Wireless voice quality evaluation method and device

Abstract

The invention discloses a wireless voice quality evaluation method and device. The method comprises the following steps: acquiring a to-be-evaluated voice slice according to a preset period through an audio detection module built in a terminal; under the condition that the duration of the voice slice is greater than a preset duration, obtaining a multi-dimensional evaluation index of the voice slice, the multi-dimensional evaluation index comprising at least one of the following: a voice quality measurement index, a mute frame detection index and a voice quality index; and determining a voice quality evaluation result of the voice slice according to the multi-dimensional evaluation index. The method provided by the invention at least solves the technical problem of low voice quality evaluation accuracy caused by focusing single parameter detection in related technologies.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, in particular to a wireless voice quality evaluation method and device. BACKGROUND

[0002] In the related art, wireless voice quality evaluation schemes are mainly divided into two categories. One is a traditional voice quality evaluation method mainly based on a MOS (Mean Opinion Score) scoring method. The MOS scoring method simulates human subjective auditory perception through a mathematical model, and maps the quality of a voice signal to a MOS value of 1-5 points (5 points being the best). The MOS test needs to be composed of an analog-to-digital conversion module (MOS box), a playback device, and speech corpus, and the scheme is complex and costly. The second category is a single-terminal voice quality evaluation technology, which mainly focuses on single parameter detection (such as packet loss rate and signal-to-noise ratio), lacks a comprehensive evaluation system with multi-module fusion, and does not establish a direct correlation analysis mechanism between voice quality and wireless environment. As a result, in the related art, when evaluating wireless voice quality, the evaluation scheme is complex and focuses on a single parameter, resulting in low evaluation accuracy. SUMMARY

[0003] Embodiments of the present application provide a wireless voice quality evaluation method and device to at least solve the technical problem of low voice quality evaluation accuracy due to focusing on single parameter detection in the related art.

[0004] According to an aspect of an embodiment of the present application, a wireless voice quality evaluation method is provided, comprising: collecting a voice slice to be evaluated according to a preset period through an audio detection module built in a terminal; in a case where a time length of the voice slice is greater than a preset time length, obtaining a multi-dimensional evaluation index of the voice slice, wherein the multi-dimensional evaluation index comprises at least one of a voice quality measurement index, a silence frame detection index, and a voice quality index; and determining a voice quality evaluation result of the voice slice according to the multi-dimensional evaluation index.

[0005] Optionally, determining the voice quality evaluation result of the voice slice according to the multi-dimensional evaluation index comprises: in a case where the voice quality measurement index is abnormal, obtaining the silence frame detection index; in a case where the silence frame detection index is greater than a first threshold value, obtaining the voice quality index; and in a case where the voice quality index is less than a second threshold value, determining that the voice quality evaluation result is that the voice slice is poor quality voice.

[0006] Optionally, the method further comprises: in a case where the silence frame detection index is not greater than the first threshold value, determining that the voice quality evaluation result is that the terminal is faulty; and in a case where the voice quality index is greater than the second threshold value, determining that the voice quality evaluation result is that the voice slice passes verification.

[0007] Optionally, the multi-dimension evaluation index of the voice slice is acquired by receiving a first field and a second field for determining the voice quality measurement index respectively, wherein the first field is used for recording abnormal phenomena of the voice slice, the second field is used for recording nodes of the abnormal phenomena, and the first field and the second field are determined according to pulse code modulation data of multiple nodes in a voice signal processing link; and the voice quality measurement index is determined according to the first field and the second field.

[0008] Optionally, the silence frame detection index is acquired by acquiring original voice signals and decoded voice signals of the voice slice, and respectively segmenting the original voice signals and the decoded voice signals into frame sequences composed of multiple voice frames of a preset length; determining time domain difference indexes, frequency domain difference indexes and perceptual difference indexes of the original voice signals and the decoded voice signals according to the frame sequences, wherein the time domain difference indexes are determined according to mean square errors of the original voice signals and the decoded voice signals, the frequency domain difference indexes are determined according to energy spectrum difference values of the original voice signals and the decoded voice signals, and the perceptual difference indexes are determined according to multiple sub-band perceptual difference metric values, the multiple sub-bands are determined according to frequency domain signals of the original voice signals and the decoded voice signals; and the silence frame detection index is determined according to the time domain difference indexes, the frequency domain difference indexes and the perceptual difference indexes.

[0009] Optionally, the voice quality index is acquired by acquiring a signal-to-noise ratio, a packet loss rate and an efficiency score of the voice slice, wherein the efficiency score is determined according to an acoustic echo cancellation index and an automatic noise suppression index of a communication link for transmitting the voice slice; and the voice quality index is determined according to the signal-to-noise ratio, the packet loss rate and the efficiency score of the voice slice.

