Frequency band adaptive switching method and device based on star flash wireless microphone

Abstract

The invention relates to a frequency band adaptive switching method and device based on a star flash wireless microphone. The method comprises the following steps: extracting a spectrum occupancy rate feature, a signal-to-noise ratio feature and a multipath fading feature from a voice frame of a target microphone through short-time Fourier transform and a neural network; calculating a frequency band quality score based on the spectrum occupancy rate feature, the signal-to-noise ratio feature and the multipath fading feature; the Dueling DQN is trained on the basis of the state space and the reward function; controlling the physical layer of the target microphone to be seamlessly switched to the optimal frequency band through the trained Dueling DQN; estimating a real-time packet loss rate through a grouping spraying method based on channel response; on the basis of the real-time packet loss rate, forward error correction and coding are carried out on the voice frames according to grades through Polar codes; and the voice frame in the switching process is recovered through the satellite flash broadcast frame and the SC-List decoder. According to the method, dynamic balance of transmission fault-tolerant capability and efficiency is realized through a scoring function, reinforcement learning and dynamic coding rate adjustment of driving Polar codes.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless microphones, and particularly relates to a star flash-based wireless microphone frequency band adaptive switching method and device. BACKGROUND

[0002] Currently, wireless microphones have problems such as spectrum congestion, signal-to-noise ratio fluctuation and multipath fading in the communication process. The existing technology mainly adopts the following scheme: first, a traditional spectrum analysis method, such as a channel scanning based on a fixed threshold, is used to evaluate the occupancy rate and signal-to-noise ratio of the 2.4GHz / 5.8GHz frequency band to select a better frequency band; second, a static or simple rule-based frequency band switching strategy, such as a switching mechanism based on signal strength, is used; and finally, a forward error correction (FEC) coding with a fixed redundancy, such as an RS code or an LDPC code, is used to reduce data loss.

[0003] However, these schemes have significant defects: the traditional spectrum analysis method lacks real-time modeling of dynamic multipath fading characteristics, and the frequency band quality evaluation accuracy is low; the static or simple rule-based switching strategy cannot adapt to complex network environments, and the switching delay is high (usually more than 100 microseconds), affecting real-time communication; and the FEC coding with a fixed redundancy cannot be dynamically adjusted according to the actual packet loss, and the redundancy efficiency is low or the error correction capability is insufficient. SUMMARY

[0004] To solve the problems of wireless microphones in spectrum congestion, signal-to-noise ratio fluctuation and multipath fading scenarios, and to improve the frequency switching stability and signal quality of wireless microphones, in a first aspect of the application, a star flash-based wireless microphone frequency band adaptive switching method is provided, comprising:

[0005] extracting spectrum occupancy rate features, signal-to-noise ratio features and multipath fading features from the voice frames of the target microphone through short-time Fourier transform and neural networks; calculating a frequency band quality score based on the spectrum occupancy rate features, the signal-to-noise ratio features and the multipath fading features; determining a state space and a reward function of the Dueling DQN based on the frequency band quality score; training the Dueling DQN based on the state space and the reward function; controlling the physical layer of the target microphone to seamlessly switch to the optimal frequency band through the trained Dueling DQN; estimating the real-time packet loss rate through a channel response-based packet spraying method; performing forward error correction and encoding on the voice frames according to the level through the Polar code based on the real-time packet loss rate; and recovering the voice frames in the switching process through the star flash broadcast frame and the SC-List decoder.

[0006] In some embodiments of the present application, the calculating the frequency band quality score based on the spectrum occupancy feature, the signal-to-noise ratio feature and the multipath fading feature comprises: determining the weights of the spectrum occupancy feature, the signal-to-noise ratio feature and the multipath fading feature based on the application scenario of the target microphone; determining a frequency band score function based on the weights of the spectrum occupancy feature, the signal-to-noise ratio feature and the multipath fading feature; and calculating the frequency band quality score based on the real-time spectrum occupancy feature, the signal-to-noise ratio feature and the multipath fading feature.

