling operations, the data dimension can be gradually reduced, and more abstract and advanced feature representations can be extracted. Next, the feature sequence extracted by CNN is input into the long short-term memory module (LSTM) for time series modeling and classification. The long short-term memory module (LSTM) can effectively solve the defects of traditional RNN (recurrent neural network) in the problem of gradient disappearance or explosion, and can process variable-length sequence data. By training the LSTM network parameters, the classification and recognition of different types of sound signals can be achieved.

[0072] In summary, the CNN-LSTM neural network recognition model utilizes the latest advances in deep learning technology and has broad application prospects in the field of sound signal processing. The hardware environment in which the CNN-LSTM neural network recognition model works mainly includes the following:

[0073] CPU: Used to perform computing tasks of neural networks, including operations such as forward propagation and back propagation.

[0074] GPU: Used to accelerate the computing tasks of neural networks, especially in large-scale deep learning models, GPU can significantly increase the computing speed.

[0075] Memory: used to store neural network parameters and intermediate results, as well as input and output data.

[0076] Storage: used to store neural network model files and training data.

[0077] Network interface: used to connect other hardware devices, such as sensors, cameras, etc., to obtain input data.

[0078] Display: used to display the running status and results of the neural network.

[0079] Power supply: used to provide power to hardware devices.

[0080] In order to verify the usability and effectiveness of the algorithm, experimental tests were carried out in the experimental environment. The operating system of the experimental computer is Windows 10. The hardware configuration in the test environment is shown in Table 1.

[0081] Table 1 Hardware configuration parameters

[0082]

[0083] Before the experiment, the sample data set was preprocessed to correct the erroneous data and fill the missing data in the data set, and the attribute data was taken out when the model was trained. The network model structure parameters are shown in Table 2.

[0084] Table 2 Network model structure parameters

[0085]

[0086] In order to better compare the accuracy of different algorithm models and avoid the impact of different data sets on model accuracy, the sample data sets were randomly arranged, and the RF model and FCN model were used for comparative testing. The training set was set to contain 5,000 sound signal samples, and the training time was set to 10 minutes.

[0087] When conducting the sound signal detection experiment, the test set was divided into 5 groups. The classification accuracy of the model under different data scale training is shown in Table 3.

[0088] Table 3 Classification accuracy of the model

[0089]

[0090] Comparative test results show that the research system model has the highest data classification accuracy under different data volumes, and can avoid the impact of data randomness on detection accuracy. With the increase in the amount of training data, the monitoring accuracy growth rate is the fastest among all models. The RF model has a minimum accuracy of 83.5% in the third set of data, and a maximum accuracy of 97.2% in the fourth set of data. The overall accuracy of the FCN model does not exceed 95%, with a minimum accuracy of 88.2% and a maximum accuracy of 94.9% in the fourth set of data.

[0091] In the above embodiment, the quality management DQM framework obtains the sound signal after noise reduction processing through the native interface API and performs quality detection on the sound signal through the rule engine. The rule engine evaluates the clarity, stability and noise suppression effect of the sound signal according to preset rules and thresholds. When there are errors in the evaluation indicators of the sound signal, the quality management DQM framework adopts quality monitoring mechanism and error handling mechanism to perform secondary processing on the sound signal.

[0092] In a specific embodiment, the DQM framework detects the quality of the sound signal after noise reduction processing through the native interface API and the rule engine. The specific process is: the DQM framework obtains the sound signal after noise reduction processing through the native interface API; the DQM framework analyzes the obtained sound signal, including the frequency, amplitude, time domain and other characteristics of the sound; the DQM framework performs quality detection on the sound signal through the rule engine, and determines whether the quality of the sound signal meets the requirements according to preset rules and thresholds; the DQM framework evaluates the quality of the sound signal according to the judgment result of the rule engine, including the clarity, stability, noise suppression effect and other aspects of the sound; the DQM framework feeds back the quality evaluation result to the noise reduction processing module, and makes adjustments according to the evaluation result to improve the quality of the sound signal.

[0093] Through the above process, the DQM framework can realize the quality detection and evaluation of the sound signal after noise reduction processing, improve the clarity and stability of the sound signal, and thus enhance the user experience.

[0094] In the above embodiment, the operator SIM card is used as a connection medium to connect the server and the IP communication network. The converged communication gateway uses the mobile SIM card slot as a vehicle-mounted device and a temporary command center relay station to improve the flexibility of use of the converged communication gateway.