[0010] Optionally, the method further comprises: in a case where the voice slice is determined to be poor-quality voice, acquiring multiple network packet information closest to a collection period of the poor-quality voice, wherein the network packet information at least includes one of the following: a wireless environment, a resident cell, adjacent cell information, packet loss information and location information, and the network packet is acquired according to a preset measurement period; determining a cause of generation of the poor-quality voice according to the multiple network packet information; and determining an adjustment scheme according to the cause of generation of the poor-quality voice, the adjustment method being used for adjusting voice quality of the voice slice.

[0011] According to a further aspect of the embodiments of the present application, a wireless voice quality evaluation device is also provided, comprising: a collection module, configured to collect a voice slice to be evaluated according to a preset period through an audio detection module built in a terminal; an acquisition module, configured to acquire a multi-dimension evaluation index of the voice slice in a case where a time length of the voice slice is greater than a preset time length, wherein the multi-dimension evaluation index comprises at least one of a voice quality measurement index, a silence frame detection index and a voice quality index; and an evaluation module, configured to determine a voice quality evaluation result of the voice slice according to the multi-dimension evaluation index.

[0012] According to a further aspect of the embodiments of the present application, a computer device is also provided, comprising: a memory and a processor, wherein the memory is configured to store program instructions; and the processor, connected with the memory, is configured to execute the wireless voice quality evaluation method.

[0013] According to a further aspect of the embodiments of the present application, a computer program product is also provided, comprising computer instructions, which, when executed by a processor, implement the wireless voice quality evaluation method.

[0014] In the embodiments of the present application, the voice slice to be evaluated is collected according to a preset period through an audio detection module built in a terminal; in a case where a time length of the voice slice is greater than a preset time length, a multi-dimension evaluation index of the voice slice is acquired, wherein the multi-dimension evaluation index comprises at least one of a voice quality measurement index, a silence frame detection index and a voice quality index; and a voice quality evaluation result of the voice slice is determined according to the multi-dimension evaluation index. The multi-dimension evaluation index is collected through the audio detection module built in the terminal, so as to achieve the purpose of determining the evaluation result according to the multi-dimension evaluation index, and achieve the technical effect of improving the voice quality evaluation accuracy. Thus, the technical problem of low voice quality evaluation accuracy caused by focusing on single parameter detection in the related art is solved. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings, which are included to provide a further understanding of the present application, constitute a part of the present application and illustrate the illustrative embodiments of the present application and its description, which serve to explain the present application, but do not constitute improper limitations on the present application. In the drawings:

[0016] Figure 1 FIG. 1 is a hardware structure block diagram of a computer terminal for implementing a wireless voice quality evaluation method according to an embodiment of the present application;

[0017] Figure 2 FIG. 2 is a flow chart of a wireless voice quality evaluation method according to an embodiment of the present application;

[0018] Figure 3is a structural schematic diagram of an audio detection module according to an embodiment of the application;

[0019] Figure 4 is a flow chart of another wireless voice quality evaluation method according to an embodiment of the application;

[0020] Figure 5 is a network packet information schematic diagram according to an embodiment of the application;

[0021] Figure 6 is a problem type and boundary condition schematic diagram of poor quality voice according to an embodiment of the application;

[0022] Figure 7 is a poor quality voice area schematic diagram according to an embodiment of the application;

[0023] Figure 8 is a problem type schematic diagram of a poor quality voice area according to an embodiment of the application;

[0024] Figure 9 is a structural diagram of a wireless voice quality evaluation device according to an embodiment of the application. DETAILED DESCRIPTION

[0025] In order to make the personnel in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily mean a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0027] The information collected by the embodiments of the present application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards in the relevant region, necessary security measures are taken, public order and good customs are not violated, and appropriate operation portals are provided for the user to choose to authorize or refuse automatic decision results; if the user chooses to refuse, the expert decision process is entered.

[0028] To solve the problems in the related art, the embodiments of the present application provide a wireless voice quality evaluation method, which can be run in Figure 1 The computer terminal is explained and described below.

[0029] The wireless voice quality evaluation method provided by the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 A hardware structure block diagram of a computer terminal for implementing the wireless voice quality evaluation method is shown. As Figure 1 shown, the computer terminal 10 can include one or more (shown as 102a, 102b, …, 102n in the figure) processors (the processor can include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission module 106 for communication function connected through a wired and / or wireless network. In addition, it can also include a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which can be included as one of the ports of the I / O interface), a network interface, a BUS bus. Those skilled in the art can understand that Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can include more or fewer components than Figure 1 shown, or have a different configuration than Figure 1 shown.

[0030] It should be noted that the one or more processors and / or other data processing circuits described above can be referred to as "data processing circuits" herein. The data processing circuit can be embodied in whole or in part as software, hardware, firmware or any other combination. In addition, the data processing circuit can be a single independent processing module, or all or part of any one of the other elements combined into the computer terminal 10. As referred to in the embodiments of the present application, the data processing circuit controls as a kind of processor (for example, the selection of the variable resistance terminal path connected with the interface).