[0007] In some embodiments of the present application, the determining the state space and the reward function of the Dueling DQN based on the frequency band quality score comprises: determining the state space of the Dueling DQN based on the frequency band quality score, the number of handovers and the number of locally detected interference events; and determining the reward function based on the frequency band quality score, the number of handovers and the handover time consumption. The determining the state space and the reward function of the Dueling DQN based on the frequency band quality score comprises: determining the state space of the Dueling DQN based on the frequency band quality score, the number of handovers and the number of locally detected interference events; and determining the reward function based on the frequency band quality score, the number of handovers and the handover time consumption.

[0008] In some embodiments of the present application, the estimating the real-time packet loss rate by the channel response based packet spraying method comprises: receiving an acknowledgement feedback within a preset sliding window size, or broadcasting an acknowledgement frame by SLB extension; counting the packet loss rate of the voice frame within a plurality of preset sliding windows based on the acknowledgement feedback or the acknowledgement frame; and smoothing the packet loss rate.

[0009] Further, the recovering the voice frame in the handover process by the star flash broadcast frame and the SC-List decoder comprises: sending the SLB extension broadcast frame, distinguishing the control and data channels by the TCI protocol; and recovering the voice frame in the handover process by the SC-List decoder.

[0010] In the above-mentioned embodiments, the extracting the spectrum occupancy feature, the signal-to-noise ratio feature and the multipath fading feature from the voice frame of the target microphone by the short-time Fourier transform and the neural network comprises: determining the window, the frame length and the frame shift of the short-time Fourier transform; mapping the voice frame of the target microphone to a frequency domain representation by the short-time Fourier transform based on the window, the frame length and the frame shift; and extracting the spectrum occupancy feature, the signal-to-noise ratio feature and the multipath fading feature from the frequency domain representation by the neural network.

[0011] In a second aspect of the present application, a star flash wireless microphone frequency band adaptive switching device is provided, comprising:

[0012] The calculation module is used to extract spectral occupancy features, signal-to-noise ratio features, and multipath fading features from the speech frames of the target microphone through short-time Fourier transform and neural networks; and to calculate the frequency band quality score based on the spectral occupancy features, signal-to-noise ratio features, and multipath fading features.

[0013] The training module is used to determine the state space and reward function of Dueling DQN based on the frequency band quality score; and to train Dueling DQN based on the state space and the reward function.

[0014] The control module is used to control the physical layer of the target microphone to seamlessly switch to the optimal frequency band through the trained Dueling DQN;

[0015] The recovery module is used to estimate the real-time packet loss rate using a channel response-based packet spraying method; based on the real-time packet loss rate, it performs forward error correction and encoding on the voice frames according to the level using Polar codes; and it recovers the voice frames during the handover process using star-bomb broadcast frames and SC-List decoders.

[0016] A third aspect of the present invention provides an electronic device, comprising: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the adaptive switching method for frequency bands of a star-flash wireless microphone provided in the first aspect of the present invention.

[0017] In a fourth aspect, the present invention provides a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the adaptive switching method for frequency bands of a star-flash wireless microphone provided in the first aspect of the present invention.

[0018] The beneficial effects of this invention are:

[0019] This invention supports interpretable and quantifiable switching criteria by using a frequency band scoring function to achieve adaptive frequency band decision-making and integration with the SLB physical layer via the DRL module, thereby improving synchronous switching and low-latency reconfiguration. Polar coding is seamlessly integrated with the SLB broadcast mechanism without the need to establish a connection. SC-LIST decoding has high fault tolerance and is suitable for microsecond-level error recovery. Attached Figure Description

[0020] Figure 1 This is a basic flowchart illustrating the adaptive switching method for frequency bands of a star-flash wireless microphone according to some embodiments of the present invention;

[0021] Figure 2 This is a schematic diagram of the specific process of the adaptive switching method for frequency bands of a star-flash wireless microphone in some embodiments of the present invention;

[0022] Figure 3 A structural schematic diagram of a star flash wireless microphone frequency band adaptive switching device in some embodiments of the present application;

[0023] Figure 4 A structural schematic diagram of an electronic device in some embodiments of the present application. DETAILED DESCRIPTION

[0024] The principles and features of the present application are described below in conjunction with the accompanying drawings, and the examples are only used to explain the present application and not to limit the scope of the present application.