[0095] In a specific embodiment, the operator SIM card is a smart card used for mobile communications. It is an identification card provided by the operator to verify the identity of the user and authorize the user to use the operator's network services. The SIM card stores the user's personal information, phone number, text messages, call records and other information, as well as the service information and network configuration information provided by the operator. When the user inserts the SIM card into a mobile phone or other mobile device, he can use the network services provided by the operator, such as calls, text messages, and Internet access. The types and specifications of operator SIM cards vary, and different operators and different countries and regions may use different SIM card specifications and technical standards.

[0096] In the above embodiment, the MESDM fusion communication model performs frequency modulation operation according to the signal type during the transmission of the sound signal to integrate the communication protocol. The frequency modulation operation expression is:

[0097]

[0098] In formula (4), E represents the matching probability between the sound signal and the communication protocol, m represents the amount of signal transmitted in the IP communication network, r represents the channel radius, s represents the amount of analog sound signal transmitted in the channel, q represents the degree of change of the communication protocol, and N a Represents a random frequency hopping sequence; in the process of matching the sound signal and the communication protocol, the bit error rate of different channels is different. In order to reduce the bit error rate, the average value is calculated:

[0099]

[0100] In formula (5), M 0 represents the initial channel state, P represents the mean bit error rate of the communication transmission channel, L represents the length of the established transmission channel, and P G Indicates the bit error rate before frequency modulation, P B It represents the bit error rate after the improvement of the ME-SDM communication algorithm model. The bit error data of the transmission channel is suppressed by the ME-SDM communication algorithm model, and the transmission channel is matched with the sound signal and the communication protocol layer by layer in the form of data transmission. The matching transmission amount is:

[0101] R[M 0 ]=1-P B [T(1-aL)]+P B [T(1-a(L+L f ))] (6)

[0102] In formula (6), R represents the transmission amount of the sound signal, T represents the improvement of the ME-SDM communication algorithm model after the communication protocol is integrated, a represents the change amount after the communication protocol is integrated, and L f Indicates the channel widening amount after the communication protocol is integrated.

[0103] In a specific embodiment, the IP communication network is a communication system based on digital technology, and the signal after noise reduction processing can be forwarded to the digital signal intercom through the MESDM fusion communication model to realize the fusion of communication protocols, and the signal communication is encrypted through the IPsec encryption protocol. Specifically, the MESDM (Multiple Encryption and Signaling Delay Mitigation) fusion communication model is a multiple encryption and delay compensation technology, which can effectively improve the reliability and security of the digital communication system. The model can use a variety of different encryption algorithms and protocols at the same time, such as AES, DES, RSA, etc., to enhance the security during data transmission. In addition, the model can also reduce errors and losses during data transmission by compensating for transmission delays. On this basis, the IP communication network also uses the IPsec (Internet Protocol Security) encryption protocol to encrypt and protect digital signals. The IPsec protocol is a security protocol widely used in the network layer, which can provide confidentiality, integrity and authentication services for data packets. By using the IPsec protocol to encrypt digital signals, it can effectively prevent unauthorized users or attackers from stealing or tampering with data information.

[0104] In summary, the IP communication network can achieve secure transmission of digital signals and communication protocol integration by adopting the MESDM converged communication model and IPsec encryption protocol, and has high reliability and security. The hardware environment of the MESDM converged communication model mainly includes the following:

[0105] Server: used to store and process large amounts of data, including user information, communication records, network topology, etc.

[0106] Router: used to achieve network connection and data transmission, including data packet forwarding, routing selection, congestion control, etc.

[0107] Switch: Used to realize data exchange and forwarding within the LAN, including filtering, forwarding, and broadcasting of data packets.

[0108] Network card: used to connect the computer and the network to realize data input and output.

[0109] Memory: used to store data and programs, including operating systems, applications, databases, etc.

[0110] Display: used to display the computer's operating status and results.

[0111] Power supply: used to provide power to hardware devices.

[0112] Sensor: used to obtain environmental information, such as temperature, humidity, light, etc.

[0113] Walkie-talkie mobile device: used to achieve mobile communication and data transmission.

[0114] The laboratory configuration uses i7 series computers, mechanical hard disks, 64+128GB of memory, and a strict experimental environment to ensure the success rate of operation. The on-site experimental environment is set up, using 5G wireless network communication, the computer operation speed reaches 2.5 billion times, and the algorithm program operation error is <2.0%. Experiments are conducted in this environment. The algorithm model and the comparative algorithm models A and B are used for experiments. The experimental contents are the total amount of successful communication signals within 20 minutes and the speed and success rate of fusion communication of 8100KB signal communication. The processing accuracy of this model and the comparative models A and B are statistically analyzed, and the experimental results are recorded in Table 4.

