Lightweight multi-device communication protocol intelligent adaptation conversion system for emergency rescue

Abstract

The invention relates to the technical field of wireless communication, in particular to a lightweight multi-device communication protocol intelligent adaptation conversion system for emergency rescue, which comprises a broadband signal segment acquisition step: segmenting a broadband IQ data stream into discrete signal segments based on an energy threshold and an instantaneous bandwidth parameter, extracting physical layer characteristics and packaging to generate a unified protocol data frame; a conversion scheduling control step: analyzing the metadata label to judge a source protocol type, and performing hash matching on the source protocol type and a support protocol list; a heterogeneous protocol streaming translation step: in response to known protocol determination, executing load isolation mapping or ciphertext shell reconstruction based on a protocol parameter mapping table, and generating a target protocol baseband data stream; a dynamic spectrum fingerprint routing step: in response to unknown protocol determination, establishing a transparent bridging channel, and executing digital spectrum shifting and resampling on the original IQ slice; the problems that heterogeneous protocol conversion is high in delay, large in computing power consumption and difficult to give consideration to safety are solved.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, specifically to a lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue. Background Technology

[0002] In emergency rescue communication scenarios, there are usually multiple heterogeneous communication devices of various standards on site, and the electromagnetic environment is complex and changeable, which puts forward extremely high requirements for the real-time performance and stability of cross-device interconnection. Existing technologies for handling heterogeneous protocol conversion generally employ full-frame demodulation and a complete OSI seven-layer decapsulation model, relying on external memory for frequent data read / write interactions. This architecture not only leads to high system computing power consumption and high conversion latency, but also causes congestion and power consumption issues due to frequent bus contention. Furthermore, traditional relay methods often face security risks such as complex key management and plaintext exposure in memory during cross-system conversion, and often require restarting or reloading the software stack during protocol switching, making it difficult to adapt to the rapid response requirements of lightweight devices in dynamic real-world environments.

[0003] Therefore, how to reduce the latency and energy consumption of heterogeneous protocol conversion through streaming processing under the condition of limited hardware resources, while taking into account the high security and anti-interference capability of communication, has become an urgent technical problem to be solved. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue. Specifically, the technical solution of this invention includes: Broadband signal segmentation acquisition module, conversion scheduling control module, heterogeneous protocol streaming translation module, and dynamic spectrum fingerprinting routing module; The broadband signal segmentation acquisition module is used to acquire the broadband IQ data stream input from the radio frequency front end. Based on the preset energy threshold and instantaneous bandwidth parameters, the broadband IQ data stream is segmented into discrete signal segments in real time, the physical layer signal features of the discrete signal segments are extracted, and the segments are encapsulated to generate a unified protocol data frame. The conversion scheduling control module is used to parse the metadata tags in the unified protocol data frame, determine the source protocol type of the unified protocol data frame, and match the source protocol type with the preset list of supported protocols; The heterogeneous protocol streaming translation module is used to respond to the determination result that the source protocol type is a known protocol, and based on the preset protocol parameter mapping table, to perform payload isolation mapping operation or ciphertext shell reconstruction operation on the unified protocol data frame to generate the baseband data stream of the target protocol. The dynamic spectrum fingerprinting routing module is used to establish a transparent bridging channel in response to the determination result that the source protocol type is unknown. It performs digital spectrum shifting and resampling operations on the original IQ slice contained in the unified protocol data frame to generate signal data mapped to the target frequency band.

[0005] Preferably, the process of the broadband signal segmentation acquisition module encapsulating and generating a unified protocol data frame includes: Using a neural network model running on a digital signal processing unit, modulation identification and symbol rate estimation are performed on discrete signal segments to generate modulation type labels; Construct a unified protocol data frame data structure that includes timestamps, energy feature vectors, modulation type labels, payload pointers, and raw IQ slices; The unified protocol data frame is written into the on-chip block memory of the field-programmable gate array, and the write address of the unified protocol data frame is sent to the conversion scheduling control module.

[0006] Preferably, the process of the heterogeneous protocol streaming translation module performing payload isolation mapping operation includes: When the source protocol type is digital voice protocol and the target protocol is analog voice protocol, extract the vocoder parameters from the unified protocol data frame; Based on the protocol parameter mapping table, the vocoder parameters are directly mapped to the frequency modulation deviation value or amplitude modulation depth value of the analog voice protocol, skipping the pulse code modulation audio restoration step. The baseband data stream of the target protocol is generated by modulating the frequency modulation deviation value or amplitude modulation depth value.

[0007] Preferably, the process of the heterogeneous protocol streaming translation module performing the ciphertext shell reconstruction operation includes: When the source protocol type and the target protocol use the same encryption algorithm, the link layer header and trailer of the source protocol are stripped from the unified protocol data frame, and the encrypted payload data is retained. Based on the frame structure specification of the target protocol, a new link layer header and trailer are generated; The encrypted payload data is filled between the new link layer header and trailer, reassembled to generate the target protocol data frame, and output as a baseband data stream.

