WO2024093799A1 - 数据标注方法、装置、终端设备及网络侧设备 - Google Patents
数据标注方法、装置、终端设备及网络侧设备 Download PDFInfo
- Publication number
- WO2024093799A1 WO2024093799A1 PCT/CN2023/126883 CN2023126883W WO2024093799A1 WO 2024093799 A1 WO2024093799 A1 WO 2024093799A1 CN 2023126883 W CN2023126883 W CN 2023126883W WO 2024093799 A1 WO2024093799 A1 WO 2024093799A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- terminal device
- location
- tag
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/029—Location-based management or tracking services
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
Definitions
- the present application belongs to the field of communication technology, and specifically relates to a data labeling method, device, terminal equipment and network side equipment.
- AI artificial intelligence
- AI technology is a data-driven technology that can use labeled training data to train artificial intelligence models, and then apply the trained artificial intelligence models to actual scenarios to process corresponding businesses. It can be seen that in AI-based applications, the accuracy of AI models depends largely on the quality of data labels.
- the embodiments of the present application provide a data labeling method, apparatus, terminal device, and network-side device, which can solve the problem of fluctuating accuracy and poor robustness of the AI model caused by the different quality of data labels obtained by the terminal device.
- a data annotation method comprising:
- the terminal device receives first information sent by the network side device, where the first information is used to instruct the terminal device to mark the location data;
- the terminal device labels the location data according to the first information to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- a data annotation method comprising:
- the network side device sends first information to the terminal device, and the first information is used to mark the location data to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- a data labeling device comprising:
- a first information receiving module used to receive first information sent by a network side device, wherein the first information is used to instruct the terminal device to mark the location data;
- a data labeling module is used to label the location data according to the first information to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- a data labeling device comprising:
- the first information sending module is used to send first information to the terminal device, and the first information is used to mark the location data to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- a terminal device which includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the program or instructions are executed by the processor, the steps of the method described in the first aspect are implemented.
- a network side device including a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the second aspect is implemented. The steps of the method described above.
- a data labeling system comprising: a terminal device and a network side device, wherein the terminal device can be used to execute the steps of the data labeling method as described in the first aspect above, and the network side device can be used to execute the steps of the data labeling method as described in the second aspect above.
- a readable storage medium on which a program or instruction is stored.
- the program or instruction is executed by a processor, the steps of the data labeling method described in the first aspect are implemented, or the steps of the data labeling method described in the second aspect are implemented.
- a chip comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the data labeling method as described in the first aspect, or to implement the data labeling method as described in the second aspect.
- a computer program/program product is provided, wherein the computer program/program product is stored in a storage medium, and the computer program/program product is executed by at least one processor to implement the steps of the data labeling method as described in the first aspect or the second aspect.
- a terminal device receives a first message sent by a network side device, and labels the location data according to the first message to obtain a data label; wherein the first message is used to instruct the terminal device to label the location data; the data label is used to iteratively train a machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- the network side device uniformly manages and controls the data labeling process of the terminal device through the first information, and each terminal device labels the acquired location data based on the instruction of the first information sent by the network side device, so that the quality of the data labels obtained by each terminal device is close, which is conducive to improving the accuracy and robustness of the machine learning model obtained based on data label training.
- FIG1 is a block diagram of a wireless communication system to which an embodiment of the present application can be applied;
- FIG2 is a flow chart of a data labeling method in an embodiment of the present application.
- FIG3 is a schematic diagram of a neural network in an embodiment of the present application.
- FIG4 is a schematic diagram of a neuron of a neural network in an embodiment of the present application.
- FIG5 is a flow chart of another data labeling method in an embodiment of the present application.
- FIG6 is a structural block diagram of a data labeling device in an embodiment of the present application.
- FIG7 is a structural block diagram of another data labeling device in an embodiment of the present application.
- FIG8 is a structural block diagram of a communication device in an embodiment of the present application.
- FIG9 is a block diagram of a terminal device in an embodiment of the present application.
- FIG10 is a structural block diagram of a network side device in an embodiment of the present application.
- FIG. 11 is a structural block diagram of another network-side device in an embodiment of the present application.
- first, second, etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way are interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by “first” and “second” are generally of the same type, and the number of objects is not limited.
- the first object can be one or more.
- “and/or” in the specification and claims represents at least one of the connected objects, and the character “/" generally represents that the objects associated with each other are in an "or” relationship.
- LTE Long Term Evolution
- LTE-A Long Term Evolution
- CDMA Code Division Multiple Access
- TDMA Time Division Multiple Access
- FDMA Frequency Division Multiple Access
- OFDMA Orthogonal Frequency Division Multiple Access
- SC-FDMA Single-carrier Frequency Division Multiple Access
- NR New Radio
- 6G 6th Generation
- the wireless communication system includes a terminal device 11 and a network side device 12.
- the terminal device 11 can be a mobile phone, a tablet computer (Tablet Personal Computer), a laptop computer (Laptop Computer) or a notebook computer, a personal digital assistant (Personal Digital Assistant, PDA), a handheld computer, a netbook, an ultra-mobile personal computer (ultra-mobile personal computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), augmented reality (augmented reality, AR)/virtual reality (virtual reality, VR) equipment, a robot, a wearable device (Wearable Device), a vehicle-mounted device (VUE), a pedestrian terminal (PUE), a smart home (home appliances with wireless communication functions, such as refrigerators, televisions, washing machines or furniture, etc.), a game console, a personal computer (personal computer, PC), an ATM or a self-service machine and other terminal side devices, and the wearable devices
- the network side device 12 may include an access network device or a core network device, wherein the access network device 12 may also be referred to as a radio access network device, a radio access network (RAN), a radio access network function or a radio access network unit.
- the access network device 12 may include a base station, a WLAN access point or a WiFi node, etc.
- the base station may be referred to as a node B, an evolved node B (eNB), an access point, a base transceiver station (BTS), a radio base station, a radio transceiver, a basic service set (BSS), an extended service set (ESS), a home B node, a home evolved B node, a transmitting and receiving point (TRP) or other appropriate terms in the field.
- eNB evolved node B
- BTS basic service set
- ESS extended service set
- TRP transmitting and receiving point
- the base station is not limited to a specific technical vocabulary. It should be noted that in the embodiment of the present application, only the base station in the NR system is used as an example for introduction, and the specific type of the base station is not limited.
- the core network equipment may include but is not limited to at least one of the following: core network nodes, core network functions, mobility management entity (Mobility Management Entity, MME), access mobility management function (Access and Mobility Management Function, AMF), session management function (Session Management Function, SMF), user plane function (User Plane Function, UPF), policy control function (Policy Control Function, PCF), policy and charging rules function unit (Policy and Charging Rules Function, PCRF), edge application service discovery function (Edge Application Server Discovery ...
- MME mobility management entity
- AMF Access and Mobility Management Function
- SMF Session Management Function
- SMF Session Management Function
- UPF User Plane Function
- Policy Control Function Policy Control Function
- PCRF Policy and Charging Rules Function
- edge application service discovery function Edge Application Server Discovery ...
- an embodiment of the present application provides a data labeling method.
- FIG. 2 a flow chart of a data labeling method provided by an embodiment of the present application is shown. The method is applied to a terminal device, as shown in FIG. 2 , and the method may specifically include:
- Step 201 A terminal device receives first information sent by a network side device, where the first information is used to instruct the terminal device to mark location data.
- Step 202 The terminal device labels the location data according to the first information to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- the terminal device includes a conventional terminal device and/or a positioning reference unit.
- the conventional terminal device may be the terminal device 11 in Figure 1.
- the positioning reference unit Positioning Reference Unit, PRU
- PRU Positioning Reference Unit
- the positioning reference unit may perform positioning measurements, such as RSTD, RSRP, UE Rx-Tx time difference measurements, etc., and report these measurement results to the positioning server.
- the PRU may send a positioning reference signal (Positioning reference signal, PRS) to the TRP, so that the TRP can measure and report UL positioning measurement values of the PRU from a known location, such as RTOA, UL-AoA, gNB Rx-Tx time difference, etc.
- the location server may compare the PRU measurement values with the expected measurement values at the known PRU location to determine the correction items for other nearby target devices, and then correct the DL and/or positioning of other target devices based on the correction items. Or UL position measurements.
- the network side device can be the access network device in Figure 1, such as a base station or a newly defined artificial intelligence processing node on the access network side, or it can be the core network device in Figure 1, such as a network data analysis function (Network Data Analytics Function, NWDAF), a positioning management function (Location Management Function, LMF), or a newly defined processing node on the core network side, or it can be a combination of the above multiple nodes.
- NWDAF Network Data Analytics Function
- LMF Location Management Function
- a newly defined processing node on the core network side or it can be a combination of the above multiple nodes.
- the network side device sends first information to the terminal device, and the first information is used to instruct the terminal device to mark the location data. It is understandable that the network side device can send the first information to multiple terminal devices to instruct the multiple terminal devices to mark their respective location data.
- the terminal device After receiving the first information sent by the network side device, the terminal device labels the location data according to the first information to obtain a data label, wherein the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- a terminal device with a positioning function can continuously obtain location data during movement, and when receiving the first information sent by the network-side device, the location data is labeled according to the instructions of the first information to obtain a data label.
- the terminal device reports the data label to the network-side device, and the network-side device can iteratively train the machine learning model according to the obtained data label, thereby using the trained machine learning model to execute the target service, such as positioning service, information push service, etc.
- the network-side device After the network-side device completes the training of the machine learning model, it sends the trained machine learning model to the terminal device, and the terminal device uses the trained machine learning model to execute the target service.
- the terminal device labels the location data according to the first information indication to obtain the data label, it uses the data label to iteratively train the machine learning model locally, and uses the trained machine learning model to execute the target service.
- the terminal device sends the trained machine learning model to the network side device, and the network side device uses the trained machine learning model to execute the target service.
- each terminal device labels the acquired location data based on the instructions of the first information sent by the network side device, so that the quality of the data labels obtained by each terminal device is close, which is conducive to improving the accuracy and robustness of the machine learning model obtained based on data label training.
- the machine learning model in the embodiment of the present application can be an artificial intelligence model, such as any one of a fully connected neural network, a convolutional neural network, a decision tree, a support vector machine, and a Bayesian classifier.
- a neural network model its schematic diagram can be shown in Figure 3.
- the neural network may include one or more input layers, one or more hidden layers, and an output layer.
- the data to be processed [X1, X2...Xn] are respectively input into the neural network from the corresponding input layer, and the output result Y is obtained after being processed by the input layer, the hidden layer, and the output layer.
- the neural network is composed of neurons, and a schematic diagram of the neuron is shown in Figure 4.
- a1, a2,...aK represent inputs
- w represents weights (i.e., multiplicative coefficients)
- b represents biases (i.e., additive coefficients)
- ⁇ (.) represents activation functions.
- Common activation functions include Sigmoid (mapping variables between 0 and 1), tanh (translation and contraction of Sigmoid), linear rectification function/rectified linear unit (Rectified Linear Unit, ReLU), etc.
- the model training process is introduced as follows:
- the parameters of the neural network can be optimized through the gradient optimization algorithm.
- the gradient optimization algorithm is a type of algorithm that minimizes or maximizes the objective function (sometimes also called the loss function), and the objective function is often a mathematical combination of model parameters and data. For example, given the data X and its corresponding label Y, a neural network model f(.) can be constructed, then the predicted output f(x) can be obtained based on the input x, and the difference between the predicted value and the true value (f(x)-Y) can be calculated, which is the loss function.
- the optimization goal of the gradient optimization algorithm is to find the appropriate w (i.e. weight) and b (i.e. bias) to minimize the value of the above loss function, and the smaller the loss value, the closer the model is to the actual situation.
- the common optimization algorithms are basically based on the error back propagation (BP) algorithm.
- BP error back propagation
- the basic idea of the BP algorithm is that the learning process consists of two processes: the forward propagation of the signal and the back propagation of the error.
- the input sample is passed from the input layer, processed by each hidden layer layer by layer, and then passed to the output layer. If the actual output of the output layer does not match the expected output, it enters the error back propagation stage.
- Error back propagation is to propagate the output error layer by layer through the hidden layer to the input layer in some form, and distribute the error to all units in each layer, so as to obtain the error signal of each layer unit. This error signal is used as the basis for correcting the weights of each unit.
- This process of adjusting the weights of each layer of the signal forward propagation and error back propagation is repeated.
- the process of continuous adjustment of weights is the learning and training process of the network. This process continues until the error output of the network is The error is reduced to an acceptable level, or the learning process is continued for a preset number of times.
- the above-mentioned model training method is applied to the field of wireless communications, and the terminal device and the network side device interact to realize the training of the machine learning model.
- the first information includes a data screening condition and a label content; the terminal device labels the location data according to the first information to obtain a data label, including:
- the terminal device filters the location data according to the data filtering condition to obtain first location data
- the terminal device labels the first location data according to the tag content to obtain a data tag.
- the network side device can instruct the terminal device to filter the location data that does not meet the data filtering conditions through the data filtering conditions and label content carried in the first information, and mark the location data according to the label content.
- the terminal device receives the first information sent by the network side device, it filters the location data according to the data filtering conditions to obtain the first location data.
- the first location data is the location data that meets the data filtering conditions.
- the terminal device marks the first location data according to the label content to obtain a data label.
- the data label is the marked location data that meets the data filtering conditions.
- each terminal device filters the location data according to the data filtering conditions indicated by the network side device and then marks it, which improves the quality of the final data label, and the data labels obtained by each terminal device all meet the data filtering conditions, and the data quality is similar, which is conducive to improving the accuracy and robustness of the machine learning model obtained based on data label training.
- the label content includes at least one of the following:
- a location tag where the location tag is used to reflect the location coordinates of the terminal device, where the location coordinates include absolute location coordinates or relative location coordinates;
- an area label the area label is used to reflect the area corresponding to the location data
- A3 a timestamp tag, the timestamp tag is used to reflect the acquisition time of the location data
- A4 a location marking error tag, wherein the location marking error tag is used to reflect the error between the location tag and the actual location of the terminal device;
- a terminal tag wherein the terminal tag includes a device type tag or an identity tag of the terminal device;
- A6 a data acquisition tag, the data acquisition tag is used to indicate the acquisition method of the location data
- A7 a motion state tag
- the motion state tag is used to indicate the motion speed and direction of the terminal device when acquiring the location data
- the tag content carried in the first information may include at least one of A1 to A8.
- the location tag may be a location coordinate used to reflect the location of the terminal; the location coordinate may be an absolute location coordinate, such as a location coordinate expressed in longitude and latitude, or a relative location coordinate, such as a relative location coordinate of the terminal device relative to the network side device with the network side device as a reference point.
- the area label can be the administrative division code of the area where the terminal device is located when acquiring the location data, or it can be the coding information of the area where the terminal device is located when acquiring the location data.
- the coding information can be the area code or spatial index obtained according to the area coding algorithm or the spatial index algorithm.
- a target area can be pre-selected as the area for collecting location data.
- the target area can be an area at any level such as the country, province, city, district (county), etc., or the target area can be determined according to the target business executed by the machine learning model as needed.
- the target area is divided and encoded to obtain the coding information of each divided unit area.
- the terminal device can use the coding information of the unit area where the location data is located when acquiring the location data as the location label of the location data. It should be noted that the type of area label and the method for determining the area label can be pre-set by the network side device or specified by the protocol, and the embodiment of the present application does not specifically limit this.
- the timestamp tag may be a timestamp corresponding to the time when the location data is acquired.
- the location error label is used to reflect the difference between the location label and the actual location of the terminal device. Error.
- the location error label may be an actual error value, such as the difference between the location coordinates of the terminal device reflected by the location label and the actual location coordinates of the terminal device; the location error label may also be the error level corresponding to the difference between the location coordinates of the terminal device reflected by the location label and the actual location coordinates of the terminal device, and so on.
- the terminal tag may be a device type tag of the terminal device, for example, if the device type of the terminal device is a conventional terminal device, the device type tag of the terminal device is "UE", if the device type of the terminal device is a positioning reference unit, the device type tag of the terminal device is "PRU", etc.
- the terminal tag may also be an identity tag of the terminal device, such as a terminal identifier, which may be assigned by a network-side device and used to distinguish between various terminal devices in the network.
- the data acquisition tag is used to indicate the method for acquiring the location data
- the method for acquiring the location data can be encoded to obtain the data acquisition tag.
- the method for acquiring the location data can include acquiring the location data collected by the terminal device's own positioning system, acquiring the location data based on the positioning reference signal (PRS), and so on.
- PRS positioning reference signal
- the motion status tag may include a motion speed tag and a direction tag corresponding to the terminal device when acquiring location data, wherein the motion speed tag may be a specific numerical value of the motion speed or the encoded information corresponding to the motion speed; the direction tag may be the encoded information of the motion direction or speed direction corresponding to the terminal device when acquiring location data.
- the data quality level label is used to reflect the data quality of the location data, and the data quality level of the location data can be determined based on the missing field conditions of the location data, the location annotation error label, etc.
- the missing field conditions of the location data can be determined based on whether the location data contains the information corresponding to each item from A1 to A7. The larger the number of items corresponding to A1 to A7 contained in the location data, the more complete the field of the location data is. It can be understood that the more complete the field of the location data is and the smaller the location annotation error is, the higher the data quality level of the location data is.