[0031] The memory 104 can be used to store software programs of application software and modules, such as program instructions / data storage devices corresponding to the wireless voice quality evaluation method in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the wireless voice quality evaluation method described above. The memory 104 can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor, which can be connected to the computer terminal 10 through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0032] The transmission module 106 is used to receive or send data via a network. Specific examples of the above-mentioned network can include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission module 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission module 106 can be a radio frequency (Radio Frequency, RF) module, which is used to communicate with the Internet in a wireless manner.

[0033] The display can be, for example, a touch screen type liquid crystal display (LCD), which can enable a user to interact with the user interface of the computer terminal 10.

[0034] It should be noted that, in some optional embodiments, the above-mentioned Figure 1 The computer terminal shown can include hardware elements (including circuitry), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware and software elements. It should be noted that, Figure 1 is only one example of a particular implementation and is intended to illustrate the types of components that can be present in the above-described computer terminal.

[0035] Under the above-mentioned operating environment, the embodiments of the present application provide a wireless voice quality evaluation method embodiment. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a group of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in a different order.

[0036] In order to better understand the embodiments of the present application, the technical terms involved in the embodiments of the present application are explained as follows:

[0037] MOS: Mean Opinion Score, an average opinion score, is a standard for measuring voice quality;

[0038] BSD: Bad Speech Detection, a bad speech detection scheme. During a call, the terminal may experience various poor call quality phenomena such as single tone, intermittent, noise, etc. When the above problems occur, the terminal analyzes the voice waveform to detect poor voice quality problems.

[0039] VQI: Voice Quality Index, a voice quality index. The downlink reception quality is quantified by calculating the transmission frame, mainly used to evaluate the received voice state and quality. It can reflect the influence of network state on voice call, help monitor and optimize communication quality.

[0040] HIFI VQM: High-Fidelity Voice Quality Measurement, high-fidelity voice quality measurement. By monitoring PCM (Pulse Code Modulation) data from different sources, including microphone input, encoder input, encoder output and speaker output. It aims to detect abnormal phenomena such as silence, intermittent and noise to ensure that the voice quality meets the high-fidelity standard.

[0041] VQM: Voice Quality Measurement, Voice Quality Measurement. Mainly concerned about signal quality measurement during voice call, analyzing the quality of PCM data to evaluate the continuity and clarity of voice, and then judging the overall quality of voice call.

[0042] SFD, Silence Frame Detection, silence frame detection.

[0043] Figure 2 According to an embodiment of the present application, a flowchart of a wireless voice quality evaluation method is shown in FIG. 1, which comprises the following steps: Figure 2

[0044] Step S202, collecting the voice slice to be evaluated according to the preset period through the audio detection module built in the terminal;

[0045] Step S204, in the case where the duration of the voice slice is greater than the preset duration, obtaining the multi-dimensional evaluation index of the voice slice, wherein the multi-dimensional evaluation index comprises at least one of the following: voice quality measurement index, silence frame detection index and voice quality index;

[0046] Step S206, determining the voice quality evaluation result of the voice slice according to the multi-dimensional evaluation index.​

[0047] By the steps S202 to S206, the voice slice to be evaluated is collected by the audio detection module built in the terminal according to a preset period; in the case that the time length of the voice slice is greater than a preset time length, a multi-dimension evaluation index of the voice slice is obtained, wherein the multi-dimension evaluation index comprises at least one of a voice quality measurement index, a silence frame detection index and a voice quality index; and a voice quality evaluation result of the voice slice is determined according to the multi-dimension evaluation index. The multi-dimension evaluation index is collected by the audio detection module built in the terminal, so as to determine the evaluation result according to the multi-dimension evaluation index, improve the voice quality evaluation accuracy, and solve the technical problem of low voice quality evaluation accuracy caused by focusing on single parameter detection in the related art. Details are described below.

[0048] Figure 3 A structural diagram of an audio detection module is shown, as shown in Figure 3 The communication side VQI module, the vocoder SFD module and the VQM module are used for monitoring, wherein the VQI module is a module for quantizing the downlink reception quality by calculating the transmission frame mode; the SFD module monitors the uplink / downlink silence through the vocoder parameters, and can be used for judging the single pass; the VQM module mainly detects abnormal phenomena such as silence, discontinuity and noise through four PCM (Pulse Code Modulation) data of Micin (microphone input), Codein (encoder input), Codeout (encoder output) and Speakout (loudspeaker output), wherein Modem represents a modem, DMA represents Direct Memory Access, Codec represents a codec, and Table 1 shows the detection contents of different detection modules. The VQI module is used for calculating the downlink communication quality score according to the communication and vocoder information through the transmission frame, the SFD module is used for detecting the streaming sound (downlink) and silence (uplink / downlink) according to the vocoder speech frame parameter characteristics, and the VQM module is used for detecting abnormal phenomena such as silence, discontinuity and noise according to the data of the four PCM points of MicIn, CodeIn, CodeOut and SpeakOut.