[0008] Preferably, the system also includes a hardware security adaptation module; The heterogeneous protocol streaming translation module is also used to: route unified protocol data frames to the hardware security adaptation module when the source protocol type and the target protocol use different encryption algorithms; The hardware security adaptation module is used to decrypt the payload of the unified protocol data frame and re-encrypt it into the ciphertext format of the target protocol using streaming encryption and decryption mode. The processed payload is then returned to the heterogeneous protocol streaming translation module for encapsulation.

[0009] Preferably, the process by which the dynamic spectrum fingerprint routing module performs digital spectrum shifting and resampling operations includes: Extract the time-frequency distribution characteristics of unified protocol data frames, and identify the periodic patterns of discrete signal segments based on the time-frequency distribution characteristics; Calculate the frequency offset based on the center frequency of the target frequency band; Using a numerically controlled oscillator and a complex multiplier, the original IQ slice in the unified protocol data frame is spectrum shifted, and the shifted data is resampled by polyphase filtering according to the target bandwidth.

[0010] Preferably, the heterogeneous protocol streaming translation module is also used to perform fuzzy reconstruction operations: When the checksum of the unified protocol data frame indicates that the data is damaged, the redundant information in the unified protocol data frame and the context features of historical frames are used to interpolate and repair the damaged payload data segment. Output the target protocol baseband data stream corresponding to the repaired payload data segment.

[0011] Preferably, the switching and scheduling control module is also used to perform protocol switching operations: Real-time monitoring of changes in the protocol type of the input signal stream; When a change in the source protocol type is detected, the protocol parameter mapping table pointer called by the heterogeneous protocol streaming translation module is updated without reloading the complete protocol stack software.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention uses a wideband signal segmentation acquisition module with a dual-threshold decision mechanism and a hardware-parallel comparison tree structure to track environmental noise in real time. Segment recording is triggered only when the signal energy and instantaneous bandwidth meet preset conditions, which can effectively filter transient interference and adapt to the slow-changing characteristics of environmental thermal noise, thereby avoiding processing invalid noise. By using a lightweight convolutional neural network model running on a field-programmable gate array to perform real-time modulation recognition and physical layer feature extraction on discrete signal segments, data encapsulation can be completed at the acquisition end and directly written to on-chip memory, thereby avoiding bus congestion caused by frequent access to external memory, significantly reducing computing power consumption and improving the real-time throughput of the system in complex electromagnetic environments. 2. This invention utilizes a heterogeneous protocol streaming translation module to perform payload isolation mapping operations based on a protocol parameter mapping table. This directly maps the vocoder parameters of the digital voice protocol to the modulation parameters of the analog protocol, skipping the traditional pulse code modulation audio restoration step and thus significantly reducing conversion latency. Furthermore, by having the conversion scheduling control module update the parameter mapping table pointer without reloading the software stack when a protocol change is detected, millisecond-level seamless switching between multiple protocols can be achieved. This resolves the contradiction between limited storage in lightweight devices and support for multiple protocols, ensuring the continuity of cross-network communication. 3. This invention performs a ciphertext shell reconstruction operation when the source and target protocols have the same encryption algorithm. It only strips and reconstructs the link layer header and tail to strictly lock the intermediate encryption payload, ensuring that plaintext data never appears in the device memory. This eliminates the complexity of key management and improves battlefield survivability. By enabling the streaming encryption and decryption mode of the hardware security adaptation module in cross-encryption scheme scenarios, and using pipeline circuits to complete decryption and re-encryption in real time, the plaintext exposure window can be limited to the chip's physical boundary and side-channel attacks can be prevented. This ensures both high security and high real-time performance in multi-department joint operations. 4. This invention establishes a transparent bridging channel for unknown protocols through a dynamic spectrum fingerprinting routing module, performs digital spectrum shifting and multiphase filtering resampling on the original signal slice, and can linearly map the energy of the source frequency band to the target frequency band, thereby achieving guaranteed forwarding of non-standard or unknown signal types; by performing fuzzy reconstruction operation when the data is damaged, and using soft decision demodulation stream, intra-frame redundancy information and historical frame context features to interpolate and repair the payload, it can generate an anti-interference emergence effect and output continuous speech with some noise but still understandable, thereby greatly improving the practical value of the system in weak signal and strong noise environments. Attached Figure Description

[0013] The present invention will be further explained below with reference to the accompanying drawings and embodiments: Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0015] Example 1: Please see Figure 1 A lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue, including a broadband signal segmentation acquisition module, a conversion scheduling and control module, a heterogeneous protocol streaming translation module, and a dynamic spectrum fingerprint routing module; The broadband signal segmentation acquisition module is used to acquire the broadband IQ data stream input from the radio frequency front end. Based on the preset energy threshold and instantaneous bandwidth parameters, the broadband IQ data stream is segmented into discrete signal segments in real time, the physical layer signal features of the discrete signal segments are extracted, and the segments are encapsulated to generate a unified protocol data frame. The conversion scheduling control module is used to parse the metadata tags in the unified protocol data frame, determine the source protocol type of the unified protocol data frame, and match the source protocol type with the preset list of supported protocols; The heterogeneous protocol streaming translation module is used to respond to the determination result that the source protocol type is a known protocol, and based on the preset protocol parameter mapping table, to perform payload isolation mapping operation or ciphertext shell reconstruction operation on the unified protocol data frame to generate the baseband data stream of the target protocol. The dynamic spectrum fingerprinting routing module is used to establish a transparent bridging channel in response to the determination result that the source protocol type is unknown. It performs digital spectrum shifting and resampling operations on the original IQ slice contained in the unified protocol data frame to generate signal data mapped to the target frequency band.