[0115] Table 4 Communication effect statistics

[0116]

[0117] The comparison shows that the total amount of successful communication signals, communication speed and fusion communication success rate of this model are much greater than those of algorithm models A and B, which proves the practicability and effectiveness of this algorithm.

[0118] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these specific embodiments are only illustrative, and those skilled in the art may omit, replace, and change the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, merging the above method steps so as to perform substantially the same functions in substantially the same manner to achieve substantially the same results is within the scope of the present invention. Therefore, the scope of the present invention is limited only by the appended claims.

Claims

1. A method for integrating communication between an analog signal intercom and a digital signal intercom, characterized in that: The following steps are involved: Step 1: Build a converged communication gateway, which is used for network integration, signal conversion and network communication of digital signal intercom and analog signal intercom. The converged communication gateway includes a network integration module, an Android operating system, an IP communication network, a signal encoding and decoding module and a function extension module. The network integration module, the Android operating system, the IP communication network, the signal encoding and decoding module and the function extension module are bidirectionally connected. Step 2: Connect the analog signal intercom to the fusion communication gateway. The network integration module integrates the analog signal intercom and the digital signal intercom into the same communication system through the aviation head interface. The audio interface of the analog signal intercom is connected to the aviation head interface through an audio signal line, and the sound signal is input to the fusion communication gateway for processing. Step 3: Real-time input detection and processing of sound signals, the fusion communication gateway performs real-time detection and processing of sound signals through the Android operating system; Step 4: Encoding and decoding of sound signals and fusion communication. The signal encoding and decoding module performs encoding and decoding operations on the sound signals through the error correction code LDPC to improve the reliability of signal transmission. The IP communication network is connected to the server through the operator's SIM card. The IP communication network forwards the noise-reduced signal to the digital signal walkie-talkie through the ME-SDM fusion communication model to achieve communication protocol fusion, and encrypts the signal communication through the IPsec encryption protocol; Step 5: The digital signal intercom conducts voice intercom to the analog signal intercom. When the digital signal intercom receives the voice intercom, the server forwards the digital sound signal to the fusion communication gateway through the IP communication network. The fusion communication gateway is connected to the analog PTT button switch through the aviation head interface to simulate the PTT button operation. The aviation head interface uses the digital sound signal as the analog intercom microphone input signal through the audio signal line to realize the forwarding of the sound of the digital intercom to the analog intercom. Step six, function expansion, the converged communication gateway performs converged communication of newly added devices through the function expansion module, the function expansion module integrates different types of devices through standardized interfaces to achieve function expansion, and adopts a plug-in architecture to divide the devices into independent plug-ins to improve the flexibility of function expansion.

2. The method for integrated communication of analog signal intercom and digital signal intercom according to claim 1, characterized in that: The aviation head interface shields physical obstacles by setting a shielding cover inside or outside the device, and uses filters and isolation transformers to suppress radio frequency interference and electromagnetic noise to reduce mutual interference between devices and the impact on the surrounding environment. The aviation head interface automatically switches the signal access path through the signal detection circuit and the switch control circuit to achieve the fusion access of analog sound signals and digital sound signals. The signal detection circuit detects the type of access sound signal through the frequency range, amplitude, level or waveform characteristics of the sound signal, and the switch control circuit realizes the switching of the input path through the field effect tube FET or the transistor MOS tube.

3. The method for integrated communication of analog signal intercom and digital signal intercom according to claim 1, characterized in that: The audio signal line uses multiple strands of fine copper wires as signal transmission conductors and is coated with polytetrafluoroethylene for outer insulation to reduce the impact of interference in the external environment on signal transmission.

4. The method for integrated communication of analog signal intercom and digital signal intercom according to claim 1, characterized in that: The Android operating system includes a sound signal detection module, a format conversion module, an anti-interference module, a quality management module and a signal transmission module. The sound signal detection module performs input detection and recognition on the sound signal through a CNN-LSTM neural network recognition model. The format conversion module converts the analog sound signal into a digital sound signal through an audio format converter. The anti-interference module performs noise reduction processing through a frequency selective filter. The quality management module performs quality evaluation on the sound signal after noise reduction processing through a quality management DQM framework. The signal transmission module forwards the processed sound signal to the server through a 5G signal antenna. The output end of the sound signal detection module is connected to the input end of the format conversion module, the output end of the format conversion module is connected to the input end of the anti-interference module, the anti-interference module is bidirectionally connected to the quality management module, and the output end of the quality management module is connected to the input end of the signal transmission module.