- the method further includes: the terminal device sending labeling response information to the network side device, the labeling response information is used to reflect whether the terminal device supports labeling of location data, or the labeling response information is used to reflect the tag content supported by the terminal device.
- the terminal device may send a labeling response message to the network side device to reflect whether the labeling of the location data is supported.
- the labeling response message may include a first response message or a second response message, wherein the first response message is used to indicate that the terminal device supports the labeling of the location data, and the first response message may be "1"; the second response message is used to indicate that the terminal device does not support the labeling of the location data, and the second response message may be "0".
- the labeling response information sent by the terminal device to the network-side device can also be used to reflect the tag content supported by the terminal device.
- the labeling response information can carry 8 labeling response tags, corresponding to A1 to A8 respectively.
- the labeling response tag corresponding to the item is the first tag, which is used to indicate that the terminal device supports labeling of the corresponding tag content, and the first tag can be "1"; if the terminal device does not support one of A1 to A8, the labeling response tag corresponding to the item is the second tag, which is used to indicate that the terminal device does not support labeling of the corresponding tag content, and the second tag can be "0".
- the data screening condition includes at least one of the following:
- the position marking error of the position data is less than the preset threshold
- the area corresponding to the location data belongs to the preset area
- the area label corresponding to the location data is a preset area label
- the method for obtaining the location data is the preset method
- the data quality level of the location data is greater than the preset quality level
- the terminal tag of the terminal device is a preset terminal tag
- the terminal capability level of the terminal device is greater than the preset capability level
- the movement speed of the terminal device is less than a preset speed.
- the preset threshold, preset area, preset area label, preset acquisition method, preset time range, preset quality level, preset terminal label, preset capability level, and preset speed can be set by the network side device or specified by the protocol.
- Whether there are missing fields in the location data can be determined according to the above items A1 to A8. If the location data contains the information corresponding to each item from A1 to A8, it means that there is no missing field in the location data; conversely, if the location data does not contain the information corresponding to any item from A1 to A8, it means that there is a missing field in the location data.
- Whether the location data is abnormal data can be determined based on the distribution of the location data in all the location data collected by the terminal device. For example, if the sample point corresponding to the location data is an outlier among the sample points corresponding to all the location data collected by the terminal device, then the location data is abnormal data.
- the terminal device may filter the location data according to the data filtering condition corresponding to at least one of B1 to B11 above, and then label the filtered location data, which can improve the quality of the final data label, thereby helping to improve the accuracy and robustness of the machine learning model obtained based on data label training.
- the terminal device may first label the location data, and then filter the labeled location data according to the data filtering condition corresponding to at least one of B1 to B11 above, to obtain the final data label. It should be noted that the embodiment of the present application does not specifically limit the order of data screening and data labeling.
- the terminal device labels the location data according to the first information to obtain a data label, including:
- Step S11 the terminal device preprocesses the location data according to a data preprocessing method preset or specified in a protocol to obtain preprocessed location data;
- Step S12 The terminal device labels the preprocessed location data according to the first information to obtain a data label.
- the terminal device may first pre-process the location data according to a data pre-processing method, and then mark the pre-processed location data.
- the data pre-processing method may be pre-set locally by the terminal device or may be specified by a protocol.
- the data pre-processing method includes at least one of the following: interpolation processing, smoothing processing, quantization processing, and normalization processing.
- the first information further includes a data preprocessing method, and the terminal device labels the first location data according to the tag content.
- the method further includes:
- Step S21 the terminal device preprocesses the first position data according to the data preprocessing method to obtain second position data
- Step 202 the terminal device labels the first location data according to the tag content to obtain a data tag, including:
- Step S22 The terminal device labels the second location data according to the label content to obtain a data label.
- the data preprocessing method can be specified by the network side device, and the data preprocessing method is carried in the first information.
- the terminal device can first filter the position data according to the data filtering condition carried in the first information to obtain the first position data that meets the data filtering condition; then, preprocess the first position data according to the data preprocessing method to obtain the second position data; finally, mark the second position data according to the label content in the first information to obtain the data label.
- the data filtering condition includes at least one of the aforementioned B1 to B11; the label content includes at least one of the aforementioned A1 to A8; the data preprocessing method includes at least one of the following: interpolation processing, smoothing processing, quantization processing, and normalization processing.
- the data preprocessing method includes interpolation processing.
- the terminal device preprocesses the first position data according to the data preprocessing method.
- the method further includes:
- Step S31 The terminal device receives second information sent by the network side device, where the second information carries a label density indication, and the label density indication is used to indicate a label density of the data label;
- Step S21 the terminal device preprocesses the first position data according to the data preprocessing method to obtain second position data, including:
- Step S32 The terminal device interpolates the first position data according to the tag density indication to obtain second position data.
- the network side device can send second information carrying a label density indication to the terminal device, wherein the label density indication is used to indicate the label density of the data label.
- the label density indication includes any one of the following:
- the number of data tags is N1;
- the number of data tags within the first distance range is N2;
- the label density of data labels is N3/time unit, or N4/distance unit.
- N1, N2, N3 and N4 are all positive integers.
- the time unit includes any one of the following: hour, minute, second, millisecond, microsecond, frame, subframe, time slot, orthogonal frequency-division multiplexing (OFDM) symbol.
- the distance unit may include any one of the following: kilometer, meter, decimeter, centimeter.
- the terminal device After receiving the second information sent by the network side device, the terminal device interpolates the first position data according to the label density indication to obtain the second position data, and the data density of the second position data meets the label density indication, that is, one of C1 to C3.
- the interpolation processing may include but is not limited to linear interpolation and quadratic interpolation.
- the first location data acquired by the terminal device is:
- different second position data can be obtained by using different interpolation processing. Take the position data at time t2 obtained by using interpolation processing as an example:
- the position data at time t2 is obtained as:
- the second position data that meets the label density indication can be obtained.
- the data preprocessing method includes interpolation processing.
- the terminal device preprocesses the first position data according to the data preprocessing method.
- the method further includes:
- Step S41 The terminal device receives third information sent by the network side device, where the third information carries an interpolation indication;
- Step S21 the terminal device preprocesses the first position data according to the data preprocessing method to obtain second position data, including:
- Step S42 The terminal device performs interpolation processing on the first position data according to the target interpolation method corresponding to the interpolation indication to obtain second position data.
- the network side device when the data preprocessing method carried in the first information includes interpolation processing, can send an interpolation indication to the terminal device through the third information to indicate the target interpolation method required for the interpolation processing. After receiving the third information, the terminal device performs interpolation processing on the first position data according to the target interpolation method corresponding to the interpolation indication.
- the target interpolation method may include but is not limited to linear interpolation and quadratic interpolation.
- the data preprocessing method includes smoothing processing.
- the terminal device preprocesses the first position data according to the data preprocessing method.
- the method further includes:
- Step S51 The terminal device receives fourth information sent by the network side device, where the fourth information carries a smoothing indication
- Step S21 the terminal device preprocesses the first position data according to the data preprocessing method to obtain second position data, including:
- Step S52 The terminal device smoothes the first position data according to the target smoothing method corresponding to the smoothing indication to obtain second position data.
- the network side device can send a smoothing indication to the terminal device through the fourth information to indicate the target smoothing method required for the smoothing.
- the terminal device After receiving the fourth information, the terminal device performs smoothing on the first position data according to the target smoothing method corresponding to the smoothing indication.
- the target smoothing method may include but is not limited to the S-G (Savitzky-Golay) filter smoothing method, the moving average method, and the moving average filter method.
- the sliding window contains 2N+1 first position data
- the first position data (x, y) at time t is smoothed
- the second position data after smoothing at time t is:
- the noise in the location data can be removed, thereby improving the quality of the data labels obtained by annotating the smoothed location data.
- the data preprocessing method includes quantization processing.
- the terminal device preprocesses the first position data according to the data preprocessing method.
- the method further includes:
- Step S61 The terminal device receives fifth information sent by the network side device, where the fifth information carries data quantization auxiliary information;
- Step S21 the terminal device preprocesses the first position data according to the data preprocessing method to obtain second position data, including:
- Step S62 The terminal device quantizes the first position data according to the data quantization auxiliary information to obtain second position data.
- the data quantization auxiliary information includes at least one of the following:
- Quantization method wherein the quantization method includes uniform quantization or non-uniform quantization.
- the identifier of the quantization reference point is used to indicate the identity of the quantization reference point, and each quantization reference point corresponds to an identity identifier and a position coordinate.
- the position of quantization reference point 1 is (0, 0).
- uniform quantization refers to the quantization of the input signal's value range divided into equal intervals, also known as linear coding. Its characteristic is that the width of the quantization interval (i.e., the width order) is the same. Uniform quantization mainly indicates the quantization interval, such as a quantization interval of 10. Non-uniform quantization refers to a quantization in which the quantization intervals are unequal within the dynamic range of the input signal. In non-uniform quantization, for intervals with small signal values, the quantization interval is also small. Conversely, the larger the signal value, the larger the quantization interval.
- Non-uniform quantization can indicate a quantization codebook, such as the quantization codebook of the x dimension is ⁇ 0, 10, 30, 40, 45 ⁇ , and the quantization codebook of the y dimension is ⁇ 0, 5, 20, 40 ⁇ .
- the quantization codebook can reflect the quantization interval.
- the quantization codebook of the x dimension is ⁇ 10, 20, 10, 5 ⁇
- the quantization codebook of the y dimension is ⁇ 5, 15, 20 ⁇ .
- the position of the terminal device 1 is (8, 12)
- it is mapped to the quantization codebook
- the quantization interval of the x dimension is 10
- the quantization interval of the y dimension is 10.
- the network side device can send data quantization auxiliary information to the terminal device through the fifth information. After receiving the fifth information, the terminal device quantizes the first position data according to the data quantization auxiliary information.
- the data preprocessing method includes normalization processing.
- the terminal device preprocesses the first position data according to the data preprocessing method.
- the method further includes:
- Step S71 The terminal device receives sixth information sent by the network side device, where the sixth information carries a normalization indication
- Step S21 the terminal device preprocesses the first position data according to the data preprocessing method to obtain second position data, including:
- Step S72 The terminal device normalizes the first position data according to the target normalization method corresponding to the normalization indication to obtain second position data.
- the network side device when the data preprocessing method carried in the first information includes normalization processing, can send a normalization indication to the terminal device through the sixth information to indicate the target normalization method required for the normalization processing. After the terminal device receives the sixth information, it normalizes the first position data according to the target normalization method corresponding to the normalization indication.
- the target normalization method may include but is not limited to the maximum and minimum normalization (X-Min/(Max-Min)) method and the zero mean normalization (X-Mean/(Standard deviation)) method.
- an embodiment of the present application provides another data labeling method.
- FIG. 5 a flow chart of another data labeling method provided by an embodiment of the present application is shown. The method is applied to a network side device, as shown in FIG. 5 , and the method may specifically include:
- Step 501 The network side device sends first information to the terminal device, wherein the first information is used to indicate the terminal device
- the location data is annotated to obtain data labels.
- Step 502 The network side device receives the data tag sent by the terminal device
- Step 503 The network side device iteratively trains the machine learning model according to the data label, and the machine learning model is used to predict the location information of the terminal device.
- the terminal device includes a conventional terminal device and/or a positioning reference unit.
- the conventional terminal device can be the terminal device 11 in Figure 1.
- the network side device can be the access network device in Figure 1, such as a base station or a newly defined artificial intelligence processing node on the access network side, or it can be the core network device in Figure 1, such as a network data analysis function (Network Data Analytics Function, NWDAF), a positioning management function (Location Management Function, LMF), or a newly defined processing node on the core network side, or it can be a combination of the above multiple nodes.
- NWDAF Network Data Analytics Function
- LMF Location Management Function
- the network side device sends first information to the terminal device, and the first information is used to instruct the terminal device to mark the location data. It is understandable that the network side device can send the first information to multiple terminal devices to instruct the multiple terminal devices to mark their respective location data.
- the terminal device After receiving the first information sent by the network side device, the terminal device labels the location data according to the first information to obtain a data label, wherein the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- a terminal device with a positioning function can continuously obtain location data during movement, and when receiving the first information sent by the network-side device, the location data is labeled according to the instructions of the first information to obtain a data label.
- the terminal device reports the data label to the network-side device, and the network-side device can iteratively train the machine learning model according to the obtained data label, thereby using the trained machine learning model to execute the target service, such as positioning service, information push service, etc.
- the network-side device After the network-side device completes the training of the machine learning model, it sends the trained machine learning model to the terminal device, and the terminal device uses the trained machine learning model to execute the target service.
- the terminal device labels the location data according to the first information indication to obtain the data label, it uses the data label to iteratively train the machine learning model locally, and uses the trained machine learning model to execute the target service.
- the terminal device sends the trained machine learning model to the network side device, and the network side device uses the trained machine learning model to execute the target service.
- the first information includes data filtering conditions and label content; wherein the data filtering conditions are used to filter the location data to filter out location data that does not meet the data filtering conditions; and the label content is used to mark the location data according to the label content.
- the label content includes at least one of the following:
- a location tag where the location tag is used to reflect the location coordinates of the terminal device, where the location coordinates include absolute location coordinates or relative location coordinates;
- an area label the area label is used to reflect the area corresponding to the location data
- A3 a timestamp tag, the timestamp tag is used to reflect the acquisition time of the location data
- A4 a location marking error tag, wherein the location marking error tag is used to reflect the error between the location tag and the actual location of the terminal device;
- a terminal tag wherein the terminal tag includes a device type tag or an identity tag of the terminal device;
- A6 a data acquisition tag, the data acquisition tag is used to indicate the acquisition method of the location data
- A7 a motion state tag
- the motion state tag is used to indicate the motion speed and direction of the terminal device when acquiring the location data
- the method further includes: the network side device receives labeling response information sent by the terminal device, the labeling response information is used to reflect whether the terminal device supports labeling of location data, or the labeling response information is used to reflect the tag content supported by the terminal device.
- the terminal device may send a labeling response message to the network side device to reflect whether the labeling of the location data is supported.
- the labeling response message may include a first response message or a second response message, wherein the first response message is used to indicate that the terminal device supports the labeling of the location data, and the first response message may be "1"; the second response message is used to indicate that the terminal device does not support the labeling of the location data, and the second response message may be "0".
- the terminal device can also be used to reflect the tag content supported by the terminal device.
- the labeling response information can carry 8 labeling response tags, corresponding to A1 to A8 respectively. If the terminal device supports one of A1 to A8, the labeling response tag corresponding to the item is the first tag, which is used to indicate that the terminal device supports labeling the corresponding tag content, and the first tag can be "1"; if the terminal device does not support one of A1 to A8, the labeling response tag corresponding to the item is the second tag, which is used to indicate that the terminal device does not support labeling the corresponding tag content, and the second tag can be "0".
- the data screening condition includes at least one of the following:
- the position marking error of the position data is less than the preset threshold
- the area corresponding to the location data belongs to the preset area
- the area label corresponding to the location data is a preset area label
- the method for obtaining the location data is the preset method
- the data quality level of the location data is greater than the preset quality level
- the terminal tag of the terminal device is a preset terminal tag
- the terminal capability level of the terminal device is greater than the preset capability level
- the movement speed of the terminal device is less than a preset speed.
- the method further includes: the network side device receiving the data label sent by the terminal device; the network side device preprocessing the data label according to a preset data preprocessing method to obtain a target data label.
- a data preprocessing operation may be performed by a network side device.
- the network side device receives a data label sent by a terminal device, and the data label is obtained by the terminal device according to the indication of the first information; the network side device preprocesses the received data label according to a pre-set data preprocessing method to obtain a target data label.
- the data preprocessing method includes at least one of the following: interpolation processing, smoothing processing, quantization processing, and normalization processing.
- the first information further includes a data preprocessing method, and the data preprocessing method is used to preprocess the location data.
- the network side device may carry a data preprocessing method in the first information to instruct the terminal device to preprocess the location data according to the data preprocessing method.
- the data preprocessing method includes at least one of the following: interpolation processing, smoothing processing, quantization processing, and normalization processing.
- the data preprocessing method includes interpolation processing, and the method further includes: the network side device sends second information to the terminal device, the second information carries a label density indication, and the label density indication is used to indicate the label density of the data label.
- the network side device can send second information carrying a label density indication to the terminal device, wherein the label density indication is used to indicate the label density of the data label.
- the label density indication includes any one of the following:
- the number of data tags is N1;
- the number of data tags within the first distance range is N2;
- the label density of data labels is N3/time unit, or N4/distance unit.
- N1, N2, N3 and N4 are all positive integers.
- the time unit includes any one of the following: hour, minute, second, millisecond, microsecond, frame, subframe, time slot, orthogonal frequency division multiplexing symbol.
- the distance unit includes any one of the following: kilometer, meter, decimeter, centimeter.
- the data preprocessing method includes interpolation processing, and the method further includes: the network side device sends third information to the terminal device, the third information carries an interpolation indication, and the interpolation indication is used for interpolation processing of the location data.
- the network side device when the data preprocessing method carried in the first information includes interpolation processing, can send an interpolation indication to the terminal device through the third information to indicate the target interpolation method to be used for the interpolation processing. After receiving the third information, the terminal device performs interpolation processing on the first position data according to the target interpolation method corresponding to the interpolation indication.
