Data transmission method based on intracranial and extracranial wireless communication system

Abstract

The application discloses a data transmission method based on an intracranial and extracranial wireless communication system, and relates to the technical field of wireless communication in and out of a living body, and comprises the following steps: dynamically selecting a carrier frequency point according to the thickness and density of a skull, and establishing a special communication frequency band between an extracorporeal master control end and an intracranial implant device; the extracorporeal master control end performs frequency agility in the special communication frequency band according to real-time channel quality, and sends a forward control instruction to the intracranial implant device through different carrier frequencies; the carrier frequency point is dynamically selected by combining the thickness and density characteristics of individual skulls, and frequency agility based on real-time channel quality is realized in the special communication frequency band, so that the penetration efficiency and link robustness of the intracranial and extracranial wireless communication are improved; meanwhile, a dynamic impedance mapping load modulation and high-precision current fluctuation decoding mechanism is adopted to realize reliable reverse data transmission under the premise of extremely low power consumption, and the forward and reverse subbands are reasonably isolated according to the skull dispersion characteristics.

Description

Technical Field

[0001] This invention relates to the field of in vivo and in vitro wireless communication technology, and in particular to a data transmission method for wireless communication systems based on intracranial and extracranial communication. Background Technology

[0002] In vivo and in vitro wireless communication technology refers to a technological system that enables data transmission, command interaction, and energy transfer between implanted medical electronic devices and external control or monitoring systems via wireless means. This technology must overcome the absorption, scattering, and dispersion effects of electromagnetic waves by biological tissues, and establish a stable and reliable two-way communication link while ensuring biosafety, low power consumption, and miniaturization. It is widely used in implantable medical scenarios such as neuromodulation, cardiac pacing, blood glucose monitoring, and brain-computer interfaces.

[0003] Traditional implantable communication systems typically use fixed carrier frequencies, failing to consider the variations in penetration loss caused by differences in the thickness, density, and other anatomical structures of different patients' skulls. This results in unstable or even interrupted communication links. Furthermore, the intracranial environment is affected by factors such as physiological activities and tissue hydration, causing channel quality to fluctuate over time. Existing systems often employ static frequency points or simple frequency hopping strategies, making it difficult to maintain reliable command transmission. In addition, most implantable devices require active radio frequency transmission to send back data, which not only significantly increases power consumption but also necessitates the use of independent antennas and power amplifiers, conflicting with the miniaturization and long-life implantation requirements. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a data transmission method for a wireless communication system based on intracranial and extracranial environments. This solves the problem that the intracranial environment is affected by factors such as physiological activities and tissue hydration, and the channel quality fluctuates over time. Existing systems mostly use static frequency points or simple frequency hopping strategies, which make it difficult to maintain reliable command transmission. At the same time, most implanted devices need to transmit data back via active radio frequency, which not only greatly increases power consumption but also requires independent antennas and power amplifiers, conflicting with the miniaturization and long lifespan implantation requirements.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] This invention provides a data transmission method for a wireless communication system based on intracranial and extracranial structures, comprising:

[0008] Based on the thickness and density of the skull, the carrier frequency is dynamically selected and a dedicated communication frequency band is established between the external main control terminal and the intracranial implanted device.

[0009] The external master control terminal performs frequency agility within the dedicated communication frequency band based on real-time channel quality, and sends positive control commands to the intracranial implanted device through different carrier frequencies;

[0010] The intracranial implanted device receives and demodulates the carrier frequency, identifies the corresponding command content, and generates feedback sensing data.

[0011] The intracranial implantable device modulates the load using dynamic impedance mapping based on feedback sensing data and encodes the data into the reflection coefficient;

[0012] The external main control unit monitors the changes in the charging current in the coil circuit, identifies the corresponding current fluctuation pattern in the reflection coefficient, and decodes the feedback sensing data.

[0013] The external master control unit corrects the carrier frequency selection and power supply strategy for the next communication cycle based on the decoded feedback sensor data.

[0014] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system of the present invention, the specific steps of dynamically selecting the carrier frequency point based on skull thickness and density are as follows:

[0015] Before system deployment, multiple sets of ex vivo skull samples with different thicknesses and densities were collected.

[0016] For each sample, an electromagnetic wave transmission test platform was built in vitro. Continuous wave signals were emitted point by point within the preset frequency band and the power of the receiving end was recorded synchronously. The penetration loss at each frequency was calculated. At the same time, several test frequency points were extended outside the upper and lower boundaries of the preset frequency band for supplementary measurements to cover the resonance valley displacement that occurs in individuals with thick skulls or high density.

[0017] Based on the penetration loss data of all test frequencies, a mapping relationship between penetration loss and frequency, skull thickness, and skull density is established, and the mapping relationship is stored in the external master control terminal in the form of a parameterized model.

[0018] During the communication initialization phase, the external master control terminal sequentially sends multiple detection signals to the intracranial implanted device, with each detection signal using a different candidate frequency.

[0019] After receiving the detection signal, the intracranial implanted device measures the received power of each signal and encodes the received power sequence, then transmits it back to the external master control terminal via the reverse link. After receiving the feedback, the external master control terminal evaluates the penetration performance of each candidate frequency under the current individual skull conditions by combining it with the pre-stored mapping relationship.

[0020] The frequency with the best penetration performance is selected from the candidate frequency set as the initial carrier frequency, where the penetration performance is inversely proportional to the penetration loss.

[0021] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system of the present invention, wherein: the external master control terminal performs frequency agility in a dedicated communication frequency band based on real-time channel quality, and the specific steps are as follows:

[0022] Pilot symbols are inserted at fixed positions in each communication frame. The pilot symbols use a known modulation format and have constant power.

[0023] The intracranial implanted device demodulates the pilot symbols, calculates the signal-to-noise ratio of the current channel, quantizes the signal-to-noise ratio into a finite state identifier, and embeds it into the payload of the feedback sensing data.

[0024] When decoding feedback sensor data, the external master control terminal extracts the status identifier and restores the current channel quality. If the channel quality is lower than the preset safety threshold, the frequency switching process is initiated.

[0025] The frequency switching process includes querying the mapping relationship, excluding frequencies that have been used recently and whose channels are degraded, and selecting the frequency with the smallest predicted penetration loss value from the remaining candidate frequencies as the new carrier frequency.

[0026] After updating the communication parameters, the next frame of positive control command is sent using the new carrier frequency.

[0027] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system of the present invention, wherein: the intracranial implanted device modulates the load using dynamic impedance mapping based on feedback sensing data, and the specific steps are as follows:

[0028] A programmable impedance network is set at the radio frequency front end of the intracranial implantable device, which supports at least two discrete impedance states.

[0029] The two impedance states correspond to binary logic 0 and 1, respectively;

[0030] When the feedback sensor data to be sent is 0, the impedance network switches to the first impedance value; when the data is 1, it switches to the second impedance value.

[0031] Impedance switching operation is synchronized with the energy transfer cycle, and one or more state switching is completed within each energy pulse window;

[0032] The switching amplitude of the impedance network can be dynamically adjusted, including increasing the impedance difference in the high current stage to enhance the change in reflection coefficient, and decreasing the impedance difference in the low current stage to avoid nonlinear distortion.