5. The method for integrated communication of analog signal intercom and digital signal intercom according to claim 4, characterized in that: The CNN-LSTM neural network recognition model realizes sound signal input detection and recognition by combining a convolutional neural network CNN and a long short-term memory module LSTM. The convolution layer of the convolutional neural network performs a convolution operation on the input sound signal with the convolution kernel, which is expressed as: In formula (1), Z o represents the convolutional layer output of the convolutional neural network, o represents the output, x i Represents the input of the convolution kernel of the recognition model, i represents the input, k i represents the input sound signal, ∑ i x i ·k i Represents the convolution set of sound signal input, * represents the convolution operation, represents the bias of the convolutional layer of the recognition model. Formula (1) can be used to integrate the input sound signal into the convolutional neural network for convolution operation and extract the initial features of the input sound signal. After the convolution operation is completed, the convolutional neural network deeply mines the features of the input sound signal through batch normalization and activation function, which is expressed as: In formula (1), y represents the batch normalization result of the convolutional neural network. Indicates batch normalization of the convolution results, relu indicates the activation function of the convolutional neural network, Indicates activation of the convolution result, and deeply mines the characteristics of the input sound signal through formula (2); After the convolutional neural network extracts the feature output, the long short-term memory module LSTM is integrated into the convolutional neural network. The long short-term memory module LSTM determines whether the input sound signal meets the requirements through the memory gate and filters it. The memory gate function model of the long short-term memory module LSTM is expressed as: In formula (3), tanh represents the activation function of the memory gate, W i , W c represents the weight matrix in the long short-term memory module LSTM, b t , b c represents the offset of the long short-term memory module LSTM, (yh) 2 ,h indicates that it does not conform to the characteristics of the input sound signal. h represents the characteristics of the input sound signal, A represents the discarded input sound signal, and C represents the retained input sound signal.

6. The method for integrated communication of analog signal intercom and digital signal intercom according to claim 4, characterized in that: The quality management DQM framework obtains the sound signal after noise reduction processing through the native interface API and performs quality detection on the sound signal through the rule engine. The rule engine evaluates the clarity, stability and noise suppression effect of the sound signal according to preset rules and thresholds. When there are errors in the evaluation indicators of the sound signal, the quality management DQM framework adopts quality monitoring mechanism and error handling mechanism to perform secondary processing on the sound signal.

7. The method for integrated communication of analog signal intercom and digital signal intercom according to claim 1, characterized in that: The operator SIM card is used as a connection medium to connect the server and the IP communication network. The fusion communication gateway is used as a vehicle-mounted device and a temporary command center relay station through the mobile SIM card slot, thereby improving the flexibility of use of the fusion communication gateway.

8. The method for integrated communication of analog signal intercom and digital signal intercom according to claim 1, characterized in that: The ME-SDM fusion communication model performs frequency modulation according to the signal type during the transmission of sound signals to integrate the communication protocols. The frequency modulation operation expression is: In formula (4), E represents the matching probability between the sound signal and the communication protocol, m represents the amount of signal transmitted in the IP communication network, r represents the channel radius, s represents the amount of analog sound signal transmitted in the channel, q represents the degree of change of the communication protocol, and N a Represents a random frequency hopping sequence; in the process of matching the sound signal and the communication protocol, the bit error rate of different channels is different. In order to reduce the bit error rate, the average value is calculated: In formula (5), M0 represents the initial channel state, P represents the mean bit error rate of the communication transmission channel, L represents the length of the established transmission channel, and P G Indicates the bit error rate before frequency modulation, P B It represents the bit error rate after the improvement of the ME-SDM communication algorithm model. The bit error data of the transmission channel is suppressed by the ME-SDM communication algorithm model, and the transmission channel is matched with the sound signal and the communication protocol layer by layer in the form of data transmission. The matching transmission amount is: R[M0]=1-P B [T(1-a L )]+P B [T(1-a(L+L f ))] (6) In formula (6), R represents the transmission amount of the sound signal, T represents the improvement of the ME-SDM communication algorithm model after the communication protocol is integrated, a represents the change amount after the communication protocol is integrated, and L f Indicates the channel widening amount after the communication protocol is integrated.

Images