- the target interpolation method may include but is not limited to linear interpolation, quadratic interpolation, Interpolation.
- the data preprocessing method includes smoothing processing, and the method further includes: the network side device sends fourth information to the terminal device, the fourth information carries a smoothing indication, and the smoothing indication is used for smoothing processing of the location data.
- the network side device can send a smoothing indication to the terminal device through the fourth information to indicate the target smoothing method required for the smoothing.
- the terminal device After receiving the fourth information, the terminal device performs smoothing on the first position data according to the target smoothing method corresponding to the smoothing indication.
- the target smoothing method may include but is not limited to the S-G (Savitzky-Golay) filter smoothing method, the moving average method, and the moving average filter method.
- the data preprocessing method includes quantization processing, and the method further includes: the network side device sending fifth information to the terminal device, the fifth information carrying data quantization auxiliary information, and the data quantization auxiliary information includes at least one of the following:
- Quantization method wherein the quantization method includes uniform quantization or non-uniform quantization.
- the network side device can send data quantization auxiliary information to the terminal device through the fifth information. After receiving the fifth information, the terminal device quantizes the first position data according to the data quantization auxiliary information.
- the data preprocessing method includes normalization processing, and the method further includes: the network side device sends sixth information to the terminal device, the sixth information carries a normalization indication, and the normalization indication is used for normalization processing of the location data.
- the network side device when the data preprocessing method carried in the first information includes normalization processing, can send a normalization indication to the terminal device through the sixth information to indicate the target normalization method required for the normalization processing. After the terminal device receives the sixth information, it normalizes the first position data according to the target normalization method corresponding to the normalization indication.
- the target normalization method may include but is not limited to the maximum and minimum normalization (X-Min/(Max-Min)) method and the zero mean normalization (X-Mean/(Standard deviation)) method.
- each terminal device labels the acquired location data based on the instructions of the first information sent by the network side device, so that the quality of the data labels obtained by each terminal device is close, which is conducive to improving the accuracy and robustness of the machine learning model obtained based on data label training.
- the data labeling method provided in the embodiment of the present application can be executed by a data labeling device.
- the data labeling device performing the data labeling method is taken as an example to illustrate the data labeling device provided in the embodiment of the present application.
- an embodiment of the present application provides a data labeling device.
- a structural block diagram of a data labeling device provided by an embodiment of the present application is shown, and the device can be applied to a terminal device.
- the device can specifically include:
- a first information receiving module 601 is used to receive first information sent by a network side device, where the first information is used to instruct the terminal device to mark the location data;
- the data labeling module 602 is used to label the location data according to the first information to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- the first information includes data screening conditions and label content;
- the data labeling module includes:
- a data screening submodule used for screening the position data according to the data screening condition to obtain first position data
- the first labeling submodule is used to label the first position data according to the label content to obtain a data label.
- the label content includes at least one of the following:
- a location tag wherein the location tag is used to reflect the location coordinates of the terminal device, and the location coordinates include absolute location coordinates or relative location coordinates;
- An area tag where the area tag is used to reflect the area corresponding to the location data
- a timestamp tag wherein the timestamp tag is used to reflect the time when the location data is acquired
- the position marking error tag is used to reflect the difference between the position tag and the terminal device. The error between the actual position of the device and the terminal device.
- a terminal tag wherein the terminal tag includes a device type tag or an identity tag of the terminal device;
- a data acquisition tag wherein the data acquisition tag is used to indicate a method for acquiring the location data
- a motion state tag is used to indicate the motion speed and direction of the terminal device when acquiring the location data
- a data quality level label wherein the data quality level label is used to indicate the data quality of the location data.
- the device also includes: a labeling response information sending module, used to send labeling response information to the network side device, the labeling response information is used to reflect whether the terminal device supports labeling of location data, or the labeling response information is used to reflect the label content supported by the terminal device.
- a labeling response information sending module used to send labeling response information to the network side device, the labeling response information is used to reflect whether the terminal device supports labeling of location data, or the labeling response information is used to reflect the label content supported by the terminal device.
- the data screening condition includes at least one of the following:
- the position marking error of the position data is less than a preset threshold
- the area corresponding to the location data belongs to the preset area
- the area label corresponding to the location data is a preset area label
- the method for obtaining the location data is the preset method
- the acquisition time of the location data falls within the preset time range
- the data quality level of the location data is greater than the preset quality level
- Location data is not considered abnormal data
- the terminal tag of the terminal device is a preset terminal tag
- the terminal capability level of the terminal device is greater than the preset capability level
- the movement speed of the terminal device is less than a preset speed.
- the data annotation module includes:
- a first preprocessing submodule used to preprocess the position data according to a data preprocessing method preset or specified in a protocol to obtain preprocessed position data;
- the second labeling submodule is used to label the preprocessed position data according to the first information to obtain a data label.
- the first information further includes a data preprocessing method
- the data labeling module further includes:
- a second preprocessing submodule configured for the terminal device to preprocess the first location data according to the data preprocessing method to obtain second location data
- the first marking submodule includes:
- a data labeling unit is used to label the second position data according to the label content to obtain a data label.
- the data preprocessing method includes at least one of the following: interpolation processing, smoothing processing, quantization processing, and normalization processing.
- the data preprocessing method includes interpolation processing, and the device further includes:
- a second information receiving module configured to receive second information sent by the network side device, wherein the second information carries a label density indication, and the label density indication is used to indicate a label density of the data label;
- the second preprocessing submodule includes:
- the first interpolation processing unit is used to perform interpolation processing on the first position data according to the label density indication to obtain second position data.
- the label density indication includes any one of the following:
- the number of tags of data tags is N1;
- the number of data tags within the first distance range is N2;
- the label density of data labels is N3/time unit, or N4/distance unit;
- N1, N2, N3 and N4 are all positive integers.
- the time unit includes any one of the following: hour, minute, second, millisecond, microsecond, frame, subframe, time slot, orthogonal frequency division multiplexing symbol.
- the data preprocessing method includes interpolation processing, and the device further includes:
- a third information receiving module configured to receive third information sent by the network side device, wherein the third information carries an interpolation indication
- the second preprocessing submodule includes:
- the second interpolation processing unit is used to perform interpolation processing on the first position data according to a target interpolation method corresponding to the interpolation indication to obtain second position data.
- the data preprocessing method includes smoothing processing
- the device further includes:
- a fourth information receiving module configured to receive fourth information sent by the network side device, wherein the fourth information carries a smoothing indication
- the second preprocessing submodule includes:
- a smoothing processing unit is used to perform smoothing processing on the first position data according to a target smoothing method corresponding to the smoothing instruction to obtain second position data.
- the data preprocessing method includes quantization processing
- the device further includes:
- a fifth information receiving module is configured to receive fifth information sent by the network side device, wherein the fifth information carries data quantization auxiliary information, and the data quantization auxiliary information includes at least one of the following:
- the quantization method includes uniform quantization or non-uniform quantization
- the second preprocessing submodule comprises:
- a quantization processing unit is used to perform quantization processing on the first position data according to the data quantization auxiliary information to obtain second position data.
- the data preprocessing method includes normalization processing, and the device further includes:
- a sixth information receiving module configured to receive sixth information sent by the network side device, wherein the sixth information carries a normalization indication
- the second preprocessing submodule comprises:
- a normalization processing unit is used to perform normalization processing on the first position data according to a target normalization method corresponding to the normalization indication to obtain second position data.
- the terminal device includes a conventional terminal device and/or a positioning reference unit.
- an embodiment of the present application provides another data labeling device.
- FIG. 7 a structural block diagram of a data labeling device provided by an embodiment of the present application is shown, and the device can be applied to a network side device. As shown in FIG. 7 , the device can specifically include:
- the first information sending module 701 is used to send first information to the terminal device, where the first information is used to mark the location data to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- the first information includes data filtering conditions and label content; wherein the data filtering conditions are used to filter the location data to filter out location data that does not meet the data filtering conditions; and the label content is used to mark the location data according to the label content.
- the label content includes at least one of the following:
- a location tag wherein the location tag is used to reflect the location coordinates of the terminal device, and the location coordinates include absolute location coordinates or relative location coordinates;
- An area tag where the area tag is used to reflect the area corresponding to the location data
- a timestamp tag wherein the timestamp tag is used to reflect the time when the location data is acquired
- a location marking error tag wherein the location marking error tag is used to reflect the error between the location tag and the actual location of the terminal device
- a terminal tag wherein the terminal tag includes a device type tag or an identity tag of the terminal device;
- a data acquisition tag wherein the data acquisition tag is used to indicate a method for acquiring the location data
- a motion state tag is used to indicate the motion speed and direction of the terminal device when acquiring the location data
- a data quality level label wherein the data quality level label is used to indicate the data quality of the location data.
- the device further comprises:
- the labeling response information receiving module is used to receive the labeling response information sent by the terminal device, and the labeling response information is used to reflect whether the terminal device supports labeling of location data, or the labeling response information is used to reflect the label content supported by the terminal device.
- the data screening condition includes at least one of the following:
- the position marking error of the position data is less than a preset threshold
- the area corresponding to the location data belongs to the preset area
- the area label corresponding to the location data is a preset area label
- the method for obtaining the location data is the preset method
- the acquisition time of the location data falls within the preset time range
- the data quality level of the location data is greater than the preset quality level
- Location data is not considered abnormal data
- the terminal tag of the terminal device is a preset terminal tag
- the terminal capability level of the terminal device is greater than the preset capability level
- the movement speed of the terminal device is less than a preset speed.
- the device further comprises:
- a data label receiving module used for receiving the data label sent by the terminal device
- the data preprocessing module is used to preprocess the data labels according to a preset data preprocessing method to obtain target data labels.
- the first information further includes a data preprocessing method, and the data preprocessing method is used to preprocess the location data.
- the data preprocessing method includes at least one of the following: interpolation processing, smoothing processing, quantization processing, and normalization processing.
- the data preprocessing method includes interpolation processing, and the device further includes:
- the second information sending module is used to send second information to the terminal device, where the second information carries a label density indication, and the label density indication is used to indicate the label density of the data label.
- the label density indication includes any one of the following:
- the number of tags of data tags is N1;
- the number of data tags within the first distance range is N2;
- the label density of data labels is N3/time unit, or N4/distance unit;
- N1, N2, N3 and N4 are all positive integers.
- the time unit includes any one of the following: hour, minute, second, millisecond, microsecond, frame, subframe, time slot, orthogonal frequency division multiplexing symbol.
- the data preprocessing method includes interpolation processing, and the device further includes:
- the third information sending module is used to send third information to the terminal device, where the third information carries an interpolation indication, and the interpolation indication is used for interpolation processing of position data.
- the data preprocessing method includes smoothing processing
- the device further includes:
- the fourth information sending module is used to send fourth information to the terminal device, and the fourth information carries a smoothing indication, and the smoothing indication is used for smoothing processing of position data.
- the data preprocessing method includes quantization processing
- the device further includes:
- a fifth information sending module is used to send fifth information to the terminal device, where the fifth information carries data quantization auxiliary information, and the data quantization auxiliary information includes at least one of the following:
- a quantization method wherein the quantization method includes uniform quantization or non-uniform quantization.
- the data preprocessing method includes normalization processing, and the device further includes:
- the sixth information sending module is used to send the sixth information to the terminal device, and the sixth information carries a normalization indication, and the normalization indication is used for normalization processing of the location data.
- the terminal device includes a conventional terminal device and/or a positioning reference unit.
- the data annotation device in the embodiment of the present application can be an electronic device, such as an electronic device with an operating system, or a component in an electronic device, such as an integrated circuit or a chip.
- the electronic device can be a terminal device.
- the terminal device can include but is not limited to the types of terminal devices 11 listed above.
- the data labeling device provided in the embodiment of the present application can implement each process implemented by the method embodiment of Figure 2 or Figure 5, and achieve the same technical effect. To avoid repetition, it will not be repeated here.
- the embodiment of the present application further provides a communication device 800, including a processor 801 and a memory 802, the memory 802 storing programs or instructions that can be run on the processor 801, for example, when the communication device 800 is a network side device, the program or instruction is executed by the processor 801 to implement the various steps of the data labeling method embodiment described in the first aspect above, and can achieve the same technical effect.
- the communication device 800 is a terminal device
- the program or instruction is executed by the processor 801 to implement the various steps of the data labeling method embodiment described in the second aspect above, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.
- FIG9 it is a schematic diagram of the hardware structure of a terminal device implementing an embodiment of the present application.
- the terminal device 900 includes but is not limited to: a radio frequency unit 901, a network module 902, an audio output unit 903, an input unit 904, a sensor 905, a display unit 906, a user input unit 907, an interface unit 908, a memory 909 and at least some of the components of a processor 910.
- the terminal device 900 may also include a power source (such as a battery) for supplying power to various components, and the power source may be logically connected to the processor 910 through a power management system, thereby The system realizes functions such as managing charging, discharging, and power consumption management.
- the terminal device structure shown in FIG9 does not constitute a limitation on the terminal device.
- the terminal device may include more or fewer components than shown in the figure, or combine certain components, or arrange components differently, which will not be described in detail here.
- the input unit 904 may include a graphics processing unit (GPU) 9041 and a microphone 9042, and the graphics processor 9041 processes the image data of the static picture or video obtained by the image capture device (such as a camera) in the video capture mode or the image capture mode.
- the display unit 906 may include a display panel 9061, and the display panel 9061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc.
- the user input unit 907 includes a touch panel 9071 and at least one of other input devices 9072.
- the touch panel 9071 is also called a touch screen.
- the touch panel 9071 may include two parts: a touch detection device and a touch controller.
- Other input devices 9072 may include, but are not limited to, a physical keyboard, function keys (such as a volume control key, a switch key, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.
- the RF unit 901 can transmit the data to the processor 910 for processing; in addition, the RF unit 901 can send uplink data to the network side device.
- the RF unit 901 includes but is not limited to an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.
- the memory 909 can be used to store software programs or instructions and various data.
- the memory 909 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instruction required for at least one function (such as a sound playback function, an image playback function, etc.), etc.
- the memory 909 may include a volatile memory or a non-volatile memory, or the memory 909 may include both volatile and non-volatile memories.
- the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory.
- the volatile memory may be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDRSDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM) and a direct memory bus random access memory (DRRAM).
- the memory 909 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.
- the processor 910 may include one or more processing units; optionally, the processor 910 integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor 910.
- the radio frequency unit 901 is used to receive first information sent by a network side device, where the first information is used to instruct the terminal device to mark the location data;
- Processor 910 is used to label the location data according to the first information to obtain a data label; the data label is used to iteratively train the machine learning model, and the machine learning model is used to predict the location information of the terminal device.
- the first information includes a data screening condition and a tag content; the processor 910 is specifically configured to:
- the first position data is labeled according to the label content to obtain a data label.
- the label content includes at least one of the following:
- a location tag wherein the location tag is used to reflect the location coordinates of the terminal device, and the location coordinates include absolute location coordinates or relative location coordinates;
- An area tag where the area tag is used to reflect the area corresponding to the location data
- a timestamp tag wherein the timestamp tag is used to reflect the time when the location data is acquired
- a location marking error tag wherein the location marking error tag is used to reflect the error between the location tag and the actual location of the terminal device
- a terminal tag wherein the terminal tag includes a device type tag or an identity tag of the terminal device;
- a data acquisition tag wherein the data acquisition tag is used to indicate a method for acquiring the location data
- a motion status tag wherein the motion status tag is used to indicate that the terminal device acquires the location data The speed and direction of movement;
- a data quality level label wherein the data quality level label is used to indicate the data quality of the location data.
- the radio frequency unit 901 is further used to: send labeling response information to the network side device, the labeling response information is used to reflect whether the terminal device supports labeling of location data, or the labeling response information is used to reflect the tag content supported by the terminal device.
- the data screening condition includes at least one of the following:
- the position marking error of the position data is less than a preset threshold
- the area corresponding to the location data belongs to the preset area
- the area label corresponding to the location data is a preset area label
- the method for obtaining the location data is the preset method
- the acquisition time of the location data falls within the preset time range
- the data quality level of the location data is greater than the preset quality level
- Location data is not considered abnormal data
- the terminal tag of the terminal device is a preset terminal tag
- the terminal capability level of the terminal device is greater than the preset capability level
- the movement speed of the terminal device is less than a preset speed.
- processor 910 is specifically configured to:
- the preprocessed position data is labeled according to the first information to obtain a data label.
- the first information further includes a data preprocessing method, and the first position data is labeled according to the tag content.
- the processor 910 is further used to: preprocess the first position data according to the data preprocessing method to obtain the second position data;
- the processor 910 is specifically configured to: mark the second position data according to the tag content to obtain a data tag.
- the data preprocessing method includes at least one of the following: interpolation processing, smoothing processing, quantization processing, and normalization processing.
- the data preprocessing method includes interpolation processing, and the radio frequency unit 901 is further used to receive second information sent by the network side device, where the second information carries a label density indication, and the label density indication is used to indicate a label density of the data label;
- the processor 910 is specifically configured to perform interpolation processing on the first position data according to the label density indication to obtain second position data.