[0025] Reference Figure 1 With Figure 2 In a first aspect of the present application, a star flash wireless microphone frequency band adaptive switching method is provided, comprising:

[0026] S100. Extracting spectral occupancy features, signal-to-noise ratio features and multipath fading features from the speech frames of the target microphone through short-time Fourier transform and neural network; calculating the frequency band quality score based on the spectral occupancy features, the signal-to-noise ratio features and the multipath fading features;

[0027] S200. Determining the state space and the reward function of the Dueling DQN based on the frequency band quality score; training the Dueling DQN based on the state space and the reward function;

[0028] S300. Controlling the physical layer seamless switching of the target microphone to the optimal frequency band through the trained Dueling DQN;

[0029] S400. Estimating the real-time packet loss rate through the channel response-based packet spraying method; based on the real-time packet loss rate, performing forward error correction and encoding on the speech frames according to the level through the Polar code; recovering the speech frames in the switching process through the star flash broadcast frame and the SC-List decoder.

[0030] In step S100 of some embodiments of the present application, spectral occupancy features, signal-to-noise ratio features and multipath fading features are extracted from the speech frames of the target microphone through short-time Fourier transform and neural network.

[0031] Specifically, the window, frame length, and frame shift of the Short-Time Fourier Transform (SFT) are determined. Based on these parameters, the speech frames of the target microphone are mapped to a frequency domain representation using the SFT. A neural network is then used to extract spectral occupancy features, signal-to-noise ratio (SNR) features, and multipath fading features from the frequency domain representation. For example, 128 sampling points with a 75% overlap are used in the frequency domain to obtain a frequency map. In the energy dimension, the spectrum map is input into the gSNR estimation network. Occupancy is determined using sliding window averaging and binarized spectrum to assess the activity of the current frequency band. Stability features are characterized by the envelope consistency coefficient or delay spread function, and multipath fading features are also considered. Without loss of generality, the feature vector is input as a subband energy tensor of the form [B, C, T, F], where B = batch, C = number of channels, T = number of frames, and F = frequency bins. A neural network is used to encode and extract the feature space.

[0032] In step S100 of some embodiments of the present invention, calculating the frequency band quality score based on spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics includes:

[0033] S101. Based on the application scenario of the target microphone, determine the weights of spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics;

[0034] S102. Determine the frequency band scoring function based on the weights of spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics;

[0035] S103. Calculate the frequency band quality score based on real-time spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics.

[0036] Specifically, the quality scoring function is expressed as:

[0037] Q(f i )=α·gSNR(f i )+β·(1-Occ(f i ))+γ·Stab(f i ),

[0038] Where gSNR represents the signal-to-noise ratio feature, normalized to [0,1]; O(f i S(f) represents the spectrum occupancy estimate, where 1 indicates a long-term active frequency band and 0 indicates idle frequency; i The value ) represents the stability score of multipath fading. α, β, and γ represent the weights of the signal-to-noise ratio feature, spectrum occupancy feature, and multipath fading feature, respectively, with reference values ​​of 0.5, 0.3, and 0.2.

[0039] In step S200 of some embodiments of the present invention, determining the state space and reward function of Dueling DQN based on the frequency band quality score includes: determining the state space of Dueling DQN based on the frequency band quality score, the number of handovers and the number of locally detected interference events; and determining the reward function based on the frequency band quality score, the number of handovers and the handover time.

