Processing method and apparatus for ai model switching, and communication device

Abstract

This application discloses a method, apparatus, and communication device for AI model switching, belonging to the field of communication technology. The AI ​​model switching method of this application includes: a first communication device receiving switching instruction information sent by a second communication device, the switching instruction information being used to instruct the first communication device to switch AI models; when there is a switching interval gap between the switching of the first AI model and the second AI model, the first communication device performs a first operation during the switching gap; the first operation includes any one of the following: not performing calculations on the first AI model and the second AI model; stopping the use of the AI ​​computing unit; performing calculations on the first AI model, and the calculation result is usable or valid; performing calculations on the first AI model and / or the second AI model, and the calculation result is unusable or invalid; stopping the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit.

Description

Technical Field

[0001] This application belongs to the field of communication technology, specifically relating to a processing method, apparatus, and communication equipment for switching AI models. Background Technology

[0002] Artificial intelligence (AI) has been widely applied in various fields. AI models can be implemented in various ways, such as neural networks, decision trees, support vector machines, and Bayesian classifiers.

[0003] Take neural networks as an example (but this doesn't limit the specific type of AI model). A neural network consists of neurons. The parameters of the neural network are optimized using optimization algorithms. An optimization algorithm is a class of algorithms that minimizes or maximizes an objective function (or loss function). The objective function is a mathematical combination of the model parameters and the data. For example, given data X and its corresponding label Y, we build a neural network model f(.). With the model, we can obtain the predicted output f(x) based on the input x, and calculate the difference between the predicted value and the true value (f(x) - Y), which is the loss function. The goal of the optimization algorithm is to find suitable parameters that minimize the value of the loss function. The smaller the loss value, the closer the model is to the reality.

[0004] Most common optimization algorithms are based on backpropagation. The basic idea of ​​backpropagation is that the learning process consists of two parts: forward propagation of the signal and backward propagation of the error. During forward propagation, input samples are introduced from the input layer, processed layer by layer by the hidden layers, and then transmitted to the output layer. If the actual output of the output layer does not match the expected output, the process transitions to backpropagation. Backpropagation transmits the output error in some form back through the hidden layers to the input layer, distributing the error to all units in each layer, thus obtaining the error signal for each unit. This error signal serves as the basis for adjusting the weights of each unit. This process of adjusting the weights through forward and backward propagation is cyclical. This continuous adjustment of weights is the learning and training process of the network. This process continues until the error in the network output is reduced to an acceptable level, or until a predetermined number of learning iterations are completed.

[0005] During AI model switching, hardware such as Neural Network Processing Unit (NPU), Graphics Processing Unit (GPU), and Field Programmable Gate Array (FPGA) may be affected. Furthermore, AI model switching takes time, and the new AI model cannot take effect immediately. Current technology does not restrict the behavior of communication devices during AI model switching, which can easily lead to communication chaos. Summary of the Invention

[0006] This application provides a method, apparatus, and communication device for processing AI model switching, which can solve the problem in the prior art that the behavior of the communication device during AI model switching is not restricted, which can easily lead to communication chaos.

[0007] Firstly, a method for switching artificial intelligence (AI) models is provided, including:

[0008] The first communication device receives a switching instruction information sent by the second communication device, the switching instruction information being used to instruct the first communication device to switch AI models;

[0009] When there is a switching gap between the first AI model and the second AI model, the first communication device performs a first operation during the switching gap; the first operation includes any one of the following:

[0010] The calculations for the first AI model and the second AI model are not performed;

[0011] Stop using the AI ​​computing unit;

[0012] Perform the calculation of the first AI model, and the calculation result is available or valid;

[0013] The calculation of the first AI model and / or the second AI model is performed, and the calculation