[0016] This embodiment details the core architecture of the system, aiming to solve the technical problems of high latency, high computing power consumption, and difficulty in ensuring security in the existing technology. The broadband signal segmentation acquisition module serves as the sensing entry point of the system and directly connects to the radio frequency front end to acquire broadband IQ data streams. Instead of performing traditional full-frame demodulation, it segments the signal stream into discrete signal segments in the time domain based on energy detection and spectrum analysis. Specifically, the module employs a dual-threshold decision mechanism, meaning it only determines whether the input signal's real-time energy exceeds a preset noise floor threshold (e.g., noise floor +6dB) and the instantaneous bandwidth calculated via Fast Fourier Transform falls within a preset parameter range. The calculation uses the energy span method: performing the calculation on the signal segment. Point FFT yields the spectral sequence Set a valid signal determination threshold, and calculate it using the following formula: ; in, The system estimates the ambient noise floor; specifically, this is achieved through statistical measurements: during the initialization phase, the system collects idle spectrum data to calculate the average power as a baseline, and during operation, it utilizes a sliding window. To adapt to the slowly varying characteristics of ambient thermal noise, the system first optimizes the broadband... Data Stream Execution The energy is downsampled and accumulated by a factor of 1, and then the length is maintained on the downsampled energy sequence. A sliding window is used to statistically analyze the minimum energy value within the window as the real-time value. The specific statistical implementation is as follows: The system maintains a first-in, first-out circular buffer of depth N. Whenever a new energy value is enqueued and overwrites an old value, a hardware-parallel comparison tree structure is used to obtain the absolute minimum value in the buffer in real time, avoiding the latency caused by serial traversal. This strategy effectively filters transient interference to adapt to the slow-changing characteristics of environmental thermal noise. ; Traversing the spectrum to extract amplitude exceeding Effective subcarrier index set If it is detected that the set is caused by the coexistence of multiple valid signals in different frequency bands, When discontinuous intervals occur, a clustering algorithm is used to group the sets. Divided into multiple independent contiguous subsets The instantaneous bandwidth is calculated independently for each independent continuous subset, and the corresponding calculation formula is adjusted as follows: ; in, For the hardware sampling rate of the RF front-end analog-to-digital converter, e.g. As an inherent time base parameter of the system, it is used to map the FFT index to the physical frequency; The number of points for the FFT transform is 512 in this embodiment, used to determine the frequency resolution; only when Slice recording is triggered only when the instantaneous bandwidth of at least one subset meets the above conditions; otherwise, it is considered invalid noise. For cases with multiple independent continuous subsets, slice recording is triggered as long as the instantaneous bandwidth of at least one subset meets the above conditions. The module extracts physical layer features and encapsulates them to generate a unified protocol data frame. This data frame serves as a standardized intermediate data structure, shielding the differences between various protocols at the physical layer feature level. The conversion scheduling control module, acting as the system's central nervous system, receives the write address of the unified protocol data frame, accesses the on-chip shared memory area based on this address, reads and parses the metadata tags within it to determine the source protocol type of the frame, and performs a hash match between the identified source protocol type and a preset list of supported protocols. The specific matching logic is as follows: extract the modulation type ID and normalized symbol rate from the metadata tag. The formula for calculating hash key values ​​is as follows: ; in, The normalized symbol rate is defined in Example 2; The preset fixed-point quantization coefficient, for example, 10000, is used to map the normalized sign rate to the integer field to avoid hash collisions; in the formula, the sign... Represents the binary bitwise left shift operator, symbol This represents the binary bitwise XOR operator; it searches for the corresponding protocol descriptor in the hash table indexed by this key value; this descriptor explicitly defines the entry address of the mapping table required for subsequent processing, thereby determining the subsequent processing path, and passing the address pointer of the data frame to the subsequent module to achieve zero-copy data flow; The heterogeneous protocol streaming translation module, as the core computing engine, responds to the determination that the source protocol is of a known type. Based on the received address pointer, it obtains the data frame content. Instead of performing a complete OSI seven-layer decapsulation, it directly performs payload isolation mapping or ciphertext shell reconstruction on the data frame at the atomic level based on a preset protocol parameter mapping table. That is, it skips the complete multi-layer decapsulation steps of the network communication protocol stack, treats the data frame as the basic minimum processing unit, and directly reads and modifies the basic data blocks such as vocoder parameters, link layer headers and trailers, or encrypted payloads inside the frame through a fixed address offset, generating a baseband data stream that conforms to the target protocol specification. The dynamic spectrum fingerprinting routing module, as the system's backup mechanism, responds to the determination of an unknown protocol, establishes a transparent bridging channel, and directly performs digital spectrum shifting and resampling on the data frame containing the original IQ slice, linearly mapping the energy of the source frequency band to the target frequency band.