[0040] The state space and reward function of Dueling DQN are determined based on the frequency band quality score, the number of handovers, and the number of locally detected interference events. The state space of Dueling DQN is determined based on the frequency band quality score, the number of handovers, and the number of handover times. The reward function is determined based on the frequency band quality score, the number of handovers, and the handover time.

[0041] Specifically, the system updates its state once per time slot (e.g., 20 μs), including:

[0042] The state vector is represented as:

[0043] S t ={Q(f1), Q(f2), ..., Q(f n ), H(f i ), B t L t D t},

[0044] SNR t (f i ) represents the signal-to-noise ratio of the frequency band, I t (f i ) indicates frequency band f i Interference power, D t L represents the number of frequency band switchings within the most recent 1ms. t Indicates the remaining time of the delayed budget, B t Indicates the buffer fill level.

[0045] Action space definition:

[0046] A={a i |Select frequency band f i Or maintain the current frequency band}, i = 1, 2, ..., n,

[0047] Where a0 indicates maintaining the current frequency band, a i Indicates switching to frequency band f i .

[0048] To maximize band quality and minimize the risk of handover delay disruptions, the reward function is set as follows:

[0049]

[0050] Wherein, Δ represents the switching time (less than 20 μs), α, β, γ, λ are 1.0, 0.5, 0.1, 2.0 respectively.

[0051] The Dueling structure is used to reinforce the state value estimation, and the Dueling network includes:

[0052] Shared convolutional layer: abstract the spectrum state; Branch 1 (Value Stream): FC (convolutional layer) → 64 down-sampling → output state value V(s t ); Branch 2 (Advantage Stream): FC (convolutional layer) → 64 down-sampling |A|(output each action advantage A(s t , a t );

[0053] Output Q value:

[0054]

[0055] Wherein, V(s t ) represents the state value function; s t , a t represent state and action respectively.

[0056] Wherein, represents using average instead of maximizing operation to ensure the stability of the Q value function.

[0057] In order to further optimize the switching strategy, the target objective function is defined as:

[0058]

[0059] Wherein, the policy π defines the probability distribution of taking action a t in state s t . Constraint: switching delay limit: T switch ≤ 20 μs; audio non-terminal constraint: b t ≥ b min = 5%; broadcast band synchronization constraint: broadcast cycle = 5 ms, ≥ 3 recommended frequency bands per broadcast; energy consumption constraint: E switch ≤ 20 μJ.

[0060] The reward function is represented as:

[0061]

[0062] Wherein, α, β and γ are hyperparameters, used to adjust the weight of different terms; Q(fa t ) represents the Q value of switching frequency band action a t corresponding to frequency band f; H(fa t) represents the frequency band f corresponding to the switching frequency band action a t The penalty term is, for example, the frequency band switching failure; Δt represents the change amount of the state.

[0063] In step S300 of some embodiments of the application, the physical layer seamless switching of the target microphone to the optimal frequency band is controlled by the Dueling DQN trained.

[0064] Specifically, the action a t is evaluated every 20μs window t If a t =a0, the current frequency band is maintained; if a i =a t , the frequency band switching is initiated, and it is checked whether the SLB broadcast synchronization window is valid. The broadcast mechanism is as follows: each G node publishes an extended broadcast message every 5ms, and the content (field) includes:

[0065] Message type, for example, extended broadcast is 0x0004; TCID, recommended frequency band list, start offset time, switching confirmation flag; that is, the whole process is: the wireless microphone receives audio, the self-frequency band score calculator calculates the frequency band quality function value, the Dueling DQN decision outputs the switching action a t , the SLB broadcast receives / acknowledges the message with the frequency band ID, and then the frequency band switching scheduler completes the frequency band reconfiguration within 10μs.