[0049] In some embodiments of the present application, the specific steps of determining the voice quality evaluation result of the voice slice according to the multi-dimension evaluation index are as follows: in the case that the voice quality measurement index is abnormal, the silence frame detection index is obtained; in the case that the silence frame detection index is greater than a first threshold value, the voice quality index is obtained; and in the case that the voice quality index is less than a second threshold value, it is determined that the voice quality evaluation result is that the voice slice is poor quality voice.

[0050] wherein, in a case that the silence frame detection index is not greater than the first threshold, determining the voice quality evaluation result as the terminal failure; and in a case that the voice quality index is greater than the second threshold, determining the voice quality evaluation result as the voice slice passing the verification.

[0051] In some embodiments of the present application, the specific steps of obtaining the voice quality measurement index in the multi-dimension evaluation index of the voice slice are as follows: respectively receiving a first field and a second field used for determining the voice quality measurement index, wherein the first field is used for recording abnormal phenomena of the voice slice, the second field is used for recording nodes of the abnormal phenomena, and the first field and the second field are determined according to pulse code modulation data of multiple nodes in a voice signal processing link; and determining the voice quality measurement index according to the first field and the second field.

[0052] It needs to be explained that the voice quality measurement index VQM can be obtained through the first field uBSD_Phenomenon and the second field uBSD_Position in the CHR log of the terminal.

[0053] Table 1 shows a first field, as shown in Table 1, the abnormal phenomena of the voice slice include but are not limited to: single pass, intermittent, noise, current sound, water sound.

[0054] It needs to be explained that single pass means that the call only flows in one direction, that is, one party can hear the sound of the other party, but the sound emitted by oneself cannot be heard by the other party. Intermittent means that during the voice call process, the sound appears intermittent interruption or jump, resulting in the conversation becoming intermittent, which is not easy to hear clearly. Intermittent is usually caused by network packet loss, unstable signal, change of wireless environment (such as rapid movement causing signal fluctuation) or terminal buffer mechanism problem. Noise means that in the voice call, there are unexpected background noise, interference sound or other irrelevant audio components, which affect the clarity and intelligibility of the call. Noise may come from environmental noise, internal device noise, signal interference or noise adding mechanism of the encoder. Current sound is a common type of audio interference, which is manifested as a continuous low-frequency humming sound similar to the sound produced when current flows through a conductor. In the call, the current sound usually occurs due to poor signal grounding, electromagnetic interference or the working noise of some electronic components. Water sound means that a continuous background noise similar to water flow appears in the call.

[0055] Table 1

[0056] Bit position Content 1 Single tone 2 Interrupted tone 3 Hum 4 Current hum 5 Reserved 6 Waterfall 7 Reserved 8 Reserved

[0057] Table 2 shows a second field, which indicates the reporting location of abnormal phenomena. As shown in Table 2, MicIn (microphone input) refers to the raw audio signal input captured from the terminal microphone. SpkOut (speaker output) refers to the audio signal output played by the terminal speaker. CodeIn (encoder input) refers to the audio signal before the microphone audio is processed and prepared for encoding in a voice communication system. CodeOut (encoder output) refers to the audio signal after encoding processing, ready for transmission over the wireless network. ECRef (Echo Canceller Reference) refers to the reference signal needed by the echo canceller to identify and eliminate echoes in voice communication. ECRef refers to this reference signal, which is usually derived from the output of the encoder, used for AEC (Acoustic Echo Cancellation) algorithm. LineIn (line input) refers to the audio signal input to the terminal through the line interface (such as earphone line). LineOut (line output) refers to the audio signal output through the terminal line interface, such as the sound played through the earphone. Modem (modem) is an important component in the terminal, responsible for the transmission and reception of wireless signals, including signal modulation and demodulation, as well as communication control with other network devices.

[0058] Table 2

[0059]

[0060]

[0061] Figure 4 A method for evaluating wireless voice quality is shown, as shown in Figure 4 Step 1: Call initialization, when the call starts, the BSD audio detection module of the terminal is started, preparing for real-time monitoring of voice quality. Step 2: Generate BSD slices, the BSD audio detection module generates a voice slice according to the preset period, for example: 20 seconds, to collect and analyze voice quality data in that time period. Step 3: Detect VQM abnormalities.

[0062] During each slice, check the reporting phenomena and location of VQM (Voice Quality Measurement) to identify any voice quality abnormalities.

[0063] If VQM detects abnormalities (such as single tone, intermittent, noise), proceed to the next step of analysis; if VQM does not detect abnormalities, the call continues until the next slice is generated or the call ends.

[0064] Step 4: Calculate SFD value and VQI value: compare SFD value: if ∑SFD≤ first threshold, for example: 0.1, it indicates that audio decoding is basically normal, and the abnormality is likely caused by terminal itself problem. If ∑SFD> first threshold, further analysis is made in combination with VQI value, and average VQI value is calculated to determine the severity of wireless voice problem. Step 5: Determine wireless voice problem: determine whether there is a serious wireless voice problem according to VQI average value and SFD value: if average VQI< 60 and SFD abnormality exists, it is considered that the voice quality problem is on the wireless side. If average VQI> 60, it indicates that the problem may not be serious, but still needs attention. According to the bit information of VQM, specific voice abnormal phenomena (single pass, intermittent, noise, etc.) are identified.