[0017] Example 2: The process of generating a unified protocol data frame by the broadband signal segmentation acquisition module includes: using a neural network model running on a digital signal processing unit to perform modulation identification and symbol rate estimation on discrete signal segments, and generating modulation type labels; constructing a unified protocol data frame data structure containing timestamps, energy feature vectors, modulation type labels, payload pointers, and original IQ slices; writing the unified protocol data frame into the on-chip block memory of the field-programmable gate array, and sending the write address of the unified protocol data frame to the conversion scheduling control module.

[0018] This embodiment further illustrates that the broadband signal segmentation acquisition module is not a simple sampler, but a preprocessing unit with edge intelligence. The module uses a lightweight convolutional neural network model running on a field-programmable gate array digital signal processing unit to perform real-time inference on the segmented discrete signal segments. The input is the time-frequency diagram of IQ data, and the output is the modulation type label and the estimated symbol rate, thereby realizing feature extraction at the physical layer. To meet the code-level reproduction requirements, the input time-frequency graph generation process is as follows: The system processes discrete signal segments... Perform a short-time Fourier transform using the Hanning window. and 50% overlap After calculating the amplitude spectrum, the spectrum matrix is ​​adjusted using bilinear interpolation. Dimensions; Min-Max normalization is performed, and the calculation formula is as follows: ; in, For time-frequency plots on coordinates The original amplitude value at that location, and These represent the maximum and minimum amplitude values ​​in the time-frequency plot matrix corresponding to the discrete signal segment, respectively. To prevent division by zero errors, a minimum value protection quantity is used; a single-channel grayscale image with a value range between [0,1] is generated; specifically, the neural network model adopts a MobileNetV2 variant architecture that is tailored to adapt to FPGA resources—the specific tailoring implementation includes: removing the 4th to 17th inverse residual blocks used to extract high-level semantics in the standard model, and retaining only the first 3 blocks to focus on physical layer texture features; Meanwhile, the 1x1 extended convolutional layer before the global average pooling layer was removed, and the original number of channels was 1280, reducing the number of model parameters to less than 5% of the standard version; in addition, the INT8 dynamic fixed-point quantization strategy was adopted to map the 32-bit floating-point weights to 8-bit integers to match the DSP48E1 slice resources of the FPGA. The input layer receives the single-channel time-frequency grayscale image normalized to 64x64 pixels. The main body of the network includes a 3x3 standard convolutional layer with a stride of 2 as the stem, and its output channel number is configured to 16 to balance the computational power consumption. Three inverse residual blocks are cascaded to extract deep features. To ensure the determinism of the model in hardware implementation, the specific configuration parameters of these three residual blocks, i.e., the expansion factor, are determined. Number of output channels and step length The first block is set sequentially as follows: The second piece The third piece ; A fully connected layer is connected via a global average pooling layer. This fully connected layer is designed as a multi-task output structure, containing a softmax classification branch to output probability vector labels corresponding to preset modulation types such as BPSK, QPSK, and 16QAM, and a linear regression branch to output normalized symbol rate estimates. To ensure that the hash key value calculation formula in Example 1 is accurate... To ensure uniqueness and reproducibility, the normalized mapping function of the symbol rate is explicitly defined here as: ; in, The actual physical symbol rate is expressed in Baud; the denominator is... This represents the Nyquist frequency limit, which is the highest signal frequency that the system can theoretically resolve. The hardware sampling rate of the analog-to-digital converter, such as This definition ensures And with subsequent fixed-point quantization coefficients Matching; Based on this, the system constructs a unified protocol data frame data structure inside the FPGA. This structure rigorously integrates a timestamp for timing reassembly, an energy feature vector for assisting silence detection, a modulation type label generated by a neural network, a payload pointer pointing to the original data buffer location, and an IQ slice that retains the original waveform data. In order to match the high throughput of the FPGA, the generated unified protocol data frame is not passed through external DDR memory, but is directly written into the on-chip block memory of the FPGA. Only the write address of the data frame is sent to the conversion scheduling control module via a lightweight bus, thus achieving physical separation of the data flow and the control flow; This embodiment effectively avoids bus congestion and power consumption problems caused by frequent access to external memory by completing physical layer feature extraction and encapsulation at the acquisition end. In the complex electromagnetic environment of multiple concurrent signals, the high-speed read and write characteristics of on-chip memory are utilized to significantly improve the real-time throughput of the system, laying a solid data foundation for subsequent streaming processing and ensuring system stability under high load.

[0019] Example 3: The process of the heterogeneous protocol streaming translation module performing payload isolation mapping includes: when the source protocol type is a digital voice protocol and the target protocol is an analog voice protocol, extracting vocoder parameters from the unified protocol data frame; based on the protocol parameter mapping table, directly mapping the vocoder parameters to the frequency modulation deviation value or amplitude modulation depth value of the analog voice protocol, skipping the pulse code modulation audio restoration step; and generating the baseband data stream of the target protocol based on the frequency modulation deviation value or amplitude modulation depth value.