- the label density indication includes any one of the following:
- the number of tags of data tags is N1;
- the number of data tags within the first distance range is N2;
- the label density of data labels is N3/time unit, or N4/distance unit;
- N1, N2, N3 and N4 are all positive integers.
- the time unit includes any one of the following: hour, minute, second, millisecond, microsecond, frame, subframe, time slot, orthogonal frequency division multiplexing symbol.
- the data preprocessing method includes interpolation processing, and the radio frequency unit 901 is further used to receive third information sent by the network side device, where the third information carries an interpolation indication;
- the processor 910 is configured to perform interpolation processing on the first position data according to a target interpolation method corresponding to the interpolation indication to obtain second position data.
- the data preprocessing method includes smoothing processing, and the radio frequency unit 901 is further used to receive fourth information sent by the network side device, and the fourth information carries a smoothing indication;
- the processor 910 is configured to perform smoothing processing on the first position data according to a target smoothing method corresponding to the smoothing indication to obtain second position data.
- the data preprocessing method includes quantization processing
- the radio frequency unit 901 is further used to receive fifth information sent by the network side device, where the fifth information carries data quantization auxiliary information, and the data quantization auxiliary information includes at least one of the following:
- the quantization method includes uniform quantization or non-uniform quantization
- the processor 910 is configured to perform quantization processing on the first position data according to the data quantization auxiliary information.
- the second position data is obtained.
- the data preprocessing method includes normalization processing, and the radio frequency unit 901 is further used to receive sixth information sent by the network side device, where the sixth information carries a normalization indication;
- the processor 910 is used to normalize the first position data according to the target normalization method corresponding to the normalization indication to obtain second position data.
- the terminal device includes a conventional terminal device and/or a positioning reference unit.
- the network side device 1000 includes: an antenna 101, a radio frequency device 102, a baseband device 103, a processor 104 and a memory 105.
- the antenna 101 is connected to the radio frequency device 102.
- the radio frequency device 102 receives information through the antenna 101 and sends the received information to the baseband device 103 for processing.
- the baseband device 103 processes the information to be sent and sends it to the radio frequency device 102, and the radio frequency device 102 processes the received information and sends it out through the antenna 101.
- the method executed by the network-side device in the above embodiment may be implemented in the baseband device 103, which includes a baseband processor.
- the baseband device 103 may include, for example, at least one baseband board, on which a plurality of chips are arranged, as shown in FIG10 , wherein one of the chips is, for example, a baseband processor, which is connected to the memory 105 through a bus interface to call a program in the memory 105 and execute the network device operations shown in the above method embodiment.
- the network side device may also include a network interface 106, which is, for example, a common public radio interface (CPRI).
- a network interface 106 which is, for example, a common public radio interface (CPRI).
- CPRI common public radio interface
- the network side device 1000 of the embodiment of the present invention also includes: instructions or programs stored in the memory 105 and executable on the processor 104.
- the processor 104 calls the instructions or programs in the memory 105 to execute the aforementioned second aspect method and achieve the same technical effect. To avoid repetition, it will not be repeated here.
- the embodiment of the present application also provides a network side device.
- the network side device 1100 includes: a processor 1101, a network interface 1102 and a memory 1103.
- the network interface 1102 is, for example, a common public radio interface (CPRI).
- CPRI common public radio interface
- the network side device 1100 of the embodiment of the present invention also includes: instructions or programs stored in the memory 1103 and executable on the processor 1101.
- the processor 1101 calls the instructions or programs in the memory 1103 to execute the method of the aforementioned second aspect and achieve the same technical effect. To avoid repetition, it will not be repeated here.
- An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored.
- a program or instruction is stored.
- each process of the above-mentioned data labeling method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
- the processor is the processor in the terminal device described in the above embodiment.
- the readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.
- An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned data labeling method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.
- the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.
- the embodiments of the present application further provide a computer program/program product, which is stored in a storage medium.
- the computer program/program product is executed by at least one processor to implement the various processes of the above-mentioned data labeling method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.
- An embodiment of the present application also provides a data labeling system, including: a terminal device and a network side device, wherein the terminal can be used to execute the steps of the data labeling method described in the second aspect above, and the network side device can be used to execute the steps of the data labeling method described in the first aspect above.
- the technical solution of the present application can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM/RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.
- a storage medium such as ROM/RAM, a magnetic disk, or an optical disk
- a terminal which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
本申请公开了一种数据标注方法、装置、终端设备及网络侧设备,属于通信技术领域,本申请实施例的数据标注方法包括:终端设备接收网络侧设备发送的第一信息,并根据所述第一信息对位置数据进行标注,得到数据标签;其中,所述第一信息用于指示所述终端设备对位置数据进行标注;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
Description
相关申请的交叉引用
本申请主张在2022年10月31日在中国提交的申请号为202211350294.1的中国专利的优先权,其全部内容通过引用包含于此。
本申请属于通信技术领域,具体涉及一种数据标注方法、装置、终端设备及网络侧设备。
目前,人工智能(Artificial Intelligence,AI)技术在各领域得到了广泛应用,并起到了较好作用。由此可见,将AI技术融入到无线通信网络中,显著提升吞吐量、时延以及用户容量等技术指标,是未来的无线通信网络的重要任务。
其中,AI技术是一种数据驱动技术,可以利用有标签的训练数据对人工智能模型进行训练,然后将训练得到的人工智能模型应用到实际场景中,进行相应业务的处理。由此可见,在基于AI的应用中,AI模型的准确性很大程度上依赖于数据标签的质量。
然而,对于AI定位技术,数据标签只能由终端设备采集,由于终端能力和标签获取方法等方面存在差异,导致不同终端设备获取的数据标签的质量并不相同,网络侧设备基于多个终端设备上报的数据标签对AI模型进行训练时,AI模型的准确性存在浮动,AI模型的鲁棒性较差。
发明内容
本申请实施例提供一种数据标注方法、装置、终端设备及网络侧设备,能够解决终端设备获取的数据标签的质量不相同导致的AI模型的准确性浮动、鲁棒性差的问题。
第一方面,提供了一种数据标注方法,包括:
终端设备接收网络侧设备发送的第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注;
所述终端设备根据所述第一信息对位置数据进行标注,得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
第二方面,提供了一种数据标注方法,包括:
网络侧设备向终端设备发送第一信息,所述第一信息用于对位置数据进行标注得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
第三方面,提供了一种数据标注装置,包括:
第一信息接收模块,用于接收网络侧设备发送的第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注;
数据标注模块,用于根据所述第一信息对位置数据进行标注,得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
第四方面,提供了一种数据标注装置,包括:
第一信息发送模块,用于向终端设备发送第一信息,所述第一信息用于对位置数据进行标注得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
第五方面,提供了一种终端设备,该终端包括处理器和存储器,所述存储器存储可在所述处理器上运行的程序或指令,所述程序或指令被所述处理器执行时实现如第一方面所述的方法的步骤。
第六方面,提供了一种网络侧设备,包括处理器和存储器,所述存储器存储可在所述处理器上运行的程序或指令,所述程序或指令被所述处理器执行时实现如第二方
面所述的方法的步骤。
第七方面,提供了一种数据标注系统,包括:终端设备和网络侧设备,所述终端设备可用于执行如上述第一方面所述的数据标注方法的步骤,所述网络侧设备可用于执行如上述第二方面所述的数据标注方法的步骤。
第八方面,提供了一种可读存储介质,所述可读存储介质上存储程序或指令,所述程序或指令被处理器执行时实现如第一方面所述的数据标注方法的步骤,或者实现如第二方面所述的数据标注方法的步骤。
第九方面,提供了一种芯片,所述芯片包括处理器和通信接口,所述通信接口和所述处理器耦合,所述处理器用于运行程序或指令,实现如第一方面所述的数据标注方法,或实现如第二方面所述的数据标注方法。
第十方面,提供了一种计算机程序/程序产品,所述计算机程序/程序产品被存储在存储介质中,所述计算机程序/程序产品被至少一个处理器执行以实现如第一方面或者第二方面所述的数据标注方法的步骤。
在本申请实施例中,终端设备接收网络侧设备发送的第一信息,并根据所述第一信息对位置数据进行标注,得到数据标签;其中,所述第一信息用于指示所述终端设备对位置数据进行标注;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。本申请实施例中,网络侧设备通过第一信息对终端设备的数据标注过程进行统一管理和控制,各个终端设备均基于网络侧设备发送的第一信息的指示,对获取的位置数据进行标注,从而各个终端设备得到的数据标签的质量接近,有利于提升基于数据标签训练得到的机器学习模型的准确性和鲁棒性。
图1是本申请实施例可应用的一种无线通信系统的框图;
图2是本申请实施例中的一种数据标注方法的流程图;
图3是本申请实施例中神经网络的示意图;
图4是本申请实施例中神经网络的神经元的示意图;
图5是本申请实施例中的另一种数据标注方法的流程图;
图6是本申请实施例中的一种数据标注装置的结构框图;
图7是本申请实施例中的另一种数据标注装置的结构框图;
图8是本申请实施例中的一种通信设备的结构框图;
图9是本申请实施例中的一种终端设备的结构框图;
图10是本申请实施例中的一种网络侧设备的结构框图;
图11是本申请实施例中另一种网络侧设备的结构框图。
下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员所获得的所有其他实施例,都属于本申请保护的范围。