[0065] It needs to be noted that the average VQI< 60 and the VQM report single pass and intermittent problem on the wireless side, and the terminal side voice detection has a low detection rate through the packet loss rate feedback of the packet loss rate for the uplink problem, which can be considered as a wireless downlink voice quality problem. Wireless uplink voice quality is mainly determined by the base station side.

[0066] In some embodiments of the present application, the specific manner of obtaining the silence frame detection index is: obtaining original voice signals and decoded voice signals of the voice slice, and respectively segmenting the original voice signals and the decoded voice signals into frame sequences composed of a plurality of voice frames of a plurality of preset lengths; determining time domain difference indexes, frequency domain difference indexes and perceptual difference indexes of the original voice signals and the decoded voice signals according to the frame sequences, wherein the time domain difference indexes are determined according to mean square errors of the original voice signals and the decoded voice signals, the frequency domain difference indexes are determined according to energy spectrum difference values of the original voice signals and the decoded voice signals, and the perceptual difference indexes are determined according to a plurality of sub-band perceptual difference metric values, the plurality of sub-bands being determined according to frequency domain signals of the original voice signals and the decoded voice signals; and determining the silence frame detection index according to the time domain difference indexes, the frequency domain difference indexes and the perceptual difference indexes.

[0067] Specifically, the calculation of SFD is based on the differential calculation of the original voice frame and the decoded voice frame. The process is divided into the following two steps: 1) preprocessing: frame segmentation and windowing: the original voice signal and the decoded voice signal are divided into frames of a preset length, for example: 20ms, and a Hamming window is applied to reduce the boundary effect. Time domain alignment: the time domain waveforms of the two signals are aligned through a cross-correlation function to eliminate the time delay difference. 2) difference calculation: time domain difference: calculate the mean square error (MSE) of the two signals:

[0068] MSE=N*(xi-yi) 2 ;

[0069] where xi is the original speech signal, yi is the decoded signal, N is the frame length, and i represents the number of input / output audio channels.

[0070] Frequency domain difference: Convert the signals to the frequency domain by FFT, and calculate the energy spectrum difference:

[0071] Frequency domain difference

[0072] X k and Y k are the frequency energy of the original speech signal and the frequency energy of the decoded speech signal, respectively, K represents the total number of frequency points, and k represents the frequency point number.

[0073] Perceptual difference: Based on Bark sub-band division, calculate the perceptual distortion (such as masking effect) of each sub-band.

[0074] Weighted aggregation of time domain difference, frequency domain difference and perceptual difference to generate SFD value. SDF value ranges from 0 to 1, and the smaller the value, the lower the distortion.

[0075] In some embodiments of the present application, the voice quality index is obtained as follows: the signal-to-noise ratio, packet loss rate and performance score of the voice slice are obtained, wherein the performance score is determined according to the acoustic echo cancellation index and automatic noise suppression index of the communication link transmitting the voice slice; the voice quality index is determined according to the signal-to-noise ratio, packet loss rate and performance score of the voice slice.

[0076] Specifically, VQI = a · SNR + b · (1-packet loss rate) + g · algorithm performance score. Wherein a, b, g are weight coefficients, SNR represents the signal-to-noise ratio of the voice slice, and the algorithm performance score covers AEC echo suppression ratio, ANR noise cancellation rate, etc. BSDVQI is a percentage system, with a value of 0-100. Wherein:

[0077] 0-40: poor quality, there may be single, intermittent or noise problems.

[0078] 40-60: general quality, speech is understandable but there are obvious problems.

[0079] 60-80: good quality, clear speech and fewer problems.

[0080] 80-100: excellent quality, speech is completely clear and problem-free.

[0081] When the speech slice is determined to be of poor quality, multiple network packet information closest to the acquisition time of the poor quality speech is obtained. The network packet information includes at least one of the following: wireless environment, stationary cell, neighbor cell information, packet loss information, and location information. The network packets are acquired according to a preset measurement period. The cause of the poor quality speech is determined based on the multiple network packet information. An adjustment scheme is determined based on the cause of the poor quality speech. The adjustment method is used to adjust the speech quality of the speech slice.