[0020] This embodiment details the payload isolation mapping operation in the scenario of converting from a digital voice protocol to an analog voice protocol. This process employs direct vocoder parameter mapping technology. In response to the source protocol being identified as a digital voice protocol, the module only decrypts the link layer and extracts vocoder parameters from the unified protocol data frame, such as pitch period, codebook index, and gain factor. Based on a preset protocol parameter mapping table, the system uses a mapping transfer function to directly map the extracted digital features to the modulation parameters of the analog protocol. The calculation formula for the gain factor is as follows: ; in, The modulation parameters for the target analog protocol are physically represented as frequency modulation deviation or amplitude modulation depth, expressed in Hertz or as a percentage. The vocoder energy gain parameter of the source protocol is explicitly defined here as a normalized linear scalar value, dimensionless, with a value range of [0,1]. It is derived from the vocoder quantization gain index extracted from the unified protocol data frame, after lookup table inverse quantization and normalization. These are gain mapping coefficients, derived from a preset protocol parameter mapping table stored in Flash. The system can dynamically load the corresponding row configuration based on the target protocol ID to adjust volume sensitivity. Their physical units are consistent with the modulation parameter units of the target protocol; that is, the unit is Hz when the target is FM, and the unit is percentage when the target is AM. Since it is a dimensionless value, therefore here It directly undertakes the function of assigning dimensions to ensure the dimensional balance of the formula; The basic modulation bias, derived from the standard specifications of the target protocol, has its physical meaning and units adaptively adjusted according to the target protocol standard: when the target protocol is a frequency modulation standard... The silent frequency offset, expressed in Hertz (Hz), represents the carrier center frequency; when the target protocol is amplitude modulation (AM). This represents the baseline amplitude modulation depth offset in silent mode, expressed as a percentage, to ensure dimensional balance in the formula calculations. For the fundamental period parameter, the system maps it to the audio frequency control word that drives the numerically controlled oscillator; the specific mapping formula is as follows: ; in, For the extracted pitch period, For FPGA system clock frequency, Given the bit width of the NCO phase accumulator, such as 32 bits, the calculated... This causes the pitch of the generated baseband signal to vary with the fundamental frequency, while the calculated pitch above... The value is then used as the control variable for the output amplitude of the NCO; the numerically controlled oscillator is directly driven according to the calculated modulation parameters to generate the analog baseband data stream of the target protocol, thereby skipping the traditional pulse code modulation audio restoration step.

[0021] Example 4: The process of the heterogeneous protocol streaming translation module performing ciphertext shell reconstruction includes: when the source protocol type and the target protocol use the same encryption algorithm, stripping the link layer header and trailer of the source protocol in the unified protocol data frame and retaining the encrypted payload data; Based on the frame structure specification of the target protocol, a new link layer header and trailer are generated; encrypted payload data is filled between the new link layer header and trailer, reassembled to generate a data frame of the target protocol, and output as a baseband data stream.

[0022] This embodiment describes in detail the ciphertext shell reconstruction operation in a scenario where the source and target protocols use the same encryption algorithm but have different link layer encapsulation formats. The module accesses the memory to read the data frame according to the received address pointer. The system identifies and strips the link layer header and trailer of the source protocol in the unified protocol data frame and marks them as discarded data. At the same time, it strictly locks the encrypted payload data in the middle to ensure that it is not tampered with or read. Based on the frame structure specification of the target protocol, the FPGA logic generates a new link layer header containing the target device ID and synchronization word, as well as a trailer containing a new checksum. During this process, the system executes a payload length adaptation algorithm to obtain the length of the source encrypted payload. Payload length allowed by the target protocol's maximum transmission unit ;like Then the fragmentation mechanism is activated, and the source payload is divided into... Each sub-fragment is processed, and an extended header containing the fragment sequence number is generated for each sub-fragment; if Then, a padding operation is performed, generating a length of: ; Pseudo-random filling sequence Append it to the end of the payload and mark the payload length field in the new header; execute the zero-trust reassembly logic to fill the new header and trailer with the locked encrypted payload data, or its fragmented / padded data, and reassemble it to generate a data frame for the target protocol; output the frame as a baseband data stream through parallel-to-serial conversion to complete the protocol conversion; This embodiment implements a zero-trust relay mechanism, in which the encrypted payload is never decrypted and the plaintext data never appears in the device's memory during the entire conversion process. In joint rescue operations involving highly classified communications, this design eliminates the complexity of key management, and even if the device is lost, attackers cannot extract sensitive communication content from memory, greatly improving the system's security and battlefield survivability.