本申请的说明书和权利要求书中的术语“第一”、“第二”等是用于区别类似的对象,而不用于描述特定的顺序或先后次序。应该理解这样使用的术语在适当情况下可以互换,以便本申请的实施例能够以除了在这里图示或描述的那些以外的顺序实施,且“第一”、“第二”所区别的对象通常为一类,并不限定对象的个数,例如第一对象可以是一个,也可以是多个。此外,说明书以及权利要求中“和/或”表示所连接对象的至少其中之一,字符“/”一般表示前后关联对象是一种“或”的关系。
值得指出的是,本申请实施例所描述的技术不限于长期演进型(Long Term Evolution,LTE)/LTE的演进(LTE-Advanced,LTE-A)系统,还可用于其他无线通信系统,诸如码分多址(Code Division Multiple Access,CDMA)、时分多址(Time Division Multiple Access,TDMA)、频分多址(Frequency Division Multiple Access,FDMA)、正交频分多址(Orthogonal Frequency Division Multiple Access,OFDMA)、单载波频分多址(Single-carrier Frequency Division Multiple Access,SC-FDMA)和其他系统。本申请实施例中的术语“系统”和“网络”常被可互换地使用,所描述的技术既可用于以上提及的系统和无线电技术,也可用于其他系统和无线电技术。以下描述出于示例目的描述了新空口(New Radio,NR)系统,并且在以下
大部分描述中使用NR术语,但是这些技术也可应用于NR系统应用以外的应用,如第6代(6th Generation,6G)通信系统。
图l示出本申请实施例可应用的一种无线通信系统的框图。无线通信系统包括终端设备11和网络侧设备12。其中,终端设备11可以是手机、平板电脑(Tablet Personal Computer)、膝上型电脑(Laptop Computer)或称为笔记本电脑、个人数字助理(Personal Digital Assistant,PDA)、掌上电脑、上网本、超级移动个人计算机(ultra-mobile personal computer,UMPC)、移动上网装置(Mobile Internet Device,MID)、增强现实(augmented reality,AR)/虚拟现实(virtual reality,VR)设备、机器人、可穿戴式设备(Wearable Device)、车载设备(VUE)、行人终端(PUE)、智能家居(具有无线通信功能的家居设备,如冰箱、电视、洗衣机或者家具等)、游戏机、个人计算机(personal computer,PC)、柜员机或者自助机等终端侧设备,可穿戴式设备包括:智能手表、智能手环、智能耳机、智能眼镜、智能首饰(智能手镯、智能手链、智能戒指、智能项链、智能脚镯、智能脚链等)、智能腕带、智能服装等。需要说明的是,在本申请实施例并不限定终端设备11的具体类型。网络侧设备12可以包括接入网设备或核心网设备,其中,接入网设备12也可以称为无线接入网设备、无线接入网(Radio Access Network,RAN)、无线接入网功能或无线接入网单元。接入网设备12可以包括基站、WLAN接入点或WiFi节点等,基站可被称为节点B、演进节点B(eNB)、接入点、基收发机站(Base Transceiver Station,BTS)、无线电基站、无线电收发机、基本服务集(Basic Service Set,BSS)、扩展服务集(Extended Service Set,ESS)、家用B节点、家用演进型B节点、发送接收点(Transmitting Receiving Point,TRP)或所述领域中其他某个合适的术语,只要达到相同的技术效果,所述基站不限于特定技术词汇,需要说明的是,在本申请实施例中仅以NR系统中的基站为例进行介绍,并不限定基站的具体类型。核心网设备可以包含但不限于如下至少一项:核心网节点、核心网功能、移动管理实体(Mobility Management Entity,MME)、接入移动管理功能(Access and Mobility Management Function,AMF)、会话管理功能(Session Management Function,SMF)、用户平面功能(User Plane Function,UPF)、策略控制功能(Policy Control Function,PCF)、策略与计费规则功能单元(Policy and Charging Rules Function,PCRF)、边缘应用服务发现功能(Edge Application Server Discovery Function,EASDF)、统一数据管理(Unified Data Management,UDM),统一数据仓储(Unified Data Repository,UDR)、归属用户服务器(Home Subscriber Server,HSS)、集中式网络配置(Centralized network configuration,CNC)、网络存储功能(Network Repository Function,NRF),网络开放功能(Network Exposure Function,NEF)、本地NEF(Local NEF,或L-NEF)、绑定支持功能(Binding Support Function,BSF)、应用功能(Application Function,AF)等。需要说明的是,在本申请实施例中仅以NR系统中的核心网设备为例进行介绍,并不限定核心网设备的具体类型。
下面结合附图,通过一些实施例及其应用场景对本申请实施例提供的数据标注方法进行详细地说明。
第一方面,本申请实施例提供了一种数据标注方法。参照图2,示出了本申请实施例提供的一种数据标注方法的流程图。该方法应用于终端设备,如图2所示,该方法具体可以包括:
步骤201、终端设备接收网络侧设备发送的第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注。
步骤202、所述终端设备根据所述第一信息对位置数据进行标注,得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
需要说明的是,在本申请实施例中,所述终端设备包括常规终端设备和/或定位参考单元。其中,所述常规终端设备可以是图1中的终端设备11。所述定位参考单元(Positioning Reference Unit,PRU)可以执行定位测量,例如RSTD、RSRP、UE Rx-Tx时间差测量等,并将这些测量结果报告给定位服务器。此外,PRU可以向TRP发送定位参考信号(Positioning reference signal,PRS),使得TRP能够测量和报告来自已知位置的PRU的UL定位测量值,例如RTOA、UL-AoA、gNB Rx-Tx时间差等。位置服务器可以将PRU测量值与已知PRU位置处的预期测量值进行比较,以确定附近其他目标设备的校正项,然后基于该校正项来校正其他目标设备的DL和/
或UL位置测量值。
所述网络侧设备可以是图1中的接入网设备,如基站或接入网侧新定义的人工智能处理节点,还可以是图1中的核心网设备,如网络数据分析功能(Network Data Analytics Function,NWDAF)、定位管理功能(Location Management Function,LMF)、或者核心网侧新定义的处理节点,还可以是上述多个节点的组合。
在本申请实施例中,网络侧设备向终端设备发送第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注。可以理解的是,所述网络侧设备可以向多个终端设备发送第一信息,指示多个终端设备对各自的位置数据进行标注。
终端设备接收到网络侧设备发送的第一信息后,根据第一信息对位置数据进行标注,得到数据标签。其中,所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
在本申请的一种可能的应用场景中,具有定位功能的终端设备在移动过程中,可以不断获取位置数据,并在接收到网络侧设备发送的第一信息的情况下,按照第一信息的指示对位置数据进行标注,得到数据标签。终端设备将数据标签上报给网络侧设备,网络侧设备即可根据获得的数据标签对机器学习模型进行迭代训练,从而利用训练好的机器学习模型执行目标业务,如定位业务、信息推送业务等等。或者,网络侧设备完成对机器学习模型的训练之后,将训练好的机器学习模型发送至终端设备,由终端设备利用训练好的机器学习模型执行目标业务。
在本申请的另一种可能的应用场景中,终端设备按照第一信息指示对位置数据进行标注得到数据标签之后,利用数据标签在本地对机器学习模型进行迭代训练,并利用训练好的机器学习模型执行目标业务。或者,终端设备将训练好的机器学习模型发送给网络侧设备,由网络侧设备利用训练好的机器学习模型执行目标业务。
由上述步骤201至202可知,在本申请实施例中,各个终端设备均基于网络侧设备发送的第一信息的指示,对获取的位置数据进行标注,从而各个终端设备得到的数据标签的质量接近,有利于提升基于数据标签训练得到的机器学习模型的准确性和鲁棒性。
需要说的是,本申请实施例中的机器学习模型可以为人工智能模型,例如全连接神经网络、卷积神经网络、决策树、支持向量机、贝叶斯分类器中的任意一种。以神经网络模型为例,其示意图可如图3所示。如图3所示,神经网络可以包括一个或多个输入层、一个或多个隐层和一个输出层。待处理数据[X1,X2…Xn]分别从对应的输入层输入至神经网络中,经过输入层、隐层和输出层的处理,得到输出结果Y。另外,神经网络由神经元组成,神经元的示意图如图4所示。其中在图4中,a1,a2,…aK表示输入,w表示权值(即乘性系数),b表示偏置(即加性系数),σ(.)表示激活函数。常见的激活函数包括Sigmoid(将变量映射到0、1之间)、tanh(对Sigmoid的平移和收缩)、线性整流函数/修正线性单元(Rectified Linear Unit,ReLU)等。
此外,以神经网络模型为例,对模型训练的过程进行如下介绍:
其中,神经网络的参数可以通过梯度优化算法进行优化。梯度优化算法是一类最小化或者最大化目标函数(有时候也称为损失函数)的算法,而目标函数往往是模型参数和数据的数学组合。例如给定数据X和其对应的标签Y,可以构建一个神经网络模型f(.),则根据输入x就可以得到预测输出f(x),并且可以计算出预测值和真实值之间的差距(f(x)-Y),这个就是损失函数。其中,梯度优化算法的优化目标是找到合适的w(即权值)和b(即偏置)使上述的损失函数的值达到最小,而损失值越小,则说明模型越接近于真实情况。
目前常见的优化算法,基本都是基于误差反向传播(error Back Propagation,BP)算法。BP算法的基本思想是,学习过程由信号的正向传播与误差的反向传播两个过程组成。正向传播时,输入样本从输入层传入,经各隐层逐层处理后,传向输出层。若输出层的实际输出与期望的输出不符,则转入误差的反向传播阶段。误差反传则是将输出误差以某种形式通过隐层向输入层逐层反传,并将误差分摊给各层的所有单元,从而获得各层单元的误差信号,此误差信号即作为修正各单元权值的依据。这种信号正向传播与误差反向传播的各层权值调整过程,是周而复始地进行的。其中,权值不断调整的过程,也就是网络的学习训练过程。此过程一直进行到网络输出的误
差减少到可接受的程度,或进行到预先设定的学习次数为止。
另外,常见的优化算法有梯度下降(Gradient Descent)、随机梯度下降(Stochastic Gradient Descent,SGD)、小批量梯度下降(mini-batch gradient descent)、动量法(Momentum)、Nesterov(发明者的名字,具体为带动量的随机梯度下降)自适应梯度下降(ADAptive GRADient descent,Adagrad)、Adagrad的扩展算法(Adadelta)、均方根误差降速(root mean square prop,RMSprop)、自适应动量估计(Adaptive Moment Estimation,Adam)等。
这些优化算法在误差反向传播时,都是根据损失函数得到的误差/损失,对当前神经元求导数/偏导,加上学习速率、之前的梯度/导数/偏导等影响,得到梯度,将梯度传给上一层。
在本申请实施例中,将上述模型训练方法应用于无线通信领域,由终端设备和网络侧设备进行交互来实现对机器学习模型的训练。
可选地,所述第一信息包括数据筛选条件和标签内容;所述终端设备根据所述第一信息对位置数据进行标注,得到数据标签,包括:
所述终端设备根据所述数据筛选条件对位置数据进行筛选,得到第一位置数据;
所述终端设备按照所述标签内容对所述第一位置数据进行标注,得到数据标签。
在本申请实施例中,网络侧设备可以通过第一信息中携带的数据筛选条件和标签内容,指示终端设备对不符合数据筛选条件的位置数据进行筛选,并按照标签内容对位置数据进行标注。具体地,终端设备接收到网络侧设备发送的第一信息后,根据数据筛选条件对位置数据进行筛选,得到第一位置数据。可以理解的是,所述第一位置数据为符合数据筛选条件的位置数据。然后,终端设备再按照标签内容对第一位置数据进行标注,得到数据标签。所述数据标签为标注后的符合数据筛选条件的位置数据。相比于直接对未筛选的位置数据进行标注,本申请实施例中,各个终端设备按照网络侧设备指示的数据筛选条件对位置数据进行筛选后再标注,提升了最终得到的数据标签的质量,且各个终端设备得到的数据标签均符合数据筛选条件,数据质量相近,有利于提升基于数据标签训练得到的机器学习模型的准确性和鲁棒性。
可选地,所述标签内容包括以下至少一项:
A1、位置标签,所述位置标签用于反映所述终端设备的位置坐标,所述位置坐标包括绝对位置坐标或相对位置坐标;
A2、区域标签,所述区域标签用于反映所述位置数据对应的区域;
A3、时间戳标签,所述时间戳标签用于反映所述位置数据的获取时间;
A4、位置标注误差标签,所述位置标识误差标签用于反映所述位置标签与所述终端设备的真实位置之间的误差;
A5、终端标签,所述终端标签包括所述终端设备的设备类型标签或身份标签;
A6、数据获取标签,所述数据获取标签用于指示所述位置数据的获取方式;
A7、运动状态标签,所述运动状态标签用于指示所述终端设备在获取所述位置数据时的运动速度及方向;
A8、数据质量等级标签,所述数据质量等级标签用于指示所述位置数据的数据质量。
在本申请实施例中,第一信息中携带的标签内容可以包含A1至A8中的至少一项。其中,对于A1项,位置标签可以是用于反映终端位置的位置坐标;该位置坐标可以是绝对位置坐标,例如用经纬度表示的位置坐标,也可以是相对位置坐标,例如以网络侧设备为参照点,终端设备相对于该网络侧设备的相对位置坐标。
对于A2项,区域标签可以是终端设备获取位置数据时所处区域的行政区划代码,也可以是终端设备获取位置数据时所处区域的编码信息,该编码信息可以是根据区域编码算法或空间索引算法得到的区域编码或空间索引。示例性地,可以预先选定一个目标区域作为用于采集位置数据的区域,该目标区域可以是全国、省、市、区(县)等任一级别的区域,也可以根据需要利用机器学习模型执行的目标业务确定目标区域。接下来,对目标区域进行区域划分并编码,得到划分的各个单位区域的编码信息。终端设备可以将获取位置数据时所处的单位区域的编码信息作为位置数据的位置标签。需要说明的是,区域标签的类型及区域标签的确定方法,可以由网络侧设备预先设定,也可以由协议规定,本申请实施例对此不做具体限定。
对于A3项,时间戳标签可以是位置数据的获取时间对应的时间戳。
对于A4项,位置标注误差标签用于反映位置标签与终端设备的真实位置之间的
误差。所述位置标注误差标签可以是实际误差值,如位置标签反映的终端设备的位置坐标,与终端设备的真实位置坐标的差值;所述位置标注误差标签也可以是位置标签反映的终端设备的位置坐标与终端设备的真实位置坐标的差值对应的误差等级,等等。
对于A5项,终端标签可以是终端设备的设备类型标签,例如,若终端设备的设备类型为常规终端设备,则该终端设备的设备类型标签为“UE”,若终端设备的设备类型为定位参考单元,则该终端设备的设备类型标签为“PRU”,等等。终端标签还可以是终端设备的身份标签,如终端标识,该终端标识可以是网络侧设备分配的,用于区分网络中的各个终端设备。
对于A6项,数据获取标签用于指示位置数据的获取方式,可以对位置数据的获取方式进行编码,得到数据获取标签。其中,位置数据的获取方式可以包括基于终端设备自身的定位系统采集的定位数据获取、基于定位参考信号(Positioning reference signal,PRS)获取,等等。
对于A7项,运动状态标签可以包括终端设备在获取位置数据时对应的运动速度标签和方向标签,其中,所述运动速度标签可以是运动速度的具体数值,也可以是运动速度对应的编码信息;所述方向标签可以是终端设备在获取位置数据时对应的运动方向或速度方向的编码信息。
对于A8项,数据质量等级标签用于反映位置数据的数据质量,可以根据位置数据的字段缺失情况、位置标注误差标签等确定位置数据的数据质量等级。其中,可以根据位置数据是否包含A1至A7中各项对应的信息来确定位置数据的字段缺失情况,位置数据中包含的信息对应A1至A7的项数越大,说明位置数据的字段越完整。可以理解的是,位置数据的字段越完整、位置标注误差越小,位置数据的数据质量等级越高。
可选地,所述方法还包括:所述终端设备向所述网络侧设备发送标注响应信息,所述标注响应信息用于反映所述终端设备是否支持对位置数据进行标注,或者,所述标注响应信息用于反映所述终端设备支持的标签内容。
在本申请实施例中,终端设备在接收到网络侧设备发送的第一信息之后,可以向网络侧设备发送标注响应信息,以反映是否支持对位置数据进行标注。作为一种示例,标注响应信息可以包括第一响应信息或第二响应信息,其中,第一响应信息用于指示终端设备支持对位置数据进行标注,所述第一响应信息可以为“1”;第二响应信息用于指示终端设备不支持对位置数据进行标注,所述第二响应信息可以为“0”。
如果第一信息中携带标签内容,用于指示A1至A8中的至少一项,则终端设备向网络侧设备发送的标注响应信息还可以用于反映终端设备支持的标签内容。作为一种示例,标注响应信息中可以携带8个标注响应标签,分别对应A1至A8,如果终端设备支持A1至A8中某一项,则该项对应的标注响应标签为第一标签,第一标签用于指示终端设备支持对相应的标签内容进行标注,该第一标签可以为“1”;如果终端设备不支持A1至A8中某一项,则该项对应的标注响应标签为第二标签,第二标签用于指示终端设备不支持对相应的标签内容进行标注,该第二标签可以为“0”。
可选地,所述数据筛选条件包括以下至少一项:
B1、位置数据的位置标注误差小于预设阈值;
B2、位置数据对应的区域属于预设区域;
B3、位置数据对应的区域标签为预设区域标签;
B4、位置数据的获取方式为预设获取方式;
B5、位置数据的获取时间属于预设时间范围;
B6、位置数据的数据质量等级大于预设质量等级;
B7、位置数据不存在字段缺失;
B8、位置数据不属于异常数据;
B9、所述终端设备的终端标签为预设终端标签;
B10、所述终端设备的终端能力等级大于预设能力等级;
B11、所述终端设备的运动速度小于预设速度。
其中,预设阈值、预设区域、预设区域标签、预设获取方式、预设时间范围、预设质量等级、预设终端标签、预设能力等级、预设速度可以由网络侧设备设定,也可以由协议规定。
位置数据是否存在字段缺失可以根据前述A1至A8项确定。若位置数据中包含
A1至A8各项对应的信息,则说明位置数据不存在字段缺失;反之,若位置数据中不包含A1至A8中任一项对应的信息,则说明位置数据存在字段缺失。
位置数据是否属于异常数据可以位置数据在终端设备采集的所有位置数据中的分布情况进行判断。例如,如果根据该位置数据对应的样本点在终端设备采集的所有位置数据对应的样本点中属于离群点,则说明该位置数据为异常数据。
在本申请实施例中,终端设备在接收到第一信息之后,可以按照上述B1至B11中至少一项对应的数据筛选条件对位置数据进行筛选,再对筛选后的位置数据进行标注,可以提升最终得到的数据标签的质量,进而有利于提升基于数据标签训练得到的机器学习模型的准确性和鲁棒性。或者,终端设备在接收到第一信息之后,也可以先对位置数据进行标注,再按照上述B1至B11中至少一项对应的数据筛选条件对标注后的位置数据进行筛选,得到最终的数据标签。需要说明的是,本申请实施例对数据筛选和数据标注的先后顺序不做具体限定。
可选地,步骤202所述终端设备根据所述第一信息对位置数据进行标注,得到数据标签,包括:
步骤S11、所述终端设备根据预先设定或协议规定的数据预处理方式对位置数据进行预处理,得到预处理后的位置数据;
步骤S12、所述终端设备根据所述第一信息对预处理后的位置数据进行标注,得到数据标签。
在本申请实施例中,终端设备在接收到第一信息之后,可以先根据数据预处理方式对位置数据进行预处理,然后再对预处理后的位置数据进行标注。其中,所述数据预处理方式可以由终端设备本地预先设定,也可以由协议规定。所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
可选地,所述第一信息还包括数据预处理方式,所述终端设备按照所述标签内容对所述第一位置数据进行标注,得到数据标签之前,所述方法还包括:
步骤S21、所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据;
步骤202所述终端设备按照所述标签内容对所述第一位置数据进行标注,得到数据标签,包括:
步骤S22、所述终端设备按照所述标签内容对所述第二位置数据进行标注,得到数据标签。
在本申请实施例中,所述数据预处理方式可以由网络侧设备指定,且该数据预处理方式被携带在第一信息中。终端设备在接收到第一信息之后,可以先根据第一信息中携带的数据筛选条件对位置数据进行筛选,得到符合数据筛选条件的第一位置数据;然后,按照数据预处理方式对第一位置数据进行预处理,得到第二位置数据;最后,按照第一信息中的标签内容对第二位置数据进行标注,得到数据标签。其中,所述数据筛选条件包括前述B1至B11中的至少一项;所述标签内容包括前述A1至A8中的至少一项;所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
可选地,所述数据预处理方式包括插值处理,步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:
步骤S31、所述终端设备接收所述网络侧设备发送的第二信息,所述第二信息中携带标签密度指示,所述标签密度指示用于指示所述数据标签的标签密度;
步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:
步骤S32、所述终端设备根据所述标签密度指示对所述第一位置数据进行插值处理,得到第二位置数据。
在本申请实施例中,如果第一信息中携带的数据预处理方式包括插值处理,则网络侧设备可以向终端设备发送携带标签密度指示的第二信息,其中,标签密度指示用于指示数据标签的标签密度。
可选地,所述标签密度指示包括以下任一项:
C1、第一时间范围内,数据标签的标签数目为N1;
C2、第一距离范围内,数据标签的标签数目为N2;
C3、数据标签的标签密度为N3/时间单位,或者,N4/距离单位。
其中,N1、N2、N3和N4均为正整数。
可选地,所述时间单位包括以下任一项:小时、分钟、秒、毫秒、微秒、帧、子帧、时隙、正交频分复用(Orthogonal frequency-division multiplexing,OFDM)符号。所述距离单位可以包括以下任一项:千米、米、分米、厘米。
终端设备在接收到网络侧设备发送的第二信息之后,根据所述标签密度指示对所述第一位置数据进行插值处理,得到第二位置数据,所述第二位置数据的数据密度符合标签密度指示,也即C1至C3中的某一项。其中,所述插值处理可以包括但不限于线性插值、二次插值。
作为一种示例,假设终端设备获取的第一位置数据为:
t1(x1,y1),t3(x3,y3),t5(x5,y5)....,采用不同的插值处理,可以得到不同的第二位置数据。以利用插值处理得到t2时刻的位置数据为例:
对于线性插值,获取的t2时刻的位置数据为:
对于二次插值,也即以相邻三个样本点做插值,获取的t2时刻的位置数据为:
按照上述公式(1)或公式(2)所示的插值方式对第一位置数据进行插值处理,即可得到符合标签密度指示的第二位置数据。
可选地,所述数据预处理方式包括插值处理,步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:
步骤S41、所述终端设备接收所述网络侧设备发送的第三信息,所述第三信息中携带插值指示;
步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:
步骤S42、所述终端设备按照所述插值指示对应的目标插值方法对所述第一位置数据进行插值处理,得到第二位置数据。
在本申请实施例中,当第一信息中携带的数据预处理方式包括插值处理时,网络侧设备可以通过第三信息向终端设备发送插值指示,以指示插值处理所需采用的目标插值方法。终端设备接收到第三信息之后,按照插值指示对应的目标插值方法对第一位置数据进行插值处理。其中,所述目标插值方法可以包括但不限于线性插值、二次插值。
可选地,所述数据预处理方式包括平滑处理,步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:
步骤S51、所述终端设备接收所述网络侧设备发送的第四信息,所述第四信息中携带平滑指示;
步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:
步骤S52、所述终端设备按照所述平滑指示对应的目标平滑方法对所述第一位置数据进行平滑处理,得到第二位置数据。
在本申请实施例中,当第一信息中携带的数据预处理方式包括平滑处理时,网络侧设备可以通过第四信息向终端设备发送平滑指示,以指示平滑处理所需采用的目标平滑方法。终端设备接收到第四信息之后,按照平滑指示对应的目标平滑方法对第一位置数据进行平滑处理。其中,所述目标平滑方法可以包括但不限于S-G(Savitzky-Golay)滤波平滑方法、滑动平均(moving average)方法、滑动平均滤波方法。
以滑动平均滤波方法为例,假设滑动窗内包含2N+1个第一位置数据,对t时刻的第一位置数据(x,y)做平滑处理,则t时刻平滑后的第二位置数据为:
通过平滑处理,可以去除位置数据中的噪声,进而提升对平滑后的位置数据进行标注得到的数据标签的质量。
可选地,所述数据预处理方式包括量化处理,步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:
步骤S61、所述终端设备接收所述网络侧设备发送的第五信息,所述第五信息中携带数据量化辅助信息;
步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:
步骤S62、所述终端设备根据所述数据量化辅助信息对所述第一位置数据进行量化处理,得到第二位置数据。
其中,所述数据量化辅助信息包括以下至少一项:
D1、量化参考点的标识和位置;
D2、量化方式,所述量化方式包括均匀量化或非均匀量化。
对于D1项,量化参考点的标识用于指示量化参考点的身份,每一个量化参考点均对应一个身份标识和一个位置坐标。例如,量化参考点1的位置为(0,0)。
对于D2项,均匀量化是指对输入信号的取值域等间隔分割的量化,又称为线性编码,起特点是歌量化区间的宽度(即宽阶)相同。均匀量化主要指示量化间隔,比如量化间隔为10。非均匀量化是指一种在输入信号的动态范围内量化间隔不相等的量化,在非均匀量化中,对于信号取值小的区间,其量化间隔也小,反之,信号取值越大,其量化间隔就大。非均匀量化可以指示量化码本(codebook),比如x维度的量化码本为{0,10,30,40,45},y维度的量化码本为{0,5,20,40}。同时,量化码本可以反映量化间隔,如对应地,x维度的量化码本为{10,20,10,5},y维度的量化码本为{5,15,20},假设终端设备1的位置为(8,12),则映射到量化码本中,x维度的量化间隔为10,y维度的量化间隔为10。
在本申请实施例中,当第一信息中携带的数据预处理方式包括量化处理时,网络侧设备可以通过第五信息向终端设备发送数据量化辅助信息。终端设备接收到第五信息之后,按照数据量化辅助信息对第一位置数据进行量化处理。
可选地,所述数据预处理方式包括归一化处理,步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:
步骤S71、所述终端设备接收所述网络侧设备发送的第六信息,所述第六信息中携带归一化指示;
步骤S21所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:
步骤S72、所述终端设备按照所述归一化指示对应的目标归一化方法对所述第一位置数据进行归一化处理,得到第二位置数据。
在本申请实施例中,当第一信息中携带的数据预处理方式包括归一化处理时,网络侧设备可以通过第六信息向终端设备发送归一化指示,以指示归一化处理所需采用的目标归一化方法。终端设备接收到第六信息之后,按照归一化指示对应的目标归一化方法对第一位置数据进行归一化处理。其中,所述目标归一化方法可以包括但不限于最大最小归一化(X-Min/(Max-Min))方法、0均值归一化(X-Mean/(Standard deviation))方法。
第二方面,本申请实施例提供了另一种数据标注方法。参照图5,示出了本申请实施例提供的另一种数据标注方法的流程图。该方法应用于网络侧设备,如图5所示,该方法具体可以包括:
步骤501、网络侧设备向终端设备发送第一信息,所述第一信息用于指示终端设
备对位置数据进行标注得到数据标签。
步骤502、所述网络侧设备接收所述终端设备发送的数据标签;
步骤503、所述网络侧设备根据所述数据标签对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
需要说明的是,在本申请实施例中,所述终端设备包括常规终端设备和/或定位参考单元。其中,所述常规终端设备可以是图1中的终端设备11。所述网络侧设备可以是图1中的接入网设备,如基站或接入网侧新定义的人工智能处理节点,还可以是图1中的核心网设备,如网络数据分析功能(Network Data Analytics Function,NWDAF)、定位管理功能(Location Management Function,LMF)、或者核心网侧新定义的处理节点,还可以是上述多个节点的组合。
在本申请实施例中,网络侧设备向终端设备发送第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注。可以理解的是,所述网络侧设备可以向多个终端设备发送第一信息,指示多个终端设备对各自的位置数据进行标注。
终端设备接收到网络侧设备发送的第一信息后,根据第一信息对位置数据进行标注,得到数据标签。其中,所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
在本申请的一种可能的应用场景中,具有定位功能的终端设备在移动过程中,可以不断获取位置数据,并在接收到网络侧设备发送的第一信息的情况下,按照第一信息的指示对位置数据进行标注,得到数据标签。终端设备将数据标签上报给网络侧设备,网络侧设备即可根据获得的数据标签对机器学习模型进行迭代训练,从而利用训练好的机器学习模型执行目标业务,如定位业务、信息推送业务等等。或者,网络侧设备完成对机器学习模型的训练之后,将训练好的机器学习模型发送至终端设备,由终端设备利用训练好的机器学习模型执行目标业务。
在本申请的另一种可能的应用场景中,终端设备按照第一信息指示对位置数据进行标注得到数据标签之后,利用数据标签在本地对机器学习模型进行迭代训练,并利用训练好的机器学习模型执行目标业务。或者,终端设备将训练好的机器学习模型发送给网络侧设备,由网络侧设备利用训练好的机器学习模型执行目标业务。
可选地,所述第一信息包括数据筛选条件和标签内容;其中,所述数据筛选条件用于筛选所述位置数据,以滤除不符合所述数据筛选条件的位置数据;所述标签内容用于按照所述标签内容对位置数据进行标注。
可选地,所述标签内容包括以下至少一项:
A1、位置标签,所述位置标签用于反映所述终端设备的位置坐标,所述位置坐标包括绝对位置坐标或相对位置坐标;
A2、区域标签,所述区域标签用于反映所述位置数据对应的区域;
A3、时间戳标签,所述时间戳标签用于反映所述位置数据的获取时间;
A4、位置标注误差标签,所述位置标识误差标签用于反映所述位置标签与所述终端设备的真实位置之间的误差;
A5、终端标签,所述终端标签包括所述终端设备的设备类型标签或身份标签;
A6、数据获取标签,所述数据获取标签用于指示所述位置数据的获取方式;
A7、运动状态标签,所述运动状态标签用于指示所述终端设备在获取所述位置数据时的运动速度及方向;
A8、数据质量等级标签,所述数据质量等级标签用于指示所述位置数据的数据质量。
对于A1项至A8项,可以参照前述第一方面中关于A1至A8项的相关描述,本申请实施例在此不做进一步赘述。
可选地,所述方法还包括:所述网络侧设备接收所述终端设备发送的标注响应信息,所述标注响应信息用于反映所述终端设备是否支持对位置数据进行标注,或者,所述标注响应信息用于反映所述终端设备支持的标签内容。
在本申请实施例中,终端设备在接收到网络侧设备发送的第一信息之后,可以向网络侧设备发送标注响应信息,以反映是否支持对位置数据进行标注。作为一种示例,标注响应信息可以包括第一响应信息或第二响应信息,其中,第一响应信息用于指示终端设备支持对位置数据进行标注,所述第一响应信息可以为“1”;第二响应信息用于指示终端设备不支持对位置数据进行标注,所述第二响应信息可以为“0”。
如果第一信息中携带标签内容,用于指示A1至A8中的至少一项,则终端设备
向网络侧设备发送的标注响应信息还可以用于反映终端设备支持的标签内容。作为一种示例,标注响应信息中可以携带8个标注响应标签,分别对应A1至A8,如果终端设备支持A1至A8中某一项,则该项对应的标注响应标签为第一标签,第一标签用于指示终端设备支持对相应的标签内容进行标注,该第一标签可以为“1”;如果终端设备不支持A1至A8中某一项,则该项对应的标注响应标签为第二标签,第二标签用于指示终端设备不支持对相应的标签内容进行标注,该第二标签可以为“0”。
可选地,所述数据筛选条件包括以下至少一项:
B1、位置数据的位置标注误差小于预设阈值;
B2、位置数据对应的区域属于预设区域;
B3、位置数据对应的区域标签为预设区域标签;
B4、位置数据的获取方式为预设获取方式;
B5、位置数据的获取时间属于预设时间范围;
B6、位置数据的数据质量等级大于预设质量等级;
B7、位置数据不存在字段缺失;
B8、位置数据不属于异常数据;
B9、所述终端设备的终端标签为预设终端标签;
B10、所述终端设备的终端能力等级大于预设能力等级;
B11、所述终端设备的运动速度小于预设速度。
对于B1项至B11项,可以参照前述第一方面中关于B1至B11项的相关描述,本申请实施例在此不做进一步赘述。
可选地,所述方法还包括:所述网络侧设备接收所述终端设备发送的数据标签;所述网络侧设备按照预先设定的数据预处理方式对所述数据标签进行预处理,得到目标数据标签。
在本申请实施例中,可以由网络侧设备执行数据预处理操作。具体地,网络侧设备接收终端设备发送的数据标签,所述数据标签是终端设备根据第一信息的指示进行标注得到;网络侧设备按照预先设定的数据预处理方式对接收的数据标签进行预处理,得到目标数据标签。其中,所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
可选地,所述第一信息还包括数据预处理方式,所述数据预处理方式用于对位置数据进行预处理。
在本申请的一种实施例中,网络侧设备可以在第一信息中携带数据预处理方式,指示终端设备按照数据预处理方式对位置数据进行预处理。可选地,所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
可选地,所述数据预处理方式包括插值处理,所述方法还包括:所述网络侧设备向所述终端设备发送第二信息,所述第二信息中携带标签密度指示,所述标签密度指示用于指示所述数据标签的标签密度。
在本申请实施例中,如果第一信息中携带的数据预处理方式包括插值处理,则网络侧设备可以向终端设备发送携带标签密度指示的第二信息,其中,标签密度指示用于指示数据标签的标签密度。
可选地,所述标签密度指示包括以下任一项:
C1、第一时间范围内,数据标签的标签数目为N1;
C2、第一距离范围内,数据标签的标签数目为N2;
C3、数据标签的标签密度为N3/时间单位,或者,N4/距离单位。
其中,N1、N2、N3和N4均为正整数。
可选地,所述时间单位包括以下任一项:小时、分钟、秒、毫秒、微秒、帧、子帧、时隙、正交频分复用符号。所述距离单位可以包括以下任一项:千米、米、分米、厘米。
可选地,所述数据预处理方式包括插值处理,所述方法还包括:所述网络侧设备向所述终端设备发送第三信息,所述第三信息中携带插值指示,所述插值指示用于位置数据的插值处理。
在本申请实施例中,当第一信息中携带的数据预处理方式包括插值处理时,网络侧设备可以通过第三信息向终端设备发送插值指示,以指示插值处理所需采用的目标插值方法。终端设备接收到第三信息之后,按照插值指示对应的目标插值方法对第一位置数据进行插值处理。其中,所述目标插值方法可以包括但不限于线性插值、二次
插值。
可选地,所述数据预处理方式包括平滑处理,所述方法还包括:所述网络侧设备向所述终端设备发送第四信息,所述第四信息中携带平滑指示,所述平滑指示用于位置数据的平滑处理。
在本申请实施例中,当第一信息中携带的数据预处理方式包括平滑处理时,网络侧设备可以通过第四信息向终端设备发送平滑指示,以指示平滑处理所需采用的目标平滑方法。终端设备接收到第四信息之后,按照平滑指示对应的目标平滑方法对第一位置数据进行平滑处理。其中,所述目标平滑方法可以包括但不限于S-G(Savitzky-Golay)滤波平滑方法、滑动平均(moving average)方法、滑动平均滤波方法。
可选地,所述数据预处理方式包括量化处理,所述方法还包括:所述网络侧设备向所述终端设备发送第五信息,所述第五信息中携带数据量化辅助信息,所述数据量化辅助信息包括以下至少一项:
D1、量化参考点的标识和位置;
D2、量化方式,所述量化方式包括均匀量化或非均匀量化。
对于D1项和D2项,可以参照前述第一方面中对于D1和D2项的相关描述,本申请实施例在此不做进一步赘述。
在本申请实施例中,当第一信息中携带的数据预处理方式包括量化处理时,网络侧设备可以通过第五信息向终端设备发送数据量化辅助信息。终端设备接收到第五信息之后,按照数据量化辅助信息对第一位置数据进行量化处理。
可选地,所述数据预处理方式包括归一化处理,所述方法还包括:所述网络侧设备向所述终端设备发送第六信息,所述第六信息中携带归一化指示,所述归一化指示用于位置数据的归一化处理。
在本申请实施例中,当第一信息中携带的数据预处理方式包括归一化处理时,网络侧设备可以通过第六信息向终端设备发送归一化指示,以指示归一化处理所需采用的目标归一化方法。终端设备接收到第六信息之后,按照归一化指示对应的目标归一化方法对第一位置数据进行归一化处理。其中,所述目标归一化方法可以包括但不限于最大最小归一化(X-Min/(Max-Min))方法、0均值归一化(X-Mean/(Standard deviation))方法。
在本申请实施例提供的数据标注方法中,各个终端设备均基于网络侧设备发送的第一信息的指示,对获取的位置数据进行标注,从而各个终端设备得到的数据标签的质量接近,有利于提升基于数据标签训练得到的机器学习模型的准确性和鲁棒性。
本申请实施例提供的数据标注方法,执行主体可以为数据标注装置。本申请实施例中以数据标注装置执行数据标注方法为例,说明本申请实施例提供的数据标注装置。
第三方面,本申请实施例提供了一种数据标注装置。参照图6,示出了本申请实施例提供的一种数据标注装置的结构框图,该装置可以应用于终端设备。如图6所示,该装置具体可以包括:
第一信息接收模块601,用于接收网络侧设备发送的第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注;
数据标注模块602,用于根据所述第一信息对位置数据进行标注,得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
可选地,所述第一信息包括数据筛选条件和标签内容;所述数据标注模块,包括:
数据筛选子模块,用于根据所述数据筛选条件对位置数据进行筛选,得到第一位置数据;
第一标注子模块,用于按照所述标签内容对所述第一位置数据进行标注,得到数据标签。
可选地,所述标签内容包括以下至少一项:
位置标签,所述位置标签用于反映所述终端设备的位置坐标,所述位置坐标包括绝对位置坐标或相对位置坐标;
区域标签,所述区域标签用于反映所述位置数据对应的区域;
时间戳标签,所述时间戳标签用于反映所述位置数据的获取时间;
位置标注误差标签,所述位置标识误差标签用于反映所述位置标签与所述终端设
备的真实位置之间的误差;
终端标签,所述终端标签包括所述终端设备的设备类型标签或身份标签;
数据获取标签,所述数据获取标签用于指示所述位置数据的获取方式;
运动状态标签,所述运动状态标签用于指示所述终端设备在获取所述位置数据时的运动速度及方向;
数据质量等级标签,所述数据质量等级标签用于指示所述位置数据的数据质量。
可选地,所述装置还包括:标注响应信息发送模块,用于向所述网络侧设备发送标注响应信息,所述标注响应信息用于反映所述终端设备是否支持对位置数据进行标注,或者,所述标注响应信息用于反映所述终端设备支持的标签内容。
可选地,所述数据筛选条件包括以下至少一项:
位置数据的位置标注误差小于预设阈值;
位置数据对应的区域属于预设区域;
位置数据对应的区域标签为预设区域标签;
位置数据的获取方式为预设获取方式;
位置数据的获取时间属于预设时间范围;
位置数据的数据质量等级大于预设质量等级;
位置数据不存在字段缺失;
位置数据不属于异常数据;
所述终端设备的终端标签为预设终端标签;
所述终端设备的终端能力等级大于预设能力等级;
所述终端设备的运动速度小于预设速度。