[0082] like Figure 5 As shown, the network packet information includes: Radio environment information: RSRP (Reference Signal Received Power): A metric measuring radio signal strength, reflecting the reference signal power received by the UE (User Equipment) from the serving cell. SINR (Signal-to-Interference-plus-Noise Ratio): A parameter used to evaluate signal quality, reflecting the relative strength of the signal relative to noise and interference. RSRQ (Reference Signal Received Quality): A metric combining signal strength and signal quality, reflecting the received signal quality of the cell. Resident cell information: PCI (Physical Cell ID): A unique identifier used to distinguish different cells within the same frequency band. Frequency point: Refers to the radio frequency resources used to determine the carrier frequency of the serving cell. ECI (eNodeB Cell Identifier): In LTE networks, an ID used to uniquely identify a cell within an eNodeB. RAT (Radio Access Technology): Refers to the wireless communication technology currently used by the terminal, such as 4G (LTE) and 5G (NR). Neighbor Cell Information: Number of Effective Neighbor Cells: Helps determine if there are signal islands or inter-cell coordination problems. Number of Co-frequency Neighbor Cells: Indicates the number of neighboring cells on the same carrier frequency, related to interference management. Packet Loss Information: Packet loss rate based on RTP packet sequence numbers: Used to assess the reliability of wireless transmission; a high packet loss rate usually indicates network instability or congestion. Out-of-order Rate: The rate at which voice packets may be out of order during transmission, affecting the continuity of voice. Location Information: MDT (Measurement Data Transfer) location information: If the terminal supports it, more accurate geographical location information can be obtained.

[0083] Based on the terminal BSD report and network CHR log, three types of wireless environment problems can be identified: weak RF coverage, poor DL ​​quality, and high packet loss; and three types of interoperability problems can be identified: frequent handovers, signal islanding, and inter-system handovers. The problem types and delimitation conditions are as follows: Figure 6 The criteria include: Wireless Environment Issues: RF Weak Coverage: The threshold is an average RSRP below -110dBm, indicating insufficient wireless signal strength to maintain high-quality voice calls. DL Quality Deterioration: The threshold is an average SINR below 0dB, indicating poor downlink signal quality, severely affected by noise and interference. High Packet Loss: The threshold is an RTP packet loss rate exceeding 5%, meaning a high probability of packet loss during network transmission, directly affecting voice continuity and clarity. Interoperability Issues: Frequent Handover: The threshold is more than 5 PCI changes per cell during a 20-second slice. Frequent handovers are usually caused by unreasonable network planning, improper handover threshold settings, or excessively fast movement, leading to decreased call stability. Signal Island: The threshold is more than 2 valid measurements without neighboring cells within a 20-second slice. This indicates that the terminal may be located at the coverage edge or in a blind spot, lacking effective neighbor cell support, which can easily lead to a decline in service quality. Inter-RAT Handover is defined by: a RAT (Radio Access Technology) handover occurring, i.e., within 20 seconds, the terminal switches from one radio access technology to another (e.g., from 4G to 5G or vice versa). This can introduce additional latency and potential connectivity issues, especially when interoperability between different systems is poor.

[0084] Based on the detection results, Table 3 shows the corresponding handling solutions for different problem types, including: For weak RF coverage, priority should be given to coverage optimization, such as increasing base station density and adjusting antenna azimuth and elevation angles. For poor DL ​​quality, fine-grained SINR optimization should be performed, such as interference coordination, frequency refarming, or dynamic power adjustment strategies. For high packet loss, network congestion needs to be investigated, voice enhancement features and packet loss recovery mechanisms should be enhanced to ensure data transmission reliability. To address frequent handovers, handover sequences and thresholds should be optimized to reduce unnecessary network handovers. Signal islands should be addressed by strengthening signal coverage of surrounding base stations to ensure users maintain good network connectivity while mobile. Inter-system handover should be optimized by fine-tuning interoperability strategies and handover thresholds to reduce the risk of call interruptions caused by inter-system handovers.

[0085] Table 3

[0086]

[0087] Figure 7 A schematic diagram of a voice quality difference area is shown, as Figure 7 indicated, by the voice quality difference sampling points and latitude and longitude information reported by the terminal, the high voice quality difference area can be converged by clustering. For example, according to the convergence of the area with a quality difference rate > 5% in the detection area, the area that needs to be processed is obtained, and the circular area in the figure is a quality difference area.

[0088] Figure 8 Different problem types in the voice quality difference area are shown, as Figure 8 indicated, the first type of area is frequent switching, the second type of area is high packet loss, the third type of area is inter-system switching, the fourth type of area is high SINR < 0, and the fifth type of area is RF-weak coverage.

[0089] Figure 9 A wireless voice quality evaluation device is shown, which comprises:

[0090] The acquisition module 90 is configured to collect the voice slice to be evaluated according to a preset period through the audio detection module built in the terminal;

[0091] The acquisition module 92 is configured to acquire a multi-dimensional evaluation index of the voice slice in the case that the time length of the voice slice is greater than a preset time length, wherein the multi-dimensional evaluation index comprises at least one of the following: a voice quality measurement index, a silence frame detection index, and a voice quality index.

[0092] The evaluation module 94 is configured to determine the voice quality evaluation result of the voice slice according to the multi-dimensional evaluation index.