[0033] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system of the present invention, wherein: the external master control terminal identifies the current fluctuation pattern corresponding to the reflection coefficient by monitoring the change in charging current in the coil circuit, and the specific steps are as follows:

[0034] A high-bandwidth current sensor is integrated into the energy emission coil drive circuit of the external main control terminal;

[0035] The instantaneous value of charging current is continuously acquired at a sampling rate of not less than 10 times the reverse data rate;

[0036] The acquired current sequence is subjected to real-time sliding window differential processing to extract microsecond-level perturbation components;

[0037] The disturbance waveform is matched point by point with a pre-stored modulation template library, which contains typical current fluctuation patterns corresponding to 0 and 1.

[0038] When the cross-correlation value between the template and the current disturbance waveform exceeds the preset judgment threshold, it is determined that the corresponding bit has been received;

[0039] A majority voting mechanism is used to fuse the decision results of multiple consecutive sampling windows and output the final decoded bits.

[0040] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system described in this invention, the external master control terminal corrects the carrier frequency selection and energy supply strategy for the next communication cycle based on the decoded feedback sensor data. The specific steps are as follows:

[0041] The feedback sensor data obtained from the current decoding, together with the channel quality records from the past several communication cycles, are input into the link control module.

[0042] The link control module determines whether the current channel is in a state of continuous degradation. If the determination result is yes, the energy transmission power of the next cycle is increased and the usage time of the current carrier frequency is shortened to accelerate the frequency switching response speed. If the determination result is no, and the channel quality is stable and better than the safety threshold, the energy transmission power is reduced to reduce the thermal effect of biological tissues and the usage time of the current frequency is extended to improve the spectrum efficiency.

[0043] The updated energy supply parameters and frequency scheduling plan are written into the communication controller.

[0044] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system of the present invention, wherein: the dedicated communication frequency band is divided into a forward sub-band and a reverse sub-band, including:

[0045] The forward subband is used by the external master control terminal to send control commands to the intracranial implanted device; the reverse subband corresponds to the reflection path generated by the intracranial implanted device through load modulation; the center frequency of the forward subband is higher than that of the reverse subband, and a fixed guard interval is maintained between the two, wherein the width of the guard interval is determined according to the dispersion effect of the dielectric properties of the skull.

[0046] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system described in this invention, the setting of the protection interval needs to be compatible with the specification requirements of the International Medical Implantable Communication Service (MICS) frequency band, including within the frequency band of 402 MHz to 430 MHz, with the forward subband limited to 410 MHz to 430 MHz and the reverse subband limited to 402 MHz to 410 MHz.

[0047] In a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system of the present invention, the process of constructing the mapping relationship includes the following steps:

[0048] Multiple sets of ex vivo skull samples were selected. For each sample, an in vitro transmission test platform was built, with the transmitting antenna placed on one side of the sample and the receiving antenna placed on the other side.

[0049] Within the range of 402 MHz to 430 MHz, the transmit frequency is scanned in steps of 1 MHz, the received power at each frequency is recorded, and the penetration loss is calculated.

[0050] With penetration loss as the dependent variable and frequency, thickness, and density as independent variables, a nonlinear regression method was used to fit the physical model.

[0051] The fitted model parameters are stored in the non-volatile memory of the external master control unit for online frequency optimization during operation.

[0052] As a preferred embodiment of the data transmission method of the intracranial and extracranial wireless communication system described in this invention, in the actual communication process, the external master control terminal calls the mapping relationship according to the individual skull parameters acquired in real time to dynamically estimate the penetration loss at any candidate frequency.

[0053] When the skull thickness h increases, the penetration loss L increases positively; when the carrier frequency f increases, the penetration loss L increases superlinearly; the system avoids high-loss regions and preferentially selects low-frequency points accordingly; when the skull density ρ decreases, the penetration loss L decreases accordingly.

[0054] The beneficial effects of this invention are as follows: By dynamically optimizing carrier frequency points based on individual skull thickness and density characteristics, and achieving frequency agility based on real-time channel quality within a dedicated communication frequency band, the penetration efficiency and link robustness of intracranial and extracranial wireless communication are improved. At the same time, by adopting load modulation with dynamic impedance mapping and a high-precision current fluctuation decoding mechanism, reliable reverse data transmission is achieved under extremely low power consumption. By rationally isolating the forward and reverse subbands according to the skull dispersion characteristics, and integrating feedback data closed-loop correction frequency selection and energy supply strategies, self-interference and thermal effects are effectively suppressed, ensuring both biosafety and communication reliability and spectral efficiency. Attached Figure Description

[0055] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 This is a flowchart of a data transmission method for an intracranial-extracranial wireless communication system as described in this embodiment.

[0057] Figure 2 This is a flowchart illustrating the identification of current fluctuation patterns in the embodiment.

[0058] Figure 3 This is a flowchart illustrating the construction process of the mapping relationship described in the embodiment. Detailed Implementation

[0059] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0060] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0061] Example, refer to Figure 1 , Figure 2 and Figure 3 This embodiment provides a data transmission method for a wireless communication system based on intracranial and extracranial structures, including:

[0062] S1. Based on the thickness and density of the skull, the carrier frequency is dynamically selected, and a dedicated communication frequency band is established between the external main control terminal and the intracranial implanted device.

[0063] Furthermore, before system deployment, multiple sets of ex vivo skull samples with different thicknesses and densities are collected. The samples retain the contents of the medullary cavity and the dura mater remnants attached to the surface, or are treated with artificial perfusion to simulate body fluid to maintain physiological hydration. For each sample, an electromagnetic wave transmission test platform is built in vitro, and continuous wave signals are emitted point by point within a preset frequency band while the power at the receiving end is recorded synchronously. The penetration loss at each frequency is calculated. At the same time, several test frequency points are extended beyond the upper and lower boundaries of the preset frequency band for supplementary measurements to cover the resonance valley displacements that occur in thick skulls or high-density individuals. Based on the penetration loss data of all test frequency points, a mapping relationship between penetration loss and frequency, skull thickness, and skull density is established, and the mapping relationship is stored in the form of a parametric model at the in vitro master control terminal.

[0064] The process of constructing the mapping relationship includes selecting multiple sets of ex vivo skull samples. For each sample, an in vitro transmission test platform is built with the transmitting antenna placed on one side of the sample and the receiving antenna placed on the other side. The transmitting frequency is scanned in the range of 402 MHz to 430 MHz with a step size of 1 MHz. The received power at each frequency is recorded and the penetration loss is calculated. With the penetration loss as the dependent variable and frequency, thickness, and density as independent variables, a nonlinear regression method is used to fit the physical model.

[0065] The reason for not directly applying the theoretical attenuation formula for free space or homogeneous media is that the porosity and water content distribution inside the ex vivo skull sample are uneven, resulting in a large residual in the theoretical model. Therefore, a weighted nonlinear least squares method is used for fitting because the penetration loss is high and the signal-to-noise ratio is low in the high-frequency band, and the variance of the measured data is higher than that in the low-frequency band. If equal weights are used for fitting, the model will overfit the low-frequency data and deviate from the true trend of the high frequency. By assigning lower weights to high-loss frequencies, the prediction error of the fitting results is more balanced across the entire frequency band.