[0023] Example 5: The system also includes a hardware security adaptation module; the heterogeneous protocol streaming translation module is also used to: route unified protocol data frames to the hardware security adaptation module when the source protocol type and the target protocol use different encryption algorithms; the hardware security adaptation module is used to decrypt the payload of the unified protocol data frame and re-encrypt it into the ciphertext format of the target protocol using streaming encryption and decryption mode, and return the processed payload to the heterogeneous protocol streaming translation module for encapsulation; This embodiment introduces a hardware security adaptation module to handle cross-system encryption conversion scenarios. When the heterogeneous protocol streaming translation module detects a discrepancy between the source and target protocol encryption algorithm IDs, it immediately routes the unified protocol data frame to the dedicated data channel of the hardware security adaptation module. This module does not use the traditional interrupt service mode, but instead initiates a streaming encryption / decryption mode, receiving data like a pipeline and performing decryption and re-encryption operations in real time in the hardware circuit. Its key stream conversion logic is as follows: ; in, The encrypted data stream of the target protocol is calculated and output by the hardware security module; The encrypted payload of the source protocol originates from the unified protocol data frame; For source key The decryption operator is used to restore the plaintext; For target key Encryption operators are used to generate the target ciphertext; To implement the aforementioned streaming logic in the hardware circuitry and avoid blocking, the hardware security adaptation module internally incorporates a dual-path design. Counter mode engine; the specific circuit connection is as follows: the input data frame is stored in an asynchronous FIFO buffer with a depth of 512 bytes to span the clock domain; The decryption engine reads the FIFO data and uses the source key. and the initial vector extracted from the frame header Drive the AES core to generate a synchronization key stream Executed via XOR gate array Restoring the plaintext; The encryption engine uses the target key and locally generated random initial vector Generate a new key stream ,implement ;Will The target frame header is written, and the entire process is completed in the FPGA pipeline with a single clock cycle throughput. The processed payload is immediately returned to the heterogeneous protocol streaming translation module for final frame encapsulation, completing the cross-encryption system adaptation. This embodiment uses a dedicated hardware security anchor to handle cross-system encryption conversion, ensuring that even in scenarios where decryption is necessary, the time window for plaintext exposure is limited to the chip-level physical boundary. In multi-department joint operations, this streaming processing mode not only prevents side-channel attacks but also minimizes the introduced latency, balancing high security and high real-time performance.

[0024] Example 6: The process of the dynamic spectrum fingerprinting routing module performing digital spectrum shifting and resampling operations includes: extracting the time-frequency distribution characteristics of the unified protocol data frames, and identifying the periodic patterns of discrete signal segments based on the time-frequency distribution characteristics; Calculate the frequency offset based on the center frequency of the target frequency band; use a numerically controlled oscillator and a complex multiplier to perform spectrum shifting on the original IQ slice in the unified protocol data frame, and perform polyphase filtering and resampling on the shifted data according to the target bandwidth.

[0025] This embodiment describes in detail a transparent bridging scheme for unknown protocols or interference signals; the module receives the unified protocol data frame address pointer issued by the conversion scheduling control module, reads the data frame from the FPGA on-chip block memory, and extracts its time-frequency distribution characteristics. Specifically, the energy spectral density of the signal is calculated using short-time Fourier transform, and autocorrelation analysis is performed on the energy envelope along the time axis. The specific calculation logic is as follows: summing the STFT matrix along the frequency axis yields the time-domain energy sequence. Calculate its autocorrelation function; the formula is as follows: ; And search In the lag interval ,in, In this embodiment, the preset minimum hysteresis offset is set to 32 to avoid the autocorrelation peak at zero hysteresis; based on this, the hysteresis interval is searched. Significant peak within; Only the local maximum value The effective peak value is determined to identify the signal repetition period, thereby identifying the periodicity of discrete signal segments, such as the burst period of time division multiple access, to determine the effective duty cycle of the signal. Based on this, the system further precisely locates the center frequency of the source signal. The specific algorithm is as follows: The time-frequency matrix generated by STFT... Along the time axis Summing yields the power spectrum vector ,search Peak index Then the center frequency is: ; in, The number of FFT points used in the Short Time Fourier Transform. This refers to the ADC hardware sampling rate defined in the implementation of Example 2; Based on the center frequency of the source signal and the center frequency of the target forwarding band configured in the system, the normalized angular frequency offset is calculated using the following formula: ; in, This is the normalized angular frequency offset, in radians per sample. The target frequency band center frequency is derived from the system configuration. The center frequency of the source signal is estimated by the acquisition module. The system hardware sampling rate is a hardware intrinsic constant, and its sign is consistent with that used in embodiments 1 and 2. The original IQ slice is processed using a numerically controlled oscillator and a complex multiplier within the FPGA. The formula for calculating the mixing operation is as follows: ; in, The imaginary unit ( ), The complex signal after spectrum shifting is output by the multiplier; This refers to the raw IQ slice data in the unified protocol data frame; For discrete-time indexing; using a polyphase filter bank to index according to the target bandwidth Resampling is performed to match the receiving bandwidth of the target device; Specifically, the calculation logic for the resampling parameters is as follows: The system queries the corresponding standard sampling rate according to the target protocol specification. For example, the TETRA protocol is 36kHz, combined with the source sampling rate. , here Numerically, it is strictly equal to the system hardware sampling rate. For example, at 122.88MHz, because the input data is a slice of the original IQ, the resampling ratio is calculated. And simplify to coprime fractions ; Construct a polyphase filter bank, where the cutoff frequency of the prototype low-pass filter is set to... order This ensures that image frequency and aliasing interference are effectively suppressed during rate conversion.

[0026] Example 7: The heterogeneous protocol streaming translation module is also used to perform fuzzy reconstruction operations: when the check bit of the unified protocol data frame shows that the data is damaged, the module uses the redundant information in the unified protocol data frame and the context features of the historical frames to interpolate and repair the damaged payload data segment; and outputs the target protocol baseband data stream corresponding to the repaired payload data segment.