[0093] The wireless voice quality evaluation device described above collects the voice slice to be evaluated according to a preset period through the audio detection module built in the terminal; in the case that the time length of the voice slice is greater than a preset time length, a multi-dimensional evaluation index of the voice slice is acquired, wherein the multi-dimensional evaluation index comprises at least one of the following: a voice quality measurement index, a silence frame detection index, and a voice quality index; the voice quality evaluation result of the voice slice is determined according to the multi-dimensional evaluation index, and the multi-dimensional evaluation index is collected through the audio detection module built in the terminal, thereby achieving the purpose of determining the evaluation result according to the multi-dimensional evaluation index, achieving the technical effect of improving the voice quality evaluation accuracy, and further solving the technical problem of low voice quality evaluation accuracy caused by focusing on single parameter detection in the related art.

[0094] The evaluation module 94 comprises an evaluation submodule for determining a voice quality evaluation result of the voice slice according to the multi-dimension evaluation index, comprising: in the case that the voice quality measurement index is abnormal, obtaining the silence frame detection index; in the case that the silence frame detection index is greater than a first threshold, obtaining the voice quality index; in the case that the voice quality index is less than a second threshold, determining that the voice quality evaluation result is that the voice slice is poor quality voice.

[0095] In the case that the silence frame detection index is not greater than the first threshold, it is determined that the voice quality evaluation result is that the terminal is faulty; in the case that the voice quality index is greater than the second threshold, it is determined that the voice quality evaluation result is that the voice slice passes the verification.

[0096] The evaluation submodule comprises a first obtaining unit, a second obtaining unit and a third obtaining unit, wherein the first obtaining unit is configured to obtain the multi-dimension evaluation index of the voice slice, comprising: respectively receiving a first field and a second field for determining the voice quality measurement index, wherein the first field is configured to record abnormal phenomena of the voice slice, the second field is configured to record nodes of the abnormal phenomena, and the first field and the second field are determined according to pulse code modulation data of a plurality of nodes in a voice signal processing link; and the voice quality measurement index is determined according to the first field and the second field.

[0097] The second obtaining unit is configured to obtain the silence frame detection index, comprising: obtaining original voice signals and decoded voice signals of the voice slice, and respectively segmenting the original voice signals and the decoded voice signals into frame sequences composed of a plurality of voice frames of a plurality of preset lengths; determining a time domain difference index, a frequency domain difference index and a perceptual difference index of the original voice signals and the decoded voice signals according to the frame sequences, wherein the time domain difference index is determined according to a mean square error of the original voice signals and the decoded voice signals, the frequency domain difference index is determined according to an energy spectrum difference value of the original voice signals and the decoded voice signals, and the perceptual difference index is determined according to a plurality of sub-band perceptual difference metric values, the plurality of sub-bands being determined according to frequency domain signals of the original voice signals and the decoded voice signals; and the silence frame detection index is determined according to the time domain difference index, the frequency domain difference index and the perceptual difference index.

[0098] The third acquisition unit is configured to acquire the voice quality index, including: acquiring a signal-to-noise ratio, a packet loss rate and a performance score of the voice slice, wherein the performance score is determined according to an acoustic echo cancellation indicator and an automatic noise suppression indicator of a communication link for transmitting the voice slice; and determining the voice quality index according to the signal-to-noise ratio, the packet loss rate and the performance score of the voice slice.

[0099] The wireless voice quality evaluation device further includes an association submodule configured to, in a case where the voice slice is determined to be poor-quality voice, acquire a plurality of network packet information closest to a collection time period of the poor-quality voice, wherein the network packet information at least includes one of the following: a wireless environment, a resident cell, adjacent cell information, packet loss information and location information, and the network packet is acquired according to a preset measurement period; determine a cause of the poor-quality voice according to the plurality of network packet information; and determine an adjustment scheme according to the cause of the poor-quality voice, wherein the adjustment scheme is configured to adjust the voice quality of the voice slice.

[0100] It should be noted that, Figure 9 The wireless voice quality evaluation device is configured to execute the wireless voice quality evaluation method as shown in Figure 2 The wireless voice quality evaluation device is configured to execute the wireless voice quality evaluation method as shown in

[0101] The embodiments of the present application further provide a computer device, including a memory and a processor, wherein the memory is configured to store program instructions; the processor is connected with the memory and is configured to execute the wireless voice quality evaluation method.

[0102] The wireless voice quality evaluation method executed by the computer device adopts an audio detection module built in a terminal to collect a voice slice to be evaluated according to a preset period; in a case where a time length of the voice slice is greater than a preset time length, a multi-dimension evaluation index of the voice slice is acquired, wherein the multi-dimension evaluation index includes at least one of the following: a voice quality measurement indicator, a silence frame detection indicator and a voice quality index; and a voice quality evaluation result of the voice slice is determined according to the multi-dimension evaluation index, so as to achieve the purpose of determining the evaluation result according to the multi-dimension evaluation index, and achieve the technical effect of improving the voice quality evaluation accuracy, thereby solving the technical problem of low voice quality evaluation accuracy caused by focusing on single parameter detection in the related art.