[0066] In step S400 of some embodiments of the application, the real-time packet loss rate estimated by the channel response-based packet spraying method includes:

[0067] S401. Receiving the acknowledgement feedback within a preset sliding window size, or through the SLB extended broadcast acknowledgement frame;

[0068] S402. Based on the acknowledgement feedback or the acknowledgement frame, the packet loss rate of the voice frame within a plurality of preset sliding windows is counted; the packet loss rate is smoothed.

[0069] Specifically, the input includes: whether the acknowledgement feedback (ACK) or the SLB extended broadcast acknowledgement frame is received within the sliding window;

[0070] Counting the number of lost frames / total frames p t : the smoothed packet loss rate estimate (EMA):

[0071]

[0072] In step S400 of some embodiments of the present application, the speech frame is forward error corrected and encoded according to the level by the Polar code based on the real-time packet loss rate; the speech frame in the switching process is recovered by the star flash broadcast frame and the SC-List decoder.

[0073] Specifically, assuming that the speech frame block length k = 256, the output Polar encoding parameter (n, k) satisfies:

[0074]

[0075] Wherein, δ is the redundancy safety margin, usually set to 0.05 (5%); for example, if (n = 308, k = 256) is selected, and the redundancy is 20%; if (n = 340, k = 256) is selected.

[0076] The Polar code uses the Arikan Polar construction method to construct a bit reliability sequence; its channel selection strategy: Gaussian Approximation (GA), select the k strongest channels in the reliable sequence U to construct the information bit; the rest are frozen bits, filled with 0 or CRC.

[0077] In step S400 of some embodiments of the present application, the speech frame in the switching process is recovered by the star flash broadcast frame and the SC-List decoder, which includes: sending by the SLB extended broadcast frame, distinguishing the control and data channels by the TCI protocol; recovering the speech frame in the switching process by the SC-List decoder.

[0078] Specifically, the TCID extension mark: TCID field = 0x05: indicates "high priority speech redundancy stream"; the SLB configuration is as follows: message type 0x0005; encoding type Polar (n, k); sending period 5ms; corresponding original speech frame TCID 0x01.

[0079] The receiving end supports the soft input soft output (SISO) structure: receiving the speech packet {x1,..., x n}+LLR (log likelihood ratio); combined with CRC check, the List decoder (L = 8-16) is constructed.

[0080] In particular, in order to reduce the computing power requirement of the Polar code, the G node broadcasts every 1ms: the currently available frequency band Favail; the switching priority list (including TCID / T node ID); the T node competes for the priority bit according to the broadcast, and executes the auction type switching permission; the SLB "frequency band protection window" is introduced: after a frequency band switching, other T nodes cannot occupy the frequency band within 50μs;

[0081]

[0082] where f ** and f * curr represent the optimized frequency and the existing frequency, respectively;

[0083] Preemption control logic: each node Ti generates a switching request vector:

[0084]

[0085] The G node executes the scheduling function:

[0086] Priority i = λ1Q(f i ) + λ2(1-B t ) + λ3· historical failure rate;

[0087] Finally broadcast to all T nodes: Tmax= 2 nodes can switch at the same time; TCID separates different broadcast streams to avoid interference.

[0088] It can be understood that, through the lightweight packet loss perception system, the dynamic coding rate adjustment of the Polar code is driven, and the dynamic balance of the fault tolerance capability and the efficiency of the voice transmission is realized. The design of the embedded star flash SLB expansion frame mechanism enables the rapid deployment of the SLB broadcast scene.

[0089] Embodiment 2

[0090] With reference Figure 3 , the second aspect of the present application provides a star flash wireless microphone frequency band adaptive switching device 1, comprising:

[0091] The computing module 11 is configured to extract spectral occupancy features, signal-to-noise ratio features, and multipath fading features from the speech frames of the target microphone through short-time Fourier transform and neural networks; and calculate a frequency band quality score based on the spectral occupancy features, the signal-to-noise ratio features, and the multipath fading features.

[0092] The training module 12 is configured to determine a state space and a reward function of the Dueling DQN based on the frequency band quality score; and train the Dueling DQN based on the state space and the reward function.