[0027] This embodiment describes the fuzzy reconstruction operation logic in a high-noise environment. When the checksum of a unified protocol data frame indicates data corruption, the system does not directly discard the frame. Instead, it first performs soft-decision demodulation on the original IQ slice based on the modulation type label in the frame header, extracts the demodulated bitstream and the implicit syntactic information, and then uses the redundant information in the unified protocol data frame and the context features of historical frames to perform interpolation repair based on redundancy fusion on the damaged payload data segment. The hybrid repair formula is as follows: ; in, These are the load parameter values ​​after repair; The local estimate is obtained by attempting error correction decoding using the maximum likelihood criterion based on the soft-decision demodulation stream and the residual FEC check bits within the frame. The redundancy confidence weight is calculated as follows: ; in, The error count is obtained by calculating the syndrome vector of the demodulated bitstream. Specifically, the calculation involves: calculating the syndrome vector of the received codeword based on the BCH generator polynomial; using this vector as an index to query a pre-defined syndrome-error pattern table; and counting the number of non-zero bits in the corresponding error pattern to obtain the error count. The maximum number of error correction bits defined for the FEC algorithm; In this embodiment, the unified protocol data frame uses the BCH(15,5) encoding algorithm, which has the ability to correct 3 random bit errors, therefore it is set as follows: ; These are valid parameter values ​​from the previous historical frame, derived from the cache. The valid parameter values ​​for the next frame are obtained from the cache; in order to obtain the parameters for the next frame... The system maintains a depth of [depth value missing] within the heterogeneous protocol streaming translation module. A tiny jitter buffer for frames, which introduces a fixed processing delay. Designed to operate within the limits allowed for emergency voice communication; These are the interpolation weighting coefficients, whose values ​​are dynamically adjusted based on the current signal-to-noise ratio (SNR). The SNR calculation utilizes the energy feature vector carried in the unified protocol data frame. The calculation formula is: ; in, To prevent division by zero errors in the minimum value, The function is used to prevent negative input values ​​caused by energy fluctuations. The energy of the current frame is directly extracted from... The first component of the vector, representing the average power statistic of the current slice, is expressed in linear power values ​​(watts, W), not logarithmic units. This serves as a baseline for the background noise energy tracked using a sliding minimum filter. The specific implementation of the sliding minimum filter is as follows: The system maintains a length of... A circular buffer for frames stores historical energy values; this filter, upon the arrival of the first frame during system cold start, will... Initialize to a preset thermal noise reference value, such as Corresponding to -120dBm; at each moment The minimum value within the buffer is calculated using the following formula: ; in, The energy sequence is within the circular buffer; to avoid excessively low noise floor estimation due to transient deep fading, an asymmetric smoothing update formula is used: if The calculation formula is as follows: ; in, Conversely The above parameters and The value is selected based on a trade-off between the typical speech frame length and the rate of change of ambient noise: smaller... It ensures smooth suppression of sudden, strong interference, while larger... This ensures that the filter can converge and update quickly when the background noise actually decreases, thereby achieving fast tracking and slow release of the background noise level. The specific dynamic adjustment strategy is as follows: set two signal-to-noise ratio thresholds. and When the real-time detection signal-to-noise ratio At that time, the current channel quality is considered good, and the next frame... High credibility, settings ;when At that time, the channel quality was generally poor, so the settings were adjusted accordingly. As SNR increases linearly, the specific calculation formula is as follows: ; This includes achieving a smooth transition from 0.2 to 0.6; when At that time, it was deemed that the newly received data was extremely unreliable, and a forced setting was implemented. Maintain a minimum historical weighting, completely preserve the parameter values ​​of the previous frame, and prevent noise from being introduced; use the repaired parameters to generate the target baseband stream, and output a signal that is partially noisy but still continuous. This embodiment produces an anti-interference emergence effect. Through fuzzy reconstruction, the system outputs continuous speech with some noise but still understandable, rather than intermittent silence. In emergency command, this effect of being audible but with noise is far better than being completely inaudible, greatly enhancing the system's practical value and the continuity of information transmission in weak signal environments.

[0028] Example 8: The switching and scheduling control module is also used to perform protocol switching operations: real-time monitoring of changes in the protocol type of the input signal stream; when a change in the source protocol type is detected, updating the protocol parameter mapping table pointer called by the heterogeneous protocol streaming translation module without reloading the complete protocol stack software.

[0029] This embodiment describes in detail the efficient protocol switching operation mechanism; the system continuously monitors the protocol type label of the input signal stream in real time to sense changes in the communication environment; When a change in the source protocol type is detected, such as switching from PDT to TETRA, the module does not perform time-consuming FPGA image reload or software stack restart; the system only updates the protocol parameter mapping table pointer called by the heterogeneous protocol streaming translation module, making it instantly point to the parameter table of the new protocol; due to the universality of the mapping logic, the switching process only involves the modification of the memory pointer, achieving a seamless transition at the clock cycle level. This embodiment cleverly resolves the contradiction between limited storage and multi-protocol support in lightweight devices. By reusing general streaming translation logic, multiple protocols can be supported by storing only a small parameter table. In complex rescue networks, this enables the system to achieve seamless roaming with extremely low latency between different network standards without the need to restart or reload waveforms like traditional radios, ensuring the continuity and response speed of cross-network communication.