[0103] The embodiments of the present application further provide a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the wireless voice quality evaluation method in the present application.

[0104] The above application embodiment serial numbers are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0105] In the above embodiments of the present application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0106] In the several embodiments provided by the present application, it should be understood that the disclosed technology can be implemented in other ways. Of course, the unit described as the division herein is only a logical division, and there can be other division manners in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, unit or module, and can be electrical or other forms.

[0107] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0108] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The above integrated unit can be realized in the form of hardware or in the form of software functional unit.

[0109] The integrated unit, if realized in the form of software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of software product, which is stored in a storage medium and includes a plurality of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and various program code storage media.

[0110] The above merely describes the preferred embodiments of the present application, and it should be pointed out that, for those skilled in the art, some improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A method of wireless speech quality assessment, characterized by, The method comprises: acquiring a voice slice to be evaluated by an audio detection module built in a terminal according to a preset period; in a case where a duration of the voice slice is greater than a preset duration, acquiring a multi-dimension evaluation index of the voice slice, wherein the multi-dimension evaluation index comprises at least one of a voice quality measurement index, a silence frame detection index and a voice quality index; determining a voice quality evaluation result of the voice slice according to the multi-dimension evaluation index.

2. The method of claim 1, wherein, The method further comprises: in a case where the voice quality measurement index is abnormal, acquiring the silence frame detection index; in a case where the silence frame detection index is greater than a first threshold, acquiring the voice quality index; in a case where the voice quality index is less than a second threshold, determining that the voice quality evaluation result is that the voice slice is poor quality voice.

3. The method of claim 2, wherein, The method further comprises: in a case where the silence frame detection index is not greater than the first threshold, determining that the voice quality evaluation result is that the terminal is faulty; in a case where the voice quality index is greater than the second threshold, determining that the voice quality evaluation result is that the voice slice passes verification.

4. The method of claim 2, wherein, The method further comprises: receiving a first field and a second field for determining the voice quality measurement index respectively, wherein the first field is used for recording an abnormal phenomenon of a voice slice, the second field is used for recording a node of the abnormal phenomenon, and the first field and the second field are determined according to pulse code modulation data of a plurality of nodes in a voice signal processing link; determining the voice quality measurement index according to the first field and the second field.

5. The method of claim 2, wherein, The method further comprises: acquiring an original voice signal and a decoded voice signal of the voice slice, and dividing the original voice signal and the decoded voice signal into a frame sequence composed of a plurality of voice frames of a plurality of preset lengths respectively; determining a time domain difference index, a frequency domain difference index and a perception difference index of the original voice signal and the decoded voice signal according to the frame sequence, wherein the time domain difference index is determined according to a mean square error of the original voice signal and the decoded voice signal, the frequency domain difference index is determined according to an energy spectrum difference value of the original voice signal and the decoded voice signal, and the perception difference index is determined according to a plurality of sub-band perception difference metric values, the plurality of sub-bands being determined according to frequency domain signals of the original voice signal and the decoded voice signal; determining the silence frame detection index according to the time domain difference index, the frequency domain difference index and the perception difference index.

6. The method of claim 2, wherein, The method further comprises: acquiring a signal-to-noise ratio, a packet loss rate and an efficiency score of the voice slice, wherein the efficiency score is determined according to an acoustic echo cancellation index and an automatic noise suppression index of a communication link for transmitting the voice slice; determining the voice quality index according to the signal-to-noise ratio, the packet loss rate and the efficiency score of the voice slice.

7. The method of claim 2, wherein, The method further comprises: In a case where the voice slice is determined as poor-quality voice, a plurality of network packet information closest to a collection time period of the poor-quality voice is acquired, wherein the network packet information at least includes one of the following: wireless environment, resident cell, adjacent cell information, packet loss information, and location information, and the network packet is acquired according to a preset measurement period; A cause of the poor-quality voice is determined according to the plurality of network packet information; An adjustment scheme is determined according to the cause of the poor-quality voice, and the adjustment method is used to adjust voice quality of the voice slice.

8. A wireless voice quality assessment device, characterized by Comprise: A collection module is configured to collect a voice slice to be evaluated according to a preset period through an audio detection module built in a terminal; An acquisition module is configured to acquire a multi-dimensional evaluation index of the voice slice in a case where a time length of the voice slice is greater than a preset time length, wherein the multi-dimensional evaluation index includes at least one of the following: a voice quality measurement index, a silence frame detection index, and a voice quality index; An evaluation module is configured to determine a voice quality evaluation result of the voice slice according to the multi-dimensional evaluation index.

9. A computer device, comprising: Comprise: A memory and a processor, wherein the memory is configured to store program instructions; The processor is connected with the memory and is configured to execute the wireless voice quality evaluation method in any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that, The computer instructions are executed by the processor to implement the wireless voice quality evaluation method in any one of claims 1 to 7.

Images