可选地,所述数据标注模块,包括:
第一预处理子模块,用于根据预先设定或协议规定的数据预处理方式对位置数据进行预处理,得到预处理后的位置数据;
第二标注子模块,用于根据所述第一信息对预处理后的位置数据进行标注,得到数据标签。
可选地,所述第一信息还包括数据预处理方式,所述数据标注模块还包括:
第二预处理子模块,用于所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据;
所述第一标注子模块,包括:
数据标注单元,用于按照所述标签内容对所述第二位置数据进行标注,得到数据标签。
可选地,所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
可选地,所述数据预处理方式包括插值处理,所述装置还包括:
第二信息接收模块,用于接收所述网络侧设备发送的第二信息,所述第二信息中携带标签密度指示,所述标签密度指示用于指示所述数据标签的标签密度;
所述第二预处理子模块,包括:
第一插值处理单元,用于根据所述标签密度指示对所述第一位置数据进行插值处理,得到第二位置数据。
可选地,所述标签密度指示包括以下任一项:
第一时间范围内,数据标签的标签数目为N1;
第一距离范围内,数据标签的标签数目为N2;
数据标签的标签密度为N3/时间单位,或者,N4/距离单位;
其中,N1、N2、N3和N4均为正整数。
可选地,所述时间单位包括以下任一项:小时、分钟、秒、毫秒、微秒、帧、子帧、时隙、正交频分复用符号。
可选地,所述数据预处理方式包括插值处理,所述装置还包括:
第三信息接收模块,用于接收所述网络侧设备发送的第三信息,所述第三信息中携带插值指示;
所述第二预处理子模块,包括:
第二插值处理单元,用于按照所述插值指示对应的目标插值方法对所述第一位置数据进行插值处理,得到第二位置数据。
可选地,所述数据预处理方式包括平滑处理,所述装置还包括:
第四信息接收模块,用于接收所述网络侧设备发送的第四信息,所述第四信息中携带平滑指示;
所述第二预处理子模块,包括:
平滑处理单元,用于按照所述平滑指示对应的目标平滑方法对所述第一位置数据进行平滑处理,得到第二位置数据。
可选地,所述数据预处理方式包括量化处理,所述装置还包括:
第五信息接收模块,用于接收所述网络侧设备发送的第五信息,所述第五信息中携带数据量化辅助信息,所述数据量化辅助信息包括以下至少一项:
量化参考点的标识和位置;
量化方式,所述量化方式包括均匀量化或非均匀量化;
所述第二预处理子模块,包括:
量化处理单元,用于根据所述数据量化辅助信息对所述第一位置数据进行量化处理,得到第二位置数据。
可选地,所述数据预处理方式包括归一化处理,所述装置还包括:
第六信息接收模块,用于接收所述网络侧设备发送的第六信息,所述第六信息中携带归一化指示;
所述第二预处理子模块,包括:
归一化处理单元,用于按照所述归一化指示对应的目标归一化方法对所述第一位置数据进行归一化处理,得到第二位置数据。
可选地,所述终端设备包括常规终端设备和/或定位参考单元。
第四方面,本申请实施例提供了另一种数据标注装置。参照图7,示出了本申请实施例提供的一种数据标注装置的结构框图,该装置可以应用于网络侧设备。如图7所示,该装置具体可以包括:
第一信息发送模块701,用于向终端设备发送第一信息,所述第一信息用于对位置数据进行标注得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
可选地,所述第一信息包括数据筛选条件和标签内容;其中,所述数据筛选条件用于筛选所述位置数据,以滤除不符合所述数据筛选条件的位置数据;所述标签内容用于按照所述标签内容对位置数据进行标注。
可选地,所述标签内容包括以下至少一项:
位置标签,所述位置标签用于反映所述终端设备的位置坐标,所述位置坐标包括绝对位置坐标或相对位置坐标;
区域标签,所述区域标签用于反映所述位置数据对应的区域;
时间戳标签,所述时间戳标签用于反映所述位置数据的获取时间;
位置标注误差标签,所述位置标识误差标签用于反映所述位置标签与所述终端设备的真实位置之间的误差;
终端标签,所述终端标签包括所述终端设备的设备类型标签或身份标签;
数据获取标签,所述数据获取标签用于指示所述位置数据的获取方式;
运动状态标签,所述运动状态标签用于指示所述终端设备在获取所述位置数据时的运动速度及方向;
数据质量等级标签,所述数据质量等级标签用于指示所述位置数据的数据质量。
可选地,所述装置还包括:
标注响应信息接收模块,用于接收所述终端设备发送的标注响应信息,所述标注响应信息用于反映所述终端设备是否支持对位置数据进行标注,或者,所述标注响应信息用于反映所述终端设备支持的标签内容。
可选地,所述数据筛选条件包括以下至少一项:
位置数据的位置标注误差小于预设阈值;
位置数据对应的区域属于预设区域;
位置数据对应的区域标签为预设区域标签;
位置数据的获取方式为预设获取方式;
位置数据的获取时间属于预设时间范围;
位置数据的数据质量等级大于预设质量等级;
位置数据不存在字段缺失;
位置数据不属于异常数据;
所述终端设备的终端标签为预设终端标签;
所述终端设备的终端能力等级大于预设能力等级;
所述终端设备的运动速度小于预设速度。
可选地,所述装置还包括:
数据标签接收模块,用于接收所述终端设备发送的数据标签;
数据预处理模块,用于按照预先设定的数据预处理方式对所述数据标签进行预处理,得到目标数据标签。
可选地,所述第一信息还包括数据预处理方式,所述数据预处理方式用于对位置数据进行预处理。
可选地,所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
可选地,所述数据预处理方式包括插值处理,所述装置还包括:
第二信息发送模块,用于向所述终端设备发送第二信息,所述第二信息中携带标签密度指示,所述标签密度指示用于指示所述数据标签的标签密度。
可选地,所述标签密度指示包括以下任一项:
第一时间范围内,数据标签的标签数目为N1;
第一距离范围内,数据标签的标签数目为N2;
数据标签的标签密度为N3/时间单位,或者,N4/距离单位;
其中,N1、N2、N3和N4均为正整数。
可选地,所述时间单位包括以下任一项:小时、分钟、秒、毫秒、微秒、帧、子帧、时隙、正交频分复用符号。
可选地,所述数据预处理方式包括插值处理,所述装置还包括:
第三信息发送模块,用于向所述终端设备发送第三信息,所述第三信息中携带插值指示,所述插值指示用于位置数据的插值处理。
可选地,所述数据预处理方式包括平滑处理,所述装置还包括:
第四信息发送模块,用于向所述终端设备发送第四信息,所述第四信息中携带平滑指示,所述平滑指示用于位置数据的平滑处理。
可选地,所述数据预处理方式包括量化处理,所述装置还包括:
第五信息发送模块,用于向所述终端设备发送第五信息,所述第五信息中携带数据量化辅助信息,所述数据量化辅助信息包括以下至少一项:
量化参考点的标识和位置;
量化方式,所述量化方式包括均匀量化或非均匀量化。
可选地,所述数据预处理方式包括归一化处理,所述装置还包括:
第六信息发送模块,用于向所述终端设备发送第六信息,所述第六信息中携带归一化指示,所述归一化指示用于位置数据的归一化处理。
可选地,所述终端设备包括常规终端设备和/或定位参考单元。
本申请实施例中的数据标注装置可以是电子设备,例如具有操作系统的电子设备,也可以是电子设备中的部件,例如集成电路或芯片。该电子设备可以是终端设备。示例性的,终端设备可以包括但不限于上述所列举的终端设备11的类型。
本申请实施例提供的数据标注装置能够实现图2或图5的方法实施例实现的各个过程,并达到相同的技术效果,为避免重复,这里不再赘述。
可选地,如图8所示,本申请实施例还提供一种通信设备800,包括处理器801和存储器802,存储器802上存储有可在所述处理器801上运行的程序或指令,例如,该通信设备800为网络侧设备时,该程序或指令被处理器801执行时实现上述第一方面所述的数据标注方法实施例的各个步骤,且能达到相同的技术效果。该通信设备800为终端设备时,该程序或指令被处理器801执行时实现上述第二方面所述的数据标注方法实施例的各个步骤,且能达到相同的技术效果,为避免重复,这里不再赘述。
如图9所示,为实现本申请实施例的一种终端设备的硬件结构示意图。
该终端设备900包括但不限于:射频单元901、网络模块902、音频输出单元903、输入单元904、传感器905、显示单元906、用户输入单元907、接口单元908、存储器909以及处理器910等中的至少部分部件。
本领域技术人员可以理解,终端设备900还可以包括给各个部件供电的电源(比如电池),电源可以通过电源管理系统与处理器910逻辑相连,从而通过电源管理系
统实现管理充电、放电、以及功耗管理等功能。图9中示出的终端设备结构并不构成对终端设备的限定,终端设备可以包括比图示更多或更少的部件,或者组合某些部件,或者不同的部件布置,在此不再赘述。
应理解的是,本申请实施例中,输入单元904可以包括图形处理单元(Graphics Processing Unit,GPU)9041和麦克风9042,图形处理器9041对在视频捕获模式或图像捕获模式中由图像捕获装置(如摄像头)获得的静态图片或视频的图像数据进行处理。显示单元906可包括显示面板9061,可以采用液晶显示器、有机发光二极管等形式来配置显示面板9061。用户输入单元907包括触控面板9071以及其他输入设备9072中的至少一种。触控面板9071,也称为触摸屏。触控面板9071可包括触摸检测装置和触摸控制器两个部分。其他输入设备9072可以包括但不限于物理键盘、功能键(比如音量控制按键、开关按键等)、轨迹球、鼠标、操作杆,在此不再赘述。
本申请实施例中,射频单元901接收来自网络侧设备的下行数据后,可以传输给处理器910进行处理;另外,射频单元901可以向网络侧设备发送上行数据。通常,射频单元901包括但不限于天线、放大器、收发信机、耦合器、低噪声放大器、双工器等。
存储器909可用于存储软件程序或指令以及各种数据。存储器909可主要包括存储程序或指令的第一存储区和存储数据的第二存储区,其中,第一存储区可存储操作系统、至少一个功能所需的应用程序或指令(比如声音播放功能、图像播放功能等)等。此外,存储器909可以包括易失性存储器或非易失性存储器,或者,存储器909可以包括易失性和非易失性存储器两者。其中,非易失性存储器可以是只读存储器(Read-Only Memory,ROM)、可编程只读存储器(Programmable ROM,PROM)、可擦除可编程只读存储器(Erasable PROM,EPROM)、电可擦除可编程只读存储器(Electrically EPROM,EEPROM)或闪存。易失性存储器可以是随机存取存储器(Random Access Memory,RAM),静态随机存取存储器(Static RAM,SRAM)、动态随机存取存储器(Dynamic RAM,DRAM)、同步动态随机存取存储器(Synchronous DRAM,SDRAM)、双倍数据速率同步动态随机存取存储器(Double Data RateSDRAM,DDRSDRAM)、增强型同步动态随机存取存储器(Enhanced SDRAM,ESDRAM)、同步连接动态随机存取存储器(Synch link DRAM,SLDRAM)和直接内存总线随机存取存储器(Direct Rambus RAM,DRRAM)。本申请实施例中的存储器909包括但不限于这些和任意其它适合类型的存储器。
处理器910可包括一个或多个处理单元;可选的,处理器910集成应用处理器和调制解调处理器,其中,应用处理器主要处理涉及操作系统、用户界面和应用程序等的操作,调制解调处理器主要处理无线通信信号,如基带处理器。可以理解的是,上述调制解调处理器也可以不集成到处理器910中。
其中,射频单元901用于接收网络侧设备发送的第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注;
处理器910用于根据所述第一信息对位置数据进行标注,得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
可选地,所述第一信息包括数据筛选条件和标签内容;所述处理器910具体用于:
根据所述数据筛选条件对位置数据进行筛选,得到第一位置数据;
按照所述标签内容对所述第一位置数据进行标注,得到数据标签。
可选地,所述标签内容包括以下至少一项:
位置标签,所述位置标签用于反映所述终端设备的位置坐标,所述位置坐标包括绝对位置坐标或相对位置坐标;
区域标签,所述区域标签用于反映所述位置数据对应的区域;
时间戳标签,所述时间戳标签用于反映所述位置数据的获取时间;
位置标注误差标签,所述位置标识误差标签用于反映所述位置标签与所述终端设备的真实位置之间的误差;
终端标签,所述终端标签包括所述终端设备的设备类型标签或身份标签;
数据获取标签,所述数据获取标签用于指示所述位置数据的获取方式;
运动状态标签,所述运动状态标签用于指示所述终端设备在获取所述位置数据时
的运动速度及方向;
数据质量等级标签,所述数据质量等级标签用于指示所述位置数据的数据质量。
可选地,射频单元901还用于:向所述网络侧设备发送标注响应信息,所述标注响应信息用于反映所述终端设备是否支持对位置数据进行标注,或者,所述标注响应信息用于反映所述终端设备支持的标签内容。
可选地,所述数据筛选条件包括以下至少一项:
位置数据的位置标注误差小于预设阈值;
位置数据对应的区域属于预设区域;
位置数据对应的区域标签为预设区域标签;
位置数据的获取方式为预设获取方式;
位置数据的获取时间属于预设时间范围;
位置数据的数据质量等级大于预设质量等级;
位置数据不存在字段缺失;
位置数据不属于异常数据;
所述终端设备的终端标签为预设终端标签;
所述终端设备的终端能力等级大于预设能力等级;
所述终端设备的运动速度小于预设速度。
可选地,所述处理器910具体用于:
根据预先设定或协议规定的数据预处理方式对位置数据进行预处理,得到预处理后的位置数据;
根据所述第一信息对预处理后的位置数据进行标注,得到数据标签。
可选地,所述第一信息还包括数据预处理方式,按照所述标签内容对所述第一位置数据进行标注,得到数据标签之前,所述处理器910还用于:按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据;
所述处理器910具体用于:按照所述标签内容对所述第二位置数据进行标注,得到数据标签。
可选地,所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
可选地,所述数据预处理方式包括插值处理,射频单元901还用于接收所述网络侧设备发送的第二信息,所述第二信息中携带标签密度指示,所述标签密度指示用于指示所述数据标签的标签密度;
处理器910具体用于根据所述标签密度指示对所述第一位置数据进行插值处理,得到第二位置数据。
可选地,所述标签密度指示包括以下任一项:
第一时间范围内,数据标签的标签数目为N1;
第一距离范围内,数据标签的标签数目为N2;
数据标签的标签密度为N3/时间单位,或者,N4/距离单位;
其中,N1、N2、N3和N4均为正整数。
可选地,所述时间单位包括以下任一项:小时、分钟、秒、毫秒、微秒、帧、子帧、时隙、正交频分复用符号。
可选地,所述数据预处理方式包括插值处理,射频单元901还用于接收所述网络侧设备发送的第三信息,所述第三信息中携带插值指示;
处理器910用于按照所述插值指示对应的目标插值方法对所述第一位置数据进行插值处理,得到第二位置数据。
可选地,所述数据预处理方式包括平滑处理,射频单元901还用于接收所述网络侧设备发送的第四信息,所述第四信息中携带平滑指示;
处理器910用于按照所述平滑指示对应的目标平滑方法对所述第一位置数据进行平滑处理,得到第二位置数据。
可选地,所述数据预处理方式包括量化处理,射频单元901还用于接收所述网络侧设备发送的第五信息,所述第五信息中携带数据量化辅助信息,所述数据量化辅助信息包括以下至少一项:
量化参考点的标识和位置;
量化方式,所述量化方式包括均匀量化或非均匀量化;
处理器910用于根据所述数据量化辅助信息对所述第一位置数据进行量化处理,
得到第二位置数据。
可选地,所述数据预处理方式包括归一化处理,射频单元901还用于接收所述网络侧设备发送的第六信息,所述第六信息中携带归一化指示;
处理器910用于按照所述归一化指示对应的目标归一化方法对所述第一位置数据进行归一化处理,得到第二位置数据。
可选地,所述终端设备包括常规终端设备和/或定位参考单元。
本申请实施例还提供一种网络侧设备,如图10所示,该网络侧设备1000包括:天线101、射频装置102、基带装置103、处理器104和存储器105。天线101与射频装置102连接。在上行方向上,射频装置102通过天线101接收信息,将接收的信息发送给基带装置103进行处理。在下行方向上,基带装置103对要发送的信息进行处理,并发送给射频装置102,射频装置102对收到的信息进行处理后经过天线101发送出去。
以上实施例中网络侧设备执行的方法可以在基带装置103中实现,该基带装置103包括基带处理器。
基带装置103例如可以包括至少一个基带板,该基带板上设置有多个芯片,如图10所示,其中一个芯片例如为基带处理器,通过总线接口与存储器105连接,以调用存储器105中的程序,执行以上方法实施例中所示的网络设备操作。
该网络侧设备还可以包括网络接口106,该接口例如为通用公共无线接口(common public radio interface,CPRI)。
具体地,本发明实施例的网络侧设备1000还包括:存储在存储器105上并可在处理器104上运行的指令或程序,处理器104调用存储器105中的指令或程序执行前述第二方面方法,并达到相同的技术效果,为避免重复,故不在此赘述。
本申请实施例还提供了一种网络侧设备。如图11所示,该网络侧设备1100包括:处理器1101、网络接口1102和存储器1103。其中,网络接口1102例如为通用公共无线接口(common public radio interface,CPRI)。
具体地,本发明实施例的网络侧设备1100还包括:存储在存储器1103上并可在处理器1101上运行的指令或程序,处理器1101调用存储器1103中的指令或程序执行前述第二方面的方法,并达到相同的技术效果,为避免重复,故不在此赘述。
本申请实施例还提供一种可读存储介质,所述可读存储介质上存储有程序或指令,该程序或指令被处理器执行时实现上述数据标注方法实施例的各个过程,且能达到相同的技术效果,为避免重复,这里不再赘述。
其中,所述处理器为上述实施例中所述的终端设备中的处理器。所述可读存储介质,包括计算机可读存储介质,如计算机只读存储器ROM、随机存取存储器RAM、磁碟或者光盘等。
本申请实施例另提供了一种芯片,所述芯片包括处理器和通信接口,所述通信接口和所述处理器耦合,所述处理器用于运行程序或指令,实现上述数据标注方法实施例的各个过程,且能达到相同的技术效果,为避免重复,这里不再赘述。
应理解,本申请实施例提到的芯片还可以称为系统级芯片,系统芯片,芯片系统或片上系统芯片等。
本申请实施例另提供了一种计算机程序/程序产品,所述计算机程序/程序产品被存储在存储介质中,所述计算机程序/程序产品被至少一个处理器执行以实现上述数据标注方法实施例的各个过程,且能达到相同的技术效果,为避免重复,这里不再赘述。
本申请实施例还提供了一种数据标注系统,包括:终端设备及网络侧设备,所述终端可用于执行如上第二方面所述的数据标注方法的步骤,所述网络侧设备可用于执行如上第一方面所述的数据标注方法的步骤。
需要说明的是,在本文中,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者装置不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者装置所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括该要素的过程、方法、物品或者装置中还存在另外的相同要素。此外,需要指出的是,本申请实施方式中的方法和装置的范围不限按示出或讨论的顺序来执行功能,还可包括根据所涉及的功能按基本同时的方式或按相反的顺序来执行功能,例如,可以按不同于所描述的次序来执行所描述的方法,并且还可以
添加、省去、或组合各种步骤。另外,参照某些示例所描述的特征可在其他示例中被组合。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以计算机软件产品的形式体现出来,该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中,包括若干指令用以使得一台终端(可以是手机,计算机,服务器,空调器,或者网络设备等)执行本申请各个实施例所述的方法。
上面结合附图对本申请的实施例进行了描述,但是本申请并不局限于上述的具体实施方式,上述的具体实施方式仅仅是示意性的,而不是限制性的,本领域的普通技术人员在本申请的启示下,在不脱离本申请宗旨和权利要求所保护的范围情况下,还可做出很多形式,均属于本申请的保护之内。
Claims (37)