[0066] The objective function for weighted nonlinear least squares fitting is expressed as follows:

[0067] ;

[0068] In the formula, This represents the weighted sum of squared residuals; Indicates the first Actual measured penetration loss; This represents the predicted value from the parametric model; Represents the vector of model parameters to be fitted; Indicates the first The weights of each data point are set to a value. ,in This serves as a reference loss constant, used to smooth weight transitions. This represents the total number of samples.

[0069] Parameterized model The specific form is determined by the following steps:

[0070] Step 1: Construct a candidate model library, including single exponential decay models. Polynomial model and log-saturated models ,in All are undetermined coefficients.

[0071] Step 2: For all The calibration data were used to initially fit each candidate model using the equal-weighted least squares method, and the model parameters for each group were obtained.

[0072] Step 3: Calculate the root mean square error (RMSE) and maximum absolute error (MAE) of each candidate model across all data points.

[0073] Step 4: If the RMSE and MAE of a candidate model are both less than the preset threshold (in this embodiment, the RMSE threshold is set to 1.5dB and the MAE threshold is set to 3.0dB), then the model enters the final candidate set.

[0074] Step 5: For the models in the final candidate set, use the weighted nonlinear least squares method to perform a second fitting, select the model with the smallest RMSE as the final parameterized model, and store the corresponding parameter vector in non-volatile memory.

[0075] In this embodiment, 30 groups of ex vivo skull samples (thickness range 5-12 mm, density range 1200-1900 kg / m³) were calibrated in the 402-430 MHz frequency band, and logarithmic saturation models in the candidate model library were used. After weighted quadratic fitting, the RMSE was 0.92dB and the MAE was 2.14dB, both of which were better than the threshold, and the RMSE was the smallest. Therefore, it was selected as the final parameterized model.

[0076] The fitted model parameters are stored in the non-volatile memory of the external master control unit for online frequency optimization during operation.

[0077] During the communication initialization phase, the external master control terminal sequentially sends multiple detection signals to the intracranial implanted device, each using a different candidate frequency. After receiving the detection signals, the intracranial implanted device measures the received power of each signal, encodes the received power sequence, and transmits it back to the external master control terminal via a reverse link. After receiving the feedback, the external master control terminal, in conjunction with the pre-stored mapping relationship, evaluates the penetration performance of each candidate frequency under the current individual skull conditions.

[0078] The reverse link backhaul during the communication initialization phase adopts the following mechanism: After the system is powered on, the implanted device defaults to a preset initialization fixed frequency. (In this embodiment, 402MHz, the lowest frequency in the MICS band, is used as the temporary carrier for the reverse link, as this frequency has the lowest penetration loss under all candidate skull conditions.) The return frame adopts a fixed-length short frame format, including a frame header (8-bit preamble + 4-bit frame type identifier), and a payload (the received power quantization value for each candidate frequency, each frequency point occupying 4 bits, for a total of...). (each frequency point) and frame tail (8-bit CRC check), the total frame length does not exceed [number of frequency points], and the frame tail (8-bit CRC check). bit; reverse data rate uses the lowest setting. To ensure sufficient link margin; after sending all detection signals, the external master control unit immediately switches to... The reflection monitoring mode receives and decodes the returned frame. If the CRC check fails, the probe sequence is retransmitted, with a maximum of 3 retries.

[0079] Select the frequency with the best penetration performance from the candidate frequency set as the initial carrier frequency.

[0080] Among them, penetration performance is inversely proportional to penetration loss, and penetration loss Based on the electromagnetic physical properties of the skull, the model is as follows:

[0081] ;

[0082] In the formula, This indicates the power loss of electromagnetic waves after they pass through the skull; Indicates the thickness of the skull; The skin depth is represented by the carrier frequency. With skull density The function; Indicates the carrier frequency; This represents skull density; the model does not employ the classic exponential decay form. This is because actual measurements found that when At that time, the multiple reflections and scattering caused by the multi-layered structure of the skull make the actual attenuation rate lower than the theoretical value of pure absorption, and the logarithmic form can better fit this saturation effect; at the same time, the "+1" term ensures that hour The physical rationale. Skin depth. The calculation expression is:

[0083] ;

[0084] In the formula, Indicates the permeability of free space; Indicates the density of the skull The relevant effective conductivity, obtained by fitting calibration data, reflects the contribution of bone mineral content to dielectric loss.

[0085] Among them, effective conductivity The specific fitting process is as follows: In the calibration stage, the complex permittivity of each isolated skull sample in the 402~430MHz frequency band was measured using the coaxial probe method. According to electromagnetic theory (in , (where is the vacuum dielectric constant), calculate the center frequency of each sample. Effective conductivity at density; For independent variable, Using quadratic polynomial regression as the dependent variable, we obtain the following result: ,in These are the fitting coefficients. In this embodiment, the coefficients were obtained by fitting 30 sets of samples. , , (Units are uniformly set to S / m), goodness of fit The fitted relationship is stored in the in vitro master control terminal and used for online calculation of skin depth at arbitrary densities.

[0086] When rapid online frequency optimization based on calibration data is required, penetration loss Calculated by the following formula:

[0087] ;

[0088] In the formula, This indicates the power loss of electromagnetic waves after they pass through the skull. Indicates skull thickness, Indicates skull density, Indicates the carrier frequency. These are dimensionless regression coefficients, obtained by fitting more than 30 sets of calibration data using the least squares method. Characterizing the coupling attenuation effect between thickness and frequency; Characterizing the coupling attenuation effect between density and frequency; The term representing the cross-effect of thickness and density is motivated by the following: transforming complex electromagnetic simulation or integral operations into multiply-accumulate operations that can be executed in real time by an embedded processor, reducing the time taken for a single frequency evaluation from milliseconds to microseconds while ensuring prediction accuracy (deviation from the physical model <1 dB), thus meeting the timing requirements for rapidly traversing candidate frequency points during the initialization phase.

[0089] During communication, the external master control terminal dynamically estimates the penetration loss at any candidate frequency by calling the mapping relationship based on the individual skull parameters acquired in real time. When the skull thickness h increases, the penetration loss L increases positively. When the carrier frequency f increases, the penetration loss L increases superlinearly. Based on this, the system avoids high-loss regions and prioritizes low-frequency points. When the skull density ρ decreases, the penetration loss L decreases accordingly.

[0090] It should be noted that by constructing a mapping relationship between penetration loss and individual anatomical parameters based on a large number of ex vivo skull samples before system deployment, and by combining measured feedback for online frequency optimization during the communication initialization phase, the system can adaptively select the optimal carrier frequency for different patients' skull characteristics, thereby improving the success rate of initial link establishment and communication efficiency, and avoiding severe signal attenuation or communication failure caused by blindly using fixed frequencies.

[0091] In particular, the frequency optimization mechanism is able to achieve individualized adaptation by transforming the electromagnetic response characteristics of the skull bio-medium from uncontrollable variables into modelable parameters. The mapping relationship established through in vitro calibration is essentially an engineering abstraction of the intrinsic relationship between the skull dielectric loss and structural features. This allows the system to invert individual penetration characteristics by relying solely on power feedback during the initialization phase, even in the absence of real-time imaging methods. This enables high-confidence initial frequency locking within the compliant frequency band, laying the physical layer foundation for subsequent communication.