[0030] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue, characterized in that, It includes a broadband signal segmentation acquisition module, a conversion scheduling and control module, a heterogeneous protocol streaming translation module, and a dynamic spectrum fingerprinting routing module; The broadband signal segmentation acquisition module is used to acquire the broadband IQ data stream input from the radio frequency front end. Based on the preset energy threshold and instantaneous bandwidth parameters, the broadband IQ data stream is segmented into discrete signal segments in real time, the physical layer signal features of the discrete signal segments are extracted, and the segments are encapsulated to generate a unified protocol data frame. The conversion scheduling control module is used to parse the metadata tags in the unified protocol data frame, determine the source protocol type of the unified protocol data frame, and match the source protocol type with the preset list of supported protocols; The heterogeneous protocol streaming translation module is used to respond to the determination result that the source protocol type is a known protocol, and based on the preset protocol parameter mapping table, to perform payload isolation mapping operation or ciphertext shell reconstruction operation on the unified protocol data frame to generate the baseband data stream of the target protocol. The dynamic spectrum fingerprinting routing module is used to establish a transparent bridging channel in response to the determination result that the source protocol type is unknown. It performs digital spectrum shifting and resampling operations on the original IQ slice contained in the unified protocol data frame to generate signal data mapped to the target frequency band.

2. The lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue according to claim 1, characterized in that, The process of encapsulating and generating unified protocol data frames by the broadband signal segmentation acquisition module includes: Using a neural network model running on a digital signal processing unit, modulation identification and symbol rate estimation are performed on discrete signal segments to generate modulation type labels; Construct a unified protocol data frame data structure that includes timestamps, energy feature vectors, modulation type labels, payload pointers, and raw IQ slices; The unified protocol data frame is written into the on-chip block memory of the field-programmable gate array, and the write address of the unified protocol data frame is sent to the conversion scheduling control module.

3. The lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue according to claim 1, characterized in that, The process of the heterogeneous protocol streaming translation module performing payload isolation mapping operations includes: When the source protocol type is digital voice protocol and the target protocol is analog voice protocol, extract the vocoder parameters from the unified protocol data frame; Based on the protocol parameter mapping table, the vocoder parameters are directly mapped to the frequency modulation deviation value or amplitude modulation depth value of the analog voice protocol, skipping the pulse code modulation audio restoration step. The baseband data stream of the target protocol is generated by modulating the frequency modulation deviation value or amplitude modulation depth value.

4. The lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue according to claim 1, characterized in that, The process of the heterogeneous protocol streaming translation module performing ciphertext shell reconstruction includes: When the source protocol type and the target protocol use the same encryption algorithm, the link layer header and trailer of the source protocol are stripped from the unified protocol data frame, and the encrypted payload data is retained. Based on the frame structure specification of the target protocol, a new link layer header and trailer are generated; The encrypted payload data is filled between the new link layer header and trailer, reassembled to generate the target protocol data frame, and output as a baseband data stream.

5. The lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue according to claim 1, characterized in that, The system also includes a hardware security adaptation module; The heterogeneous protocol streaming translation module is also used to: route unified protocol data frames to the hardware security adaptation module when the source protocol type and the target protocol use different encryption algorithms; The hardware security adaptation module is used to decrypt the payload of the unified protocol data frame and re-encrypt it into the ciphertext format of the target protocol using streaming encryption and decryption mode. The processed payload is then returned to the heterogeneous protocol streaming translation module for encapsulation.

6. The lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue according to claim 1, characterized in that, The process by which the dynamic spectrum fingerprint routing module performs digital spectrum shifting and resampling operations includes: Extract the time-frequency distribution characteristics of unified protocol data frames, and identify the periodic patterns of discrete signal segments based on the time-frequency distribution characteristics; Calculate the frequency offset based on the center frequency of the target frequency band; Using a numerically controlled oscillator and a complex multiplier, the original IQ slice in the unified protocol data frame is spectrum shifted, and the shifted data is resampled by polyphase filtering according to the target bandwidth.

7. The lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue according to claim 1, characterized in that, The heterogeneous protocol streaming translation module is also used to perform fuzzy reconstruction operations: When the checksum of the unified protocol data frame indicates that the data is damaged, the redundant information in the unified protocol data frame and the context features of historical frames are used to interpolate and repair the damaged payload data segment. Output the target protocol baseband data stream corresponding to the repaired payload data segment.

8. The lightweight multi-device communication protocol intelligent adaptation and conversion system for emergency rescue according to claim 1, characterized in that, The switching and scheduling control module is also used to perform protocol switching operations: Real-time monitoring of changes in the protocol type of the input signal stream; When a change in the source protocol type is detected, the protocol parameter mapping table pointer called by the heterogeneous protocol streaming translation module is updated without reloading the complete protocol stack software.

Images