- 一种数据标注方法,所述方法包括:终端设备接收网络侧设备发送的第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注;所述终端设备根据所述第一信息对位置数据进行标注,得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
- 根据权利要求1所述的方法,其中,所述第一信息包括数据筛选条件和标签内容;所述终端设备根据所述第一信息对位置数据进行标注,得到数据标签,包括:所述终端设备根据所述数据筛选条件对位置数据进行筛选,得到第一位置数据;所述终端设备按照所述标签内容对所述第一位置数据进行标注,得到数据标签。
- 根据权利要求2所述的方法,其中,所述标签内容包括以下至少一项:位置标签,所述位置标签用于反映所述终端设备的位置坐标,所述位置坐标包括绝对位置坐标或相对位置坐标;区域标签,所述区域标签用于反映所述位置数据对应的区域;时间戳标签,所述时间戳标签用于反映所述位置数据的获取时间;位置标注误差标签,所述位置标识误差标签用于反映所述位置标签与所述终端设备的真实位置之间的误差;终端标签,所述终端标签包括所述终端设备的设备类型标签或身份标签;数据获取标签,所述数据获取标签用于指示所述位置数据的获取方式;运动状态标签,所述运动状态标签用于指示所述终端设备在获取所述位置数据时的运动速度及方向;数据质量等级标签,所述数据质量等级标签用于指示所述位置数据的数据质量。
- 根据权利要求3所述的方法,其中,所述方法还包括:所述终端设备向所述网络侧设备发送标注响应信息,所述标注响应信息用于反映所述终端设备是否支持对位置数据进行标注,或者,所述标注响应信息用于反映所述终端设备支持的标签内容。
- 根据权利要求2所述的方法,其中,所述数据筛选条件包括以下至少一项:位置数据的位置标注误差小于预设阈值;位置数据对应的区域属于预设区域;位置数据对应的区域标签为预设区域标签;位置数据的获取方式为预设获取方式;位置数据的获取时间属于预设时间范围;位置数据的数据质量等级大于预设质量等级;位置数据不存在字段缺失;位置数据不属于异常数据;所述终端设备的终端标签为预设终端标签;所述终端设备的终端能力等级大于预设能力等级;所述终端设备的运动速度小于预设速度。
- 根据权利要求1所述的方法,其中,所述终端设备根据所述第一信息对位置数据进行标注,得到数据标签,包括:所述终端设备根据预先设定或协议规定的数据预处理方式对位置数据进行预处理,得到预处理后的位置数据;所述终端设备根据所述第一信息对预处理后的位置数据进行标注,得到数据标签。
- 根据权利要求2所述的方法,其中,所述第一信息还包括数据预处理方式,所述终端设备按照所述标签内容对所述第一位置数据进行标注,得到数据标签之前,所述方法还包括:所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据;所述终端设备按照所述标签内容对所述第一位置数据进行标注,得到数据标签,包括:所述终端设备按照所述标签内容对所述第二位置数据进行标注,得到数据标签。
- 根据权利要求6或7所述的方法,其中,所述数据预处理方式包括以下至少 一项:插值处理、平滑处理、量化处理、归一化处理。
- 根据权利要求7所述的方法,其中,所述数据预处理方式包括插值处理,所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:所述终端设备接收所述网络侧设备发送的第二信息,所述第二信息中携带标签密度指示,所述标签密度指示用于指示所述数据标签的标签密度;所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:所述终端设备根据所述标签密度指示对所述第一位置数据进行插值处理,得到第二位置数据。
- 根据权利要求9所述的方法,其中,所述标签密度指示包括以下任一项:第一时间范围内,数据标签的标签数目为N1;第一距离范围内,数据标签的标签数目为N2;数据标签的标签密度为N3/时间单位,或者,N4/距离单位;其中,N1、N2、N3和N4均为正整数。
- 根据权利要求10所述的方法,其中,所述时间单位包括以下任一项:小时、分钟、秒、毫秒、微秒、帧、子帧、时隙、正交频分复用符号。
- 根据权利要求7所述的方法,其中,所述数据预处理方式包括插值处理,所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:所述终端设备接收所述网络侧设备发送的第三信息,所述第三信息中携带插值指示;所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:所述终端设备按照所述插值指示对应的目标插值方法对所述第一位置数据进行插值处理,得到第二位置数据。
- 根据权利要求7所述的方法,其中,所述数据预处理方式包括平滑处理,所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:所述终端设备接收所述网络侧设备发送的第四信息,所述第四信息中携带平滑指示;所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:所述终端设备按照所述平滑指示对应的目标平滑方法对所述第一位置数据进行平滑处理,得到第二位置数据。
- 根据权利要求7所述的方法,其中,所述数据预处理方式包括量化处理,所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:所述终端设备接收所述网络侧设备发送的第五信息,所述第五信息中携带数据量化辅助信息,所述数据量化辅助信息包括以下至少一项:量化参考点的标识和位置;量化方式,所述量化方式包括均匀量化或非均匀量化;所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:所述终端设备根据所述数据量化辅助信息对所述第一位置数据进行量化处理,得到第二位置数据。
- 根据权利要求7所述的方法,其中,所述数据预处理方式包括归一化处理,所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据之前,所述方法还包括:所述终端设备接收所述网络侧设备发送的第六信息,所述第六信息中携带归一化指示;所述终端设备按照所述数据预处理方式对所述第一位置数据进行预处理,得到第二位置数据,包括:所述终端设备按照所述归一化指示对应的目标归一化方法对所述第一位置数据进 行归一化处理,得到第二位置数据。
- 根据权利要求1至15任一项所述的方法,其中,所述终端设备包括常规终端设备和/或定位参考单元。
- 一种数据标注方法,所述方法包括:网络侧设备向终端设备发送第一信息,所述第一信息用于对位置数据进行标注得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
- 根据权利要求17所述的方法,其中,所述第一信息包括数据筛选条件和标签内容;其中,所述数据筛选条件用于筛选所述位置数据,以滤除不符合所述数据筛选条件的位置数据;所述标签内容用于按照所述标签内容对位置数据进行标注。
- 根据权利要求18所述的方法,其中,所述标签内容包括以下至少一项:位置标签,所述位置标签用于反映所述终端设备的位置坐标,所述位置坐标包括绝对位置坐标或相对位置坐标;区域标签,所述区域标签用于反映所述位置数据对应的区域;时间戳标签,所述时间戳标签用于反映所述位置数据的获取时间;位置标注误差标签,所述位置标识误差标签用于反映所述位置标签与所述终端设备的真实位置之间的误差;终端标签,所述终端标签包括所述终端设备的设备类型标签或身份标签;数据获取标签,所述数据获取标签用于指示所述位置数据的获取方式;运动状态标签,所述运动状态标签用于指示所述终端设备在获取所述位置数据时的运动速度及方向;数据质量等级标签,所述数据质量等级标签用于指示所述位置数据的数据质量。
- 根据权利要求19所述的方法,其中,所述方法还包括:所述网络侧设备接收所述终端设备发送的标注响应信息,所述标注响应信息用于反映所述终端设备是否支持对位置数据进行标注,或者,所述标注响应信息用于反映所述终端设备支持的标签内容。
- 根据权利要求18所述的方法,其中,所述数据筛选条件包括以下至少一项:位置数据的位置标注误差小于预设阈值;位置数据对应的区域属于预设区域;位置数据对应的区域标签为预设区域标签;位置数据的获取方式为预设获取方式;位置数据的获取时间属于预设时间范围;位置数据的数据质量等级大于预设质量等级;位置数据不存在字段缺失;位置数据不属于异常数据;所述终端设备的终端标签为预设终端标签;所述终端设备的终端能力等级大于预设能力等级;所述终端设备的运动速度小于预设速度。
- 根据权利要求17所述的方法,其中,所述方法还包括:所述网络侧设备接收所述终端设备发送的数据标签;所述网络侧设备按照预先设定的数据预处理方式对所述数据标签进行预处理,得到目标数据标签。
- 根据权利要求18所述的方法,其中,所述第一信息还包括数据预处理方式,所述数据预处理方式用于对位置数据进行预处理。
- 根据权利要求22或23所述的方法,其中,所述数据预处理方式包括以下至少一项:插值处理、平滑处理、量化处理、归一化处理。
- 根据权利要求23所述的方法,其中,所述数据预处理方式包括插值处理,所述方法还包括:所述网络侧设备向所述终端设备发送第二信息,所述第二信息中携带标签密度指示,所述标签密度指示用于指示所述数据标签的标签密度。
- 根据权利要求25所述的方法,其中,所述标签密度指示包括以下任一项:第一时间范围内,数据标签的标签数目为N1;第一距离范围内,数据标签的标签数目为N2;数据标签的标签密度为N3/时间单位,或者,N4/距离单位;其中,N1、N2、N3和N4均为正整数。
- 根据权利要求26所述的方法,其中,所述时间单位包括以下任一项:小时、分钟、秒、毫秒、微秒、帧、子帧、时隙、正交频分复用符号。
- 根据权利要求23所述的方法,其中,所述数据预处理方式包括插值处理,所述方法还包括:所述网络侧设备向所述终端设备发送第三信息,所述第三信息中携带插值指示,所述插值指示用于位置数据的插值处理。
- 根据权利要求23所述的方法,其中,所述数据预处理方式包括平滑处理,所述方法还包括:所述网络侧设备向所述终端设备发送第四信息,所述第四信息中携带平滑指示,所述平滑指示用于位置数据的平滑处理。
- 根据权利要求23所述的方法,其中,所述数据预处理方式包括量化处理,所述方法还包括:所述网络侧设备向所述终端设备发送第五信息,所述第五信息中携带数据量化辅助信息,所述数据量化辅助信息包括以下至少一项:量化参考点的标识和位置;量化方式,所述量化方式包括均匀量化或非均匀量化。
- 根据权利要求23所述的方法,其中,所述数据预处理方式包括归一化处理,所述方法还包括:所述网络侧设备向所述终端设备发送第六信息,所述第六信息中携带归一化指示,所述归一化指示用于位置数据的归一化处理。
- 根据权利要求17至31任一项所述的方法,其中,所述终端设备包括常规终端设备和/或定位参考单元。
- 一种数据标注装置,应用于终端设备,所述装置包括:第一信息接收模块,用于接收网络侧设备发送的第一信息,所述第一信息用于指示所述终端设备对位置数据进行标注;数据标注模块,用于根据所述第一信息对位置数据进行标注,得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
- 一种数据标注装置,应用于网络侧设备,所述装置包括:第一信息发送模块,用于向终端设备发送第一信息,所述第一信息用于对位置数据进行标注得到数据标签;所述数据标签用于对机器学习模型进行迭代训练,所述机器学习模型用于预测所述终端设备的位置信息。
- 一种终端设备,包括处理器和存储器,所述存储器存储可在所述处理器上运行的程序或指令,所述程序或指令被所述处理器执行时实现如权利要求1至16任一项所述的数据标注方法的步骤。
- 一种网络侧设备,包括处理器和存储器,所述存储器存储可在所述处理器上运行的程序或指令,所述程序或指令被所述处理器执行时实现如权利要求17至32任一项所述的数据标注方法的步骤。
- 一种可读存储介质,所述可读存储介质上存储程序或指令,所述程序或指令被处理器执行时实现如权利要求1至16任一项所述的数据标注方法,或者实现如权利要求17至32任一项所述的数据标注方法的步骤。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211350294.1A CN117998295A (zh) | 2022-10-31 | 2022-10-31 | 数据标注方法、装置、终端设备及网络侧设备 |
| CN202211350294.1 | 2022-10-31 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2024093799A1 true WO2024093799A1 (zh) | 2024-05-10 |
Family
ID=90888307
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2023/126883 Ceased WO2024093799A1 (zh) | 2022-10-31 | 2023-10-26 | 数据标注方法、装置、终端设备及网络侧设备 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN117998295A (zh) |
| WO (1) | WO2024093799A1 (zh) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20200033694A (ko) * | 2018-09-20 | 2020-03-30 | 에스케이텔레콤 주식회사 | 측위모델구성장치 및 단말측위장치, 그 장치의 동작 방법 |
| CN113449207A (zh) * | 2020-03-24 | 2021-09-28 | 华为技术有限公司 | 电子设备定位的方法、装置、服务器、系统以及存储介质 |
| US20220046385A1 (en) * | 2020-08-04 | 2022-02-10 | Qualcomm Incorporated | Selective triggering of neural network functions for positioning measurement feature processing at a user equipment |
| WO2022033094A1 (zh) * | 2020-08-14 | 2022-02-17 | 格力电器(武汉)有限公司 | 定位基站数据流处理方法、装置、设备及计算机可读介质 |
| CN114090601A (zh) * | 2021-11-23 | 2022-02-25 | 北京百度网讯科技有限公司 | 一种数据筛选方法、装置、设备以及存储介质 |
| WO2022052199A1 (zh) * | 2020-09-11 | 2022-03-17 | 南方科技大学 | 数据标注方法、网络设备、终端、系统及存储介质 |
| CN114222238A (zh) * | 2020-09-03 | 2022-03-22 | 中国电信股份有限公司 | 定位方法、装置和计算机可读存储介质 |
-
2022
- 2022-10-31 CN CN202211350294.1A patent/CN117998295A/zh active Pending
-
2023
- 2023-10-26 WO PCT/CN2023/126883 patent/WO2024093799A1/zh not_active Ceased
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20200033694A (ko) * | 2018-09-20 | 2020-03-30 | 에스케이텔레콤 주식회사 | 측위모델구성장치 및 단말측위장치, 그 장치의 동작 방법 |
| CN113449207A (zh) * | 2020-03-24 | 2021-09-28 | 华为技术有限公司 | 电子设备定位的方法、装置、服务器、系统以及存储介质 |
| US20220046385A1 (en) * | 2020-08-04 | 2022-02-10 | Qualcomm Incorporated | Selective triggering of neural network functions for positioning measurement feature processing at a user equipment |
| WO2022033094A1 (zh) * | 2020-08-14 | 2022-02-17 | 格力电器(武汉)有限公司 | 定位基站数据流处理方法、装置、设备及计算机可读介质 |
| CN114222238A (zh) * | 2020-09-03 | 2022-03-22 | 中国电信股份有限公司 | 定位方法、装置和计算机可读存储介质 |
| WO2022052199A1 (zh) * | 2020-09-11 | 2022-03-17 | 南方科技大学 | 数据标注方法、网络设备、终端、系统及存储介质 |
| CN114090601A (zh) * | 2021-11-23 | 2022-02-25 | 北京百度网讯科技有限公司 | 一种数据筛选方法、装置、设备以及存储介质 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117998295A (zh) | 2024-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116233906B (zh) | 测量方法、装置、设备及存储介质 | |
| CN116939713A (zh) | 信息反馈方法、装置及设备 | |
| WO2024017176A1 (zh) | 模型训练方法、装置、网络侧设备及终端设备 | |
| CN117639862A (zh) | Csi预测处理方法、装置、通信设备及可读存储介质 | |
| US20250227507A1 (en) | Ai model processing method and apparatus, and communication device | |
| WO2024120445A1 (zh) | 模型输入信息的确定方法、装置、设备、系统及存储介质 | |
| CN117411602A (zh) | 数据采集方法及装置 | |
| WO2024083004A1 (zh) | Ai模型配置方法、终端及网络侧设备 | |
| WO2024125525A1 (zh) | Ai算力上报方法、终端及网络侧设备 | |
| WO2024067438A1 (zh) | Ai模型推理方法、设备及可读存储介质 | |
| WO2024067280A1 (zh) | 更新ai模型参数的方法、装置及通信设备 | |
| WO2024093799A1 (zh) | 数据标注方法、装置、终端设备及网络侧设备 | |
| WO2024140442A1 (zh) | 模型更新方法、装置及设备 | |
| CN117997397A (zh) | 资源配置方法、装置、通信设备及可读存储介质 | |
| WO2025108195A1 (zh) | 模型确定方法、装置及通信设备 | |
| WO2024208167A1 (zh) | 信息处理方法、信息处理装置、终端及网络侧设备 | |
| WO2025036223A1 (zh) | 信息上报方法、信息接收方法及设备 | |
| WO2025092999A1 (zh) | 模型性能监督方法、装置及设备 | |
| CN118042399A (zh) | 模型监督方法、终端及网络侧设备 | |
| WO2025031325A1 (zh) | 基于参考模型的通信方法及设备 | |
| WO2025108390A1 (zh) | 基于ai模型的定位方法、装置、设备及可读存储介质 | |
| WO2024067665A1 (zh) | Csi预测处理方法、装置、通信设备及可读存储介质 | |
| WO2024153013A1 (zh) | 信息传输方法、装置及通信设备 | |
| CN120602970A (zh) | 测量处理方法、测量配置方法、装置、设备及可读存储介质 | |
| WO2024032695A1 (zh) | Csi预测处理方法、装置、通信设备及可读存储介质 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 23884731 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 23884731 Country of ref document: EP Kind code of ref document: A1 |
|
| 32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 08.09.2025) |