[0092] S2. The external master control terminal performs frequency agility in the dedicated communication frequency band based on the real-time channel quality, and sends positive control commands to the intracranial implanted device through different carrier frequencies.

[0093] Furthermore, pilot symbols are inserted at fixed positions in each communication frame. The pilot symbols adopt a known modulation format and have constant power. The intracranial implanted device demodulates the pilot symbols, calculates the signal-to-noise ratio of the current channel, quantizes the signal-to-noise ratio into a finite state identifier, and embeds it into the payload of the feedback sensing data.

[0094] The real-time estimation and quantization process of the signal-to-noise ratio is as follows: During the pilot symbol reception period of each communication frame, the implanted device calculates the received signal power of the pilot symbol. With noise power .in It is obtained by calculating the peak square of the cross-correlation between the received samples of the known pilot sequence and the local reference sequence; The signal-to-noise ratio (SNR) is obtained by calculating the variance of the sampled values ​​within the guard interval before and after the pilot symbol. .

[0095] The quantization mapping rule for finite state identifiers is as follows: Non-uniform quantization is used, dividing the SNR into 8 states (3-bit encoding, state values ​​0-7). Specifically, the mapping is as follows: State 0 corresponds to... State 1 corresponds to State 2 corresponds to State 3 corresponds to State 4 corresponds to State 5 corresponds to State 6 corresponds to State 7 corresponds to The basis for this non-uniform partitioning is that small changes in the low SNR range have a significant impact on the frame error rate (requiring fine-grained resolution), while the link in the high SNR range is already sufficiently reliable (coarse-grained resolution is sufficient). The quantization mapping table is stored in the ROM of the implanted device in the form of a lookup table, and the mapping can be completed with only two comparisons during runtime.

[0096] When decoding the feedback sensor data, the external master control terminal extracts the status identifier and restores the current channel quality. If the channel quality is lower than the preset safety threshold, the frequency switching process is initiated. The frequency switching process includes querying the mapping relationship, excluding frequencies that have been used recently and whose channels are degraded, and selecting the frequency with the smallest predicted penetration loss value from the remaining candidate frequencies as the new carrier frequency.

[0097] The process for determining the safety threshold is as follows:

[0098] Step A: Before system deployment, the received signal-to-noise ratio of each candidate frequency was measured for multiple groups of ex vivo skull samples under different channel conditions (including different thicknesses and different simulated body fluid water content). and corresponding frame error rate .

[0099] Step B: Drawing - Curve, with target frame error rate (This embodiment is set as follows) Using this as a benchmark, find the corresponding critical signal-to-noise ratio. .

[0100] Step C: Consider additional disturbance margins in the actual in-vivo environment (In this embodiment, 3dB is used), the safety threshold is set to... .

[0101] Step D: The quantization is based on the boundary values ​​of a finite state identifier. For example, if the state identifier is a 3-bit code (0~7), then a state value ≤ 2 is considered to be below the safety threshold. This safety threshold is stored in the external master control unit and serves as the trigger criterion for frequency switching during operation.

[0102] The exclusion of degraded frequencies here is not a permanent removal, but rather the introduction of a hysteresis penalty factor. The motivation for this is that the degraded skull channel may be caused by transient physiological disturbances (such as swallowing or vascular pulsation). If the frequency is immediately and permanently disabled, the available spectrum resources will be quickly exhausted. By using a penalty factor, the degraded frequency is ranked lower in the short term, but the penalty decays over time, allowing it to be reselected after the channel recovers, thereby avoiding ping-pong switching and preserving spectrum flexibility.

[0103] Delayed penalty factor The calculation expression is:

[0104] ;

[0105] In the formula, Indicates the first Each degraded frequency point at time The penalty value; Indicates the initial penalty intensity; This indicates the time when the frequency point was marked as degraded; The time constant representing the penalty decay rate; Indicates the shortest duration of disabling; This represents an indicator function, which takes the value 1 if the condition within the parentheses is true, and 0 otherwise. After updating the communication parameters, the next frame of positive control commands is sent using the new carrier frequency.

[0106] The dedicated communication frequency band is divided into a forward subband and a reverse subband. The forward subband is used by the external master control terminal to send control commands to the intracranial implanted device. The reverse subband corresponds to the reflection path generated by the intracranial implanted device through load modulation. The center frequency of the forward subband is higher than that of the reverse subband, and a fixed protection interval is maintained between the two.

[0107] The width of the protective interval is determined based on the dispersion effect of the dielectric properties of the skull. Delayed dispersion with skull group The relational expression is:

[0108] ;

[0109] In the formula, This represents the minimum frequency interval between the forward subband and the reverse subband. This represents the equivalent noise bandwidth of the reverse link; This represents the maximum group delay dispersion caused by electromagnetic waves passing through the skull; The duration of the reverse symbol is indicated by the reverse data rate. Decision, satisfaction .

[0110] Group Delay Dispersion The physical origin of this is that the complex permittivity of the skull varies with frequency, resulting in different propagation speeds for different spectral components. Its calculation expression is:

[0111] ;

[0112] In the formula, The phase response per unit thickness of the skull is measured from calibration data; , Indicates the lower and upper limits of the operating frequency band; This represents the current skull thickness of the individual. This formula shows that the amount of dispersion is proportional to the thickness. Therefore, the protection interval needs to be dynamically adjusted according to the individual's skull thickness, rather than using a fixed value, to ensure that the forward and reverse subbands can still be effectively isolated even under the condition of the thickest skull.

[0113] Phase response The calibration process is as follows: In the ex vivo test, a vector network analyzer was used to measure the values ​​of each skull sample. Transmission parameters within the frequency band Extract its phase components After deducting the phase shift propagated in free space, the net phase response per unit thickness of the skull is obtained. ;right Perform a third-order polynomial fitting and then differentiate to obtain The parsing expression, take The maximum absolute value within is .

[0114] In this embodiment, a typical sample with a thickness of 8 mm and a density of 1600 kg / m³ was calibrated to obtain... .

[0115] The setting of the guard interval must be compatible with the specifications of the International Medical Implant Communication Service (MICS) frequency band, including the forward subband being limited to 410 MHz to 430 MHz and the reverse subband being limited to 402 MHz to 410 MHz within the 402 MHz to 430 MHz frequency band.

[0116] It should be noted that by embedding pilot symbols in each communication frame and using them to feedback the channel quality status, the external master control terminal can perceive the dynamic changes of the skull channel in real time and quickly switch to a frequency with better predictive performance when the channel deteriorates. This effectively avoids sudden fading caused by physiological activities or environmental disturbances and ensures the continuity and reliability of positive control command transmission.

[0117] In particular, the effectiveness of the frequency agility strategy depends on the coordinated design of pilot symbols and reverse links. It relies on the timing coupling of forward commands and reverse states within the same communication frame to achieve tight coupling of channel awareness and decision-making closed loop. Since changes in the skull channel mainly stem from slow physiological displacement or changes in tissue hydration state rather than fast fading, performing channel assessment and frequency switching at the frame level satisfies both response timeliness and avoids the control overhead caused by excessive switching, ensuring the feasibility of the system under limited computing resources.

[0118] S3. The intracranial implanted device receives and demodulates the carrier frequency, identifies the corresponding instruction content, and generates feedback sensing data.

[0119] Furthermore, a bandpass filter is set at the RF front end, with its passband center aligned with the current carrier frequency, to suppress out-of-band interference. The filtered signal is input to an envelope detector or coherent demodulation module and demodulated according to the modulation method adopted by the forward control command. The demodulated baseband signal is sent to a digital signal processor, where the original bit stream is recovered through matched filtering and symbol synchronization algorithms.

[0120] The symbol synchronization algorithm employs a two-stage synchronization scheme based on the Gardner error detector:

[0121] Initial Acquisition Phase: During the preamble period at the beginning of each frame, the implanted device samples the baseband signal at an oversampling rate of 8. Utilizing the known periodicity of the preamble (in this embodiment, the preamble is an alternating "10" sequence), the offset of the zero-crossing point between adjacent sampling points is calculated. By averaging these offsets, the sampling time is locked to ±1 / 8 of the symbol center within no more than 16 symbol periods. Within the range.

[0122] Tracking phase: Switch to 2x sampling rate and use Gardner timing error detector. Calculate the timing deviation, then pass it through a second-order loop filter (proportional branch gain). Integral branch gain This method drives a numerically controlled oscillator to fine-tune the sampling clock phase, enabling continuous tracking of slow clock drift. This scheme requires only 2 samplings and 3 multiplication operations per symbol, making it suitable for low-power DSP implementations in embedded devices.

[0123] The recovered bitstream is parsed according to a predefined instruction frame structure to extract the opcode and parameter fields.

[0124] The appropriate action is executed based on the opcode type: if it is a configuration instruction, the internal register or sensor sampling strategy is updated; if it is a query instruction, the sensing unit is triggered to collect physiological signals.

[0125] The sensing unit includes a temperature sensor, a pressure sensor, a pH detector, or a neural electrical activity recording electrode, whose sampling rate and resolution are dynamically set by command parameters; the acquired raw sensing data is quantized, encoded, and optionally compressed to form a feedback sensing data frame.

[0126] The specific process of lightweight compression is as follows:

[0127] Step a: Perform differential encoding on the raw sensor data, i.e. By utilizing the temporal redundancy of physiological signals changing slowly between adjacent sampling points, the dynamic range of the data is compressed.

[0128] Step b: For the difference sequence Perform adaptive piecewise linear quantization, quantization step size Based on the variance of the current data block Dynamic adjustment, calculation formula is: ,in As the reference step size, This is the reference variance.

[0129] Step c: Perform run-length encoding (RLE) on the quantized codewords, and merge consecutive identical codewords into (codeword, repetition count) pairs.

[0130] Step d: The compressed data is encapsulated into the payload field of the feedback sensing data frame. The frame header includes a compression flag to inform the external control unit of the decoding method. In this embodiment, the average compression ratio of the neural electrical activity recording data (sampling rate 1kS / s, 12-bit quantization) after the above compression is 2.8:1.

[0131] Embed check codes (such as CRC-8 or parity bits) in the feedback sensing data frames to support integrity verification by the external master control unit.

[0132] It should be noted that by employing bandpass filtering aligned with the current carrier frequency, a matched demodulation method, and a structured command parsing process, the intracranial device can accurately recover control commands under weak received signals and strong energy carrier interference. At the same time, the sensor acquisition is dynamically triggered according to the command type to generate feedback data frames with verification, laying a data foundation for subsequent highly reliable reverse communication.

[0133] In particular, the linkage architecture of command demodulation and sensor triggering deeply integrates communication and sensing functions into the same processing pipeline, avoiding hardware redundancy of independent control channels. By predefined opcode semantics, the implanted device can achieve on-demand acquisition with extremely low power consumption, which not only reduces the burden of invalid data generation and storage, but also ensures a high degree of consistency between feedback data and current control intent, thereby improving the task orientation and energy efficiency of the entire closed-loop system.

[0134] S4. The intracranial implantable device modulates the load using dynamic impedance mapping based on feedback sensing data and encodes the data into the reflection coefficient.

[0135] Furthermore, a programmable impedance network is set in the radio frequency front end of the intracranial implanted device. This network supports at least two discrete impedance states. The two impedance states correspond to binary logic 0 and 1, respectively. When the feedback sensing data to be transmitted is 0, the impedance network switches to the first impedance value. When the data is 1, it switches to the second impedance value. The impedance switching operation is synchronized with the energy transmission cycle, and one or more state switching is completed within each energy pulse window.

[0136] The adaptive decision-making rule for single and multiple state switching is as follows: Before the start of each energy pulse window, the implanted device checks the length of the queue of feedback data to be sent. and the current rectifier capacitor voltage .like (This embodiment) (indicating sufficient energy storage), and (This embodiment) If the pulse is in burst mode, it will continuously complete the maximum number of pulses within the current energy pulse window. Impedance switching of symbols, where The input pulse width is set to the specified value; otherwise, the system enters single-symbol mode, performing impedance switching for only one symbol within the current window to maximize energy margin. This decision logic is implemented using combinational logic circuitry with a delay of no more than one clock cycle.

[0137] The switching amplitude of the impedance network can be dynamically adjusted, including increasing the impedance difference in the high current stage to enhance the change in reflection coefficient, and decreasing the impedance difference in the low current stage to avoid nonlinear distortion.

[0138] The motivation for dynamic adjustment here is that there is an inherent contradiction between load modulation depth and rectification efficiency—an excessively large impedance difference, while enhancing the reflected signal, will reduce the coil Q value, causing the rectified voltage to drop or even interrupting the power supply; and during the low-current phase of the energy pulse, the rectification margin is already tight, and using a large impedance difference at this time can easily lead to undervoltage reset. Therefore, adaptively adjusting the impedance difference according to the instantaneous current intensity is a necessary means to maintain the stability of the energy link while ensuring communication detectability.

[0139] impedance difference The adaptive adjustment calculation expression is:

[0140] ;

[0141] In the formula, Indicates the current The actual impedance difference; Indicates the maximum permissible impedance difference; This represents the minimum current threshold required to maintain normal operation of the rectifier circuit. Indicates the rated operating current, when The impedance difference reaches its maximum value. This formula ensures that the modulation depth is automatically reduced when the current approaches a critical value, preventing the energy link from collapsing.

[0142] Maximum allowable impedance difference The determination process is as follows:

[0143] Step 1: Establish the rectified output voltage at the implanted device end. With load impedance and coil coupling coefficient The analytical model between them is expressed as follows: ,in For rectification efficiency, External emission power, and These are the equivalent resistance and reactance of the transmitter, respectively. This is the load reactance.

[0144] Step 2: Set the minimum voltage required for the rectifier circuit to operate normally. (Depending on the implanted chip process, this embodiment uses 1.8V).

[0145] Step ③: Under the worst coupling conditions ( ) below, order Solve upper limit .

[0146] Step 4: ,in This represents the optimal load impedance corresponding to the point of maximum power transmission. In this embodiment, , When the solution is obtained Therefore .

[0147] This value is stored in the non-volatile memory of the implanted device and serves as the upper limit of the impedance difference during operation.

[0148] impedance difference Dynamic adjustment is achieved through the following timing synchronization mechanism:

[0149] A peak hold circuit is connected in parallel at the output of the rectifier bridge of the implanted device, and its time constant is... satisfy ,in For the energy carrier cycle, The circuit operates on a reverse symbol period; it automatically captures and maintains the current envelope peak within each energy pulse window. .

[0150] At the beginning of each reverse symbol cycle (triggered by the zero-crossing detection circuit of the energy carrier), for Perform an ADC sampling to obtain the current value for the current symbol period. Substitution Formula calculation.

[0151] Based on the calculation The register values ​​of the programmable impedance network can be configured via the SPI interface, and the configuration completion time is no more than [time missing]. This ensures that the impedance state is stable before the effective modulation window of the current symbol period begins.

[0152] In this embodiment, (1MHz energy carrier). ,Pick This not only filters out carrier ripple but also preserves the trend of symbol-level current variation, ensuring... The accuracy and real-time performance of the calculations.

[0153] It should be noted that by synchronously performing programmable impedance switching during the energy transmission cycle and dynamically adjusting the impedance difference according to the current intensity, the detectability of the reflected signal is maximized, and nonlinear distortion in the modulation process is effectively suppressed, thus achieving efficient and robust reverse data loading under passive or extremely low power conditions.

[0154] In particular, the reason why dynamic impedance mapping can work stably under passive conditions is that its modulation action is strictly synchronized with the timing window of the external energy pulse, using the energy carrier itself as the modulation clock source, eliminating the dependence on the internal high-precision oscillator; at the same time, it adaptively adjusts the impedance switching amplitude according to the instantaneous current intensity, which essentially uses the dynamic range of the energy link as the adjustment lever for the modulation gain, thus avoiding the problems of resonance mismatch or sudden drop in rectification efficiency caused by impedance abrupt changes while ensuring signal detectability.

[0155] S5. The external main control terminal monitors the changes in the charging current in the coil circuit, identifies the corresponding current fluctuation pattern in the reflection coefficient, and decodes to obtain feedback sensing data.

[0156] Furthermore, a high-bandwidth current sensor is integrated into the energy emission coil drive circuit of the external main control terminal; the instantaneous value of the charging current is continuously acquired at a sampling rate of not less than 10 times the reverse data rate.

[0157] The analog front-end and ADC configuration for charging current acquisition are as follows: a dual-channel parallel acquisition architecture is adopted; the main channel uses a 12-bit ADC to acquire the full amplitude value of the raw charging current (range 0~500mA) at a sampling rate of 500kS / s for energy transfer closed-loop control; the auxiliary channel uses an analog high-pass filter (cutoff frequency) after the main channel signal passes through the filter. After removing the fundamental frequency and low-frequency ripple using a fourth-order Butterworth amplifier, the signal is then amplified by a programmable gain amplifier (PGA). Amplified, and then processed by a 16-bit SAR ADC. The sampling rate is used for acquisition, and the range corresponds to ±250μA of the original current. The output of the auxiliary channel ADC is the input sequence for subsequent sliding window differential processing.

[0158] The acquired current sequence is subjected to real-time sliding window differential processing to extract microsecond-level perturbation components.

[0159] The motivation for differential operation here is that the original charging current contains a large-amplitude energy transfer fundamental frequency and power supply ripple, with amplitudes reaching hundreds of mA, while the reflection disturbance caused by load modulation is only on the order of μA. Direct sampling would be overwhelmed by ADC quantization noise. Differential operation eliminates the slowly varying baseline, allowing weak AC disturbances to fall within the effective dynamic range of the ADC. The calculation expression for differential operation is:

[0160] ;

[0161] In the formula, This represents the disturbance signal after differentiation; Indicates the original sampled current; Represents the length of the sliding window, with values ​​satisfying... ( (where is the sampling rate) to ensure that the baseline estimate does not include modulation information within the symbol period.

[0162] Sliding window length The optimization process is as follows: First, determine the symbol period. The reverse data rate in this embodiment ,but Sampling rate Pick ;Require To ensure baseline smoothness, while Taking integer powers of 2 facilitates bitwise shift operations, therefore we take... The corresponding baseline estimation window duration It covers approximately 6.4 symbol periods, effectively smoothing the baseline without introducing significant delay.

[0163] The disturbance waveform is matched point by point with a pre-stored modulation template library, which contains typical current fluctuation patterns corresponding to 0 and 1. When the cross-correlation value between the template and the current disturbance waveform exceeds a preset judgment threshold, the corresponding bit is determined to be received. A majority voting mechanism is used to fuse the judgment results of multiple consecutive sampling windows and output the final decoded bit.

[0164] The simple hard decision majority voting method was not used here because the signal-to-noise ratio (SNR) varies significantly within different windows, and equal-weighted voting would cause erroneous decisions from low-quality windows to contaminate the final result. Therefore, confidence-weighted soft decision fusion was adopted, with the motivation of allowing decisions from high SNR windows to dominate and improving the overall decoding robustness.

[0165] Confidence-weighted bit decision The calculation expression is:

[0166] ;

[0167] In the formula, Indicates the decoded bits after fusion; Indicates the number of windows participating in the fusion; Indicates the first The confidence level of the decision for each window is given by the aforementioned formula. Calculated; Indicates the first The local hard decision result for each window; Binary values ​​are mapped to bipolar values ​​to achieve weighted accumulation.

[0168] Among them, decoding confidence Cross-correlation peak and background noise power The relational expression is:

[0169] ;

[0170] In the formula, Indicates the confidence level of the current bit decision; This represents the peak value of the normalized cross-correlation between the perturbation waveform and the pre-stored template. This represents the variance of the background noise of the charging current within the sliding window; It is an empirical weighting factor, dimensionless, used to balance correlation gain and noise suppression effect.

[0171] The process of building a pre-stored modulation template library is as follows:

[0172] Step 1: During the system calibration phase, the external master control terminal records the background current waveform in the energy transmission circuit in a silent state without reverse modulation, which serves as the noise substrate.

[0173] Step II: Control the implanted device to modulate the impedance states corresponding to logic 0 and logic 1 respectively, and synchronously collect the charging current waveforms under the corresponding states at the external end, collecting no less than 1000 symbol cycles for each state.

[0174] Step III: Perform mean removal and normalization processing on the acquired waveforms under each state, and then use the symbol period as the time factor. Align the window into segments.

[0175] Step IV: Take a point-by-point average of the segmented and aligned waveforms to obtain the logic 0 template. AND Logic 1 Template The length of each template is One sampling point.

[0176] Step V: Copy the template , It is stored in the SRAM of the external host device. In this embodiment, Each sampling point is stored in a 16-bit fixed-point format.

[0177] Normalized cross-correlation peak The specific calculation method is as follows: Let the disturbance signal sequence in the current sliding window be... The pre-saved template is ( Then, the normalized cross-correlation peak value is defined as:

[0178] ;

[0179] Pick The peak value of the cross-correlation in the current window is used for confidence calculation. This definition makes Dimensionless and its range of values ​​is strictly limited to This eliminates the impact of signal amplitude fluctuations on the absolute value of the correlation peak, ensuring confidence level. They are comparable under different received strengths.

[0180] The normalization process during the template construction phase also uses the same denominator basis, i.e. It has been divided by its own energy during storage. During runtime, only signal energy and dot product need to be calculated, reducing the amount of embedded computation.

[0181] Preset judgment threshold The adaptive setting process is as follows: During the protection time slot with no data transmission, the external main control terminal continuously collects the background current and calculates its relationship with the template. , The cross-correlation values ​​are taken as the statistical mean of the larger of the two values. and standard deviation Set threshold (i.e., the 3σ criterion) is used to ensure that the false alarm probability is below 0.13%. This threshold is automatically updated at the start of each communication session.

[0182] It should be noted that by sampling the charging current with high bandwidth and combining it with template matching and multi-window fusion decision mechanism, the reverse bit information can be accurately extracted from the weak reflection disturbances mixed in with high power energy signals, which improves the decoding accuracy and noise resistance of load modulation communication and solves the problems of traditional reverse links being susceptible to interference and having a high bit error rate.

[0183] In particular, the decoding method based on charging current disturbance is feasible because the change in reflected impedance caused by load modulation directly modulates the equivalent load of the transmitter, thereby generating a measurable current change in the drive circuit. By capturing this change at a high sampling rate and combining it with template matching, the reverse communication signal is essentially separated from the main energy path. No additional receiving antenna or RF front-end is required, which simplifies the external hardware structure. At the same time, the multi-window fusion decision effectively suppresses the influence of power supply ripple and environmental electromagnetic interference.

[0184] S6. The external master control unit corrects the carrier frequency selection and energy supply strategy for the next communication cycle based on the decoded feedback sensor data.

[0185] Furthermore, the feedback sensor data obtained from the current decoding, along with channel quality records from several past communication cycles, are input into the link control module. The link control module determines whether the current channel is in a state of continuous degradation. This determination is not made by directly comparing instantaneous values ​​with a threshold, because a single channel quality degradation may be caused by transient interference. Immediately triggering power increases or frequency switching would result in unnecessary energy waste and spectral jitter. Therefore, a first-order exponential smoothing algorithm is used to extract the trend of the channel quality sequence. The motivation is to filter out short-term fluctuations and retain long-term degradation trends, ensuring that control decisions only respond to actual channel deterioration. (Smoothed predicted values) The calculation expression is:

[0186] ;

[0187] In the formula, express The smoothed estimate of channel quality at time 1; express The measured channel quality indicator at any given time; This represents the smoothing factor, with a value range of 1. , The smaller the value, the greater the historical weight and the slower the trend response. It can be set according to the typical time constant of the skull channel changes. This represents the smoothed estimate from the previous time step; if Below the safety threshold and continuously If the energy transmission power is below the threshold for all cycles, it is considered a continuous degradation. In this case, the energy transmission power for the next cycle is increased, and the usage time of the current carrier frequency is shortened to accelerate the frequency switching response. If the determination result is negative, and the channel quality is consistently better than the safety threshold, the energy transmission power is reduced to decrease the thermal effect on biological tissues. The energy transmission power reduction strategy is as follows: when the channel quality is consistently better than the safety threshold for all cycles... During each cycle, the power decreases in fixed steps, each... Each cycle decreases The reduction will stop when either of the following conditions is met: (1) The current transmit power has been reduced to the minimum sustain power. (This embodiment) (1) Determined by the sensitivity of the rectifier circuit of the implanted device); (2) Channel quality smoothing estimate Falling back to a safe threshold No more than 2 state units (i.e.) At this point, the current power is retained and no further reduction is made to maintain link margin.

[0188] This decreasing strategy ensures that the power decrease rate is much slower than the increase rate (decreasing rate is about 1dB / 2s vs. increasing rate can reach several dB / 10ms), avoiding link interruptions due to sudden power drops; and extending the usage time of the current frequency to improve spectrum efficiency; the updated power supply parameters and frequency scheduling plan are written into the communication controller.

[0189] Smoothing factor The optimization process is as follows: Let the typical time constant of the skull channel quality variation be... The communication cycle is ,but and , The relationship is .

[0190] In this embodiment, Considering the fastest physiological disturbance (swallowing, ),but Considering slow changes (tissue hydration), ),but Actual amount taken As a compromise, it is possible to... Respond to real channel degradation within one communication cycle (i.e., 1 second).

[0191] Number of consecutive determination periods The settings are based on the following: Must meet ,in The maximum duration of a known transient disturbance (e.g., swallowing action, approximately 200-500 ms).

[0192] In this embodiment , ,but ,Pick .

[0193] The calculation process for the increase in energy transmission power is as follows: When it is determined to be a continuous degradation, the power increment is... ,in This is the proportional gain coefficient. To ensure thermal safety, a power upper limit is set. (Due to SAR limits, this embodiment) ), Enhanced transmission power .

[0194] The dynamic adjustment rule for the current carrier frequency usage duration is as follows: Let the default dwell time for the current frequency be... (In this embodiment, 500ms is used). When continuous degradation is detected, the dwell time for the next cycle is updated to... ,in To accelerate the switching factor (in this embodiment, we take...) ), This is set as the minimum dwell time limit (50ms in this embodiment) to prevent excessive handover; when the channel quality is consistently better than the safety threshold... One cycle (in this embodiment, we take) When ), the stay duration is updated to ,in This is set as the maximum dwell time limit (2000ms in this example) to improve spectrum efficiency and reduce handover overhead. After each adjustment... Write to the timer reload register of the communication controller; the changes take effect immediately.

[0195] It should be noted that by combining the decoded feedback data with the historical channel status for closed-loop control, the system can not only dynamically optimize the communication frequency of the next cycle, but also intelligently adjust the energy transmission power, minimizing the thermal load on biological tissues while ensuring communication robustness, thus achieving a synergistic improvement in safety, energy efficiency and communication performance.

[0196] In particular, the closed-loop correction capability of the link control module is based on short-term memory and prediction of channel degradation trends. It does not simply respond to a single quality decline, but judges whether the channel has entered a continuously deteriorating range by the state sequence within the sliding window. The design enables the system to tolerate sudden interference and make forward-looking adjustments to the actual channel degradation. While ensuring communication continuity, it achieves a dynamic balance between thermal safety and spectral efficiency through joint optimization of power and frequency.

[0197] In summary, this invention dynamically selects carrier frequencies by combining individual skull thickness and density characteristics, and achieves frequency agility based on real-time channel quality within a dedicated communication band, thereby improving the penetration efficiency and link robustness of intracranial and extracranial wireless communication. Simultaneously, by employing dynamic impedance mapping load modulation and a high-precision current fluctuation decoding mechanism, reliable reverse data transmission is achieved with extremely low power consumption. By rationally isolating the forward and reverse subbands according to skull dispersion characteristics and integrating feedback data closed-loop correction frequency selection and energy supply strategies, self-interference and thermal effects are effectively suppressed, ensuring both biosafety and communication reliability and spectral efficiency.

[0198] 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, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A data transmission method based on intracranial and extracranial wireless communication systems, characterized in that: include: Based on the thickness and density of the skull, the carrier frequency is dynamically selected and a dedicated communication frequency band is established between the external main control terminal and the intracranial implanted device. The external master control terminal performs frequency agility within the dedicated communication frequency band based on real-time channel quality, and sends positive control commands to the intracranial implanted device through different carrier frequencies; The intracranial implanted device receives and demodulates the carrier frequency, identifies the corresponding command content, and generates feedback sensing data. The intracranial implantable device modulates the load using dynamic impedance mapping based on feedback sensing data and encodes the data into the reflection coefficient; The external main control unit monitors the changes in the charging current in the coil circuit, identifies the corresponding current fluctuation pattern in the reflection coefficient, and decodes the feedback sensing data. The external master control unit corrects the carrier frequency selection and power supply strategy for the next communication cycle based on the decoded feedback sensor data.

2. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 1, characterized in that: The specific steps for dynamically selecting the carrier frequency point based on skull thickness and density are as follows: Before system deployment, multiple sets of ex vivo skull samples with different thicknesses and densities were collected. For each sample, an electromagnetic wave transmission test platform was built in vitro. Continuous wave signals were emitted point by point within the preset frequency band and the power at the receiving end was recorded synchronously. The penetration loss at each frequency was calculated. At the same time, several test frequency points were extended outside the upper and lower boundaries of the preset frequency band for supplementary measurements to cover the resonance valley displacement that occurs in individuals with thick skulls or high density. Based on the penetration loss data of all test frequencies, a mapping relationship between penetration loss and frequency, skull thickness, and skull density is established, and the mapping relationship is stored in the external master control terminal in the form of a parameterized model. During the communication initialization phase, the external master control terminal sequentially sends multiple detection signals to the intracranial implanted device, with each detection signal using a different candidate frequency. After receiving the detection signal, the intracranial implanted device measures the received power of each signal, encodes the received power sequence, and transmits it back to the external master control terminal via the reverse link. After receiving the feedback, the external master control terminal evaluates the penetration performance of each candidate frequency under the current individual skull conditions by combining it with the pre-stored mapping relationship. The frequency with the best penetration performance is selected from the candidate frequency set as the initial carrier frequency, where the penetration performance is inversely proportional to the penetration loss.

3. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 2, characterized in that: The external master control terminal performs frequency agility within the dedicated communication frequency band based on real-time channel quality. The specific steps are as follows: Pilot symbols are inserted at fixed positions in each communication frame. The pilot symbols use a known modulation format and have constant power. The intracranial implanted device demodulates the pilot symbols, calculates the signal-to-noise ratio of the current channel, quantizes the signal-to-noise ratio into a finite state identifier, and embeds it into the payload of the feedback sensing data. When decoding feedback sensor data, the external master control terminal extracts the status identifier and restores the current channel quality. If the channel quality is lower than the preset safety threshold, the frequency switching process is initiated. The frequency switching process includes querying the mapping relationship, excluding frequencies that have been used recently and whose channels are degraded, and selecting the frequency with the smallest predicted penetration loss value from the remaining candidate frequencies as the new carrier frequency. After updating the communication parameters, the next frame of positive control command is sent using the new carrier frequency.

4. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 3, characterized in that: The intracranial implantable device modulates the load using dynamic impedance mapping based on feedback sensor data. The specific steps are as follows: A programmable impedance network is set at the radio frequency front end of the intracranial implantable device, which supports at least two discrete impedance states. The two impedance states correspond to binary logic 0 and 1, respectively; When the feedback sensor data to be sent is 0, the impedance network switches to the first impedance value; when the data is 1, it switches to the second impedance value. Impedance switching operation is synchronized with the energy transfer cycle, and one or more state switching is completed within each energy pulse window; The switching amplitude of the impedance network can be dynamically adjusted, including increasing the impedance difference in the high current stage to enhance the change in reflection coefficient, and decreasing the impedance difference in the low current stage to avoid nonlinear distortion.

5. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 4, characterized in that: The external main control terminal identifies the current fluctuation pattern corresponding to the reflection coefficient by monitoring the changes in the charging current in the coil circuit. The specific steps are as follows: A high-bandwidth current sensor is integrated into the energy emission coil drive circuit of the external main control terminal; The instantaneous value of charging current is continuously acquired at a sampling rate of not less than 10 times the reverse data rate; The acquired current sequence is subjected to real-time sliding window differential processing to extract microsecond-level perturbation components; The disturbance waveform is matched point by point with a pre-stored modulation template library, which contains typical current fluctuation patterns corresponding to 0 and 1. When the cross-correlation value between the template and the current disturbance waveform exceeds the preset judgment threshold, it is determined that the corresponding bit has been received; A majority voting mechanism is used to fuse the decision results of multiple consecutive sampling windows and output the final decoded bits.

6. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 5, characterized in that: The external master control terminal corrects the carrier frequency selection and power supply strategy for the next communication cycle based on the decoded feedback sensor data. The specific steps are as follows: The feedback sensor data obtained from the current decoding, together with the channel quality records from the past several communication cycles, are input into the link control module. The link control module determines whether the current channel is in a state of continuous degradation. If the determination result is yes, the energy transmission power of the next cycle is increased and the usage time of the current carrier frequency is shortened to accelerate the frequency switching response speed. If the determination result is no, and the channel quality is stable and better than the safety threshold, the energy transmission power is reduced to reduce the thermal effect of biological tissues and the usage time of the current frequency is extended to improve the spectrum efficiency. The updated energy supply parameters and frequency scheduling plan are written into the communication controller.

7. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 6, characterized in that: The dedicated communication frequency band is divided into a forward sub-band and a reverse sub-band, including: The forward subband is used to send control commands from the external master control terminal to the intracranial implanted device; the reverse subband corresponds to the reflection path generated by the intracranial implanted device through load modulation; the center frequency of the forward subband is higher than that of the reverse subband, and a fixed guard interval is maintained between the two, wherein the width of the guard interval is determined according to the dispersion effect of the dielectric properties of the skull.

8. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 7, characterized in that: The setting of the protection interval must be compatible with the specifications of the International Medical Implant Communication Service (MICS) frequency band, including the forward subband being limited to 410MHz to 430MHz and the reverse subband being limited to 402MHz to 410MHz within the 402MHz to 430MHz frequency band.

9. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 8, characterized in that: The process of constructing the mapping relationship includes the following steps: Multiple sets of ex vivo skull samples were selected. For each sample, an in vitro transmission test platform was built, with the transmitting antenna placed on one side of the sample and the receiving antenna placed on the other side. Within the range of 402MHz to 430MHz, the transmission frequency is scanned in 1MHz increments, the received power at each frequency is recorded, and the penetration loss is calculated. With penetration loss as the dependent variable and frequency, thickness, and density as independent variables, a nonlinear regression method was used to fit the physical model. The fitted model parameters are stored in the non-volatile memory of the external master control unit for online frequency optimization during operation.

10. The data transmission method of the intracranial and extracranial wireless communication system as described in claim 9, characterized in that: During actual communication, the external master control terminal dynamically estimates the penetration loss at any candidate frequency by calling the mapping relationship based on the individual skull parameters acquired in real time. When the skull thickness h increases, the penetration loss L increases positively; when the carrier frequency f increases, the penetration loss L increases superlinearly; the system avoids high-loss regions and preferentially selects low-frequency points accordingly; when the skull density ρ decreases, the penetration loss L decreases accordingly.

Images