[0093] The control module 13 is configured to control the physical layer seamless switching of the target microphone to the optimal frequency band through the trained Dueling DQN.

[0094] The recovery module 14 is configured to estimate a real-time packet loss rate by a channel response based packet spraying method; based on the real-time packet loss rate, forward error correction and coding of the speech frame by the Polar code according to the level; and recovery of the speech frame in the switching process by the star flash broadcast frame and the SC-List decoder.

[0095] Further, the calculation module 11 comprises: a first determination unit configured to determine the weights of the spectrum occupancy rate feature, the signal-to-noise ratio feature and the multipath fading feature based on the application scenario of the target microphone;

[0096] A second determination unit is configured to determine the frequency band scoring function based on the weights of the spectrum occupancy rate feature, the signal-to-noise ratio feature and the multipath fading feature; and a calculation unit is configured to determine the frequency band scoring function based on the weights of the spectrum occupancy rate feature, the signal-to-noise ratio feature and the multipath fading feature.

[0097] Embodiment 3

[0098] Reference Figure 4 In a third aspect, the present application provides an electronic device, comprising: one or more processors; a storage device configured to store one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the star flash wireless microphone frequency band adaptive switching method in the first aspect of the present application.

[0099] The electronic device 500 can include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 501 that can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 502 or loaded into a random access memory (RAM) 503 from a storage device 508. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0100] In general, the following devices can be connected to the I / O interface 505: input devices 506 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 508 including, for example, a hard disk, etc.; and communication devices 509. The communication devices 509 can allow the electronic device 500 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 4 The electronic device 500 is shown with various devices, but it should be understood that all of the shown devices are not required to be implemented or present. More or fewer devices can alternatively be implemented or present. Figure 4Each block in the flow diagrams of FIGS. 1-3 can represent a module, segment, or portion of code, which comprises one or more instructions that implement the specified logical functions. It should also be noted that each block of the flow diagrams and combinations thereof can represent a module, segment, or portion of code, which comprises one or more instructions that implement the specified logical functions.

[0101] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for performing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via the communication device 509, or installed from the storage device 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the methods of embodiments of the present disclosure are performed. It should be noted that the computer readable medium described in embodiments of the present disclosure can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In embodiments of the present disclosure, the computer readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device. In embodiments of the present disclosure, the computer readable signal medium can include a data signal carried in a baseband or as part of a carrier wave, in which the computer readable program code is carried. Such a propagated data signal can take any of a variety of forms, including but not limited to electro-magnetic, optical, or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code contained in the computer readable medium can be transmitted by any suitable medium, including but not limited to wire, cable, RF, etc., or any suitable combination of the above.

[0102] The computer readable medium described above can be included in the electronic device described above; or can exist separately from the electronic device and be accessed via the electronic device. The computer readable medium carries one or more computer programs that, when executed by the electronic device, cause the electronic device to:

[0103] Computer program code for carrying out operations of embodiments of the present disclosure can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++, Python, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0104] The computer program instructions can also be loaded onto a computer or other programmable information processing apparatus to cause a series of operations to be performed on the computer or other programmable information processing apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable information processing apparatus implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0105] The above descriptions are only the preferred embodiment of the application, not intended to limit the application and any modification, equivalent replacement and improvement made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A method for adaptive frequency band switching based on a star-flash wireless microphone, characterized in that, include: By using short-time Fourier transform and neural networks, spectral occupancy features, signal-to-noise ratio features, and multipath fading features are extracted from the speech frames of the target microphone. Based on spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics, the frequency band quality score is calculated. Based on the frequency band quality score, determine the state space and reward function of Dueling DQN; train Dueling DQN based on the state space and the reward function; Through the trained Dueling DQN, the physical layer of the target microphone can be seamlessly switched to the optimal frequency band. The real-time packet loss rate is estimated using a channel response-based group spraying method. Based on the real-time packet loss rate, forward error correction and encoding of speech frames are performed according to the level using Polar codes; Voice frames during the switching process are recovered using the Starlight Broadcast Frame and the SC-List decoder.

2. The adaptive switching method for frequency bands of a star-flash wireless microphone according to claim 1, characterized in that, The calculation of the frequency band quality score based on spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics includes: Based on the application scenario of the target microphone, the weights of spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics are determined; The frequency band scoring function is determined based on the weights of spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics. The frequency band quality score is calculated based on real-time spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics.

3. The adaptive switching method for frequency bands of a star-flash wireless microphone according to claim 1, characterized in that, The determination of the state space and reward function of Dueling DQN based on frequency band quality scoring includes: The state space of Dueling DQN is determined based on the frequency band quality score, the number of handovers, and the number of locally detected interference events. The reward function is determined based on the frequency band quality score, the number of handovers, and the handover time.

4. The adaptive switching method for frequency bands of a star-flash wireless microphone according to claim 1, characterized in that, The method for estimating the real-time packet loss rate using a channel response-based packet spraying method includes: Receive confirmation feedback within the preset sliding window size, or broadcast confirmation frames via SLB; Based on the confirmation feedback or the confirmation frame, the packet loss rate of the voice frame within multiple preset sliding windows is calculated. The packet loss rate is smoothed.

5. The adaptive switching method for frequency bands of a star-flash wireless microphone according to claim 4, characterized in that, The recovery of voice frames during the handover process using the Starlight Broadcast Frame and SC-List Decoder includes: SLB extended broadcast frames are sent, and the TCI protocol is used to distinguish between control and data channels. The SC-List decoder is used to recover the audio frames during the switching process.

6. The adaptive switching method for frequency bands of a star-flash wireless microphone according to claim 1, characterized in that, The extraction of spectral occupancy features, signal-to-noise ratio features, and multipath fading features from the speech frames of the target microphone using short-time Fourier transform and neural networks includes: Determine the window, frame length, and frame shift for the short-time Fourier transform; Based on the window, frame length, and frame shift, the speech frames of the target microphone are mapped to a frequency domain representation through short-time Fourier transform. The spectral occupancy, signal-to-noise ratio, and multipath fading features are extracted from the frequency domain representation using a neural network.

7. A frequency band adaptive switching device based on a star-flash wireless microphone, characterized in that, include: The computation module is used to extract spectral occupancy features, signal-to-noise ratio features, and multipath fading features from the speech frames of the target microphone through short-time Fourier transform and neural networks. Based on spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics, the frequency band quality score is calculated. The training module is used to determine the state space and reward function of Dueling DQN based on the frequency band quality score; and to train Dueling DQN based on the state space and the reward function. The control module is used to control the physical layer of the target microphone to seamlessly switch to the optimal frequency band through the trained Dueling DQN; The recovery module is used to estimate the real-time packet loss rate using a channel response-based packet spraying method; Based on the real-time packet loss rate, forward error correction and encoding of speech frames are performed according to the level using Polar codes; Voice frames during the switching process are recovered using the Starlight Broadcast Frame and the SC-List decoder.

8. The adaptive switching device for frequency bands of a star-flash wireless microphone according to claim 7, characterized in that, The computing module includes: The first determining unit is used to determine the weights of spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics based on the application scenario of the target microphone. The second determining unit is used to determine the frequency band scoring function based on the weights of spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics. The calculation unit is used to determine the frequency band scoring function based on the weights of spectrum occupancy characteristics, signal-to-noise ratio characteristics, and multipath fading characteristics.

9. An electronic device, comprising: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the adaptive switching method for frequency bands based on a star-flash wireless microphone as described in any one of claims 1 to 6.

10. A computer-readable medium having a computer program stored thereon, wherein, When the computer program is executed by the processor, it implements the adaptive switching method for frequency bands of a star-flash wireless microphone as described in any one of claims 1 to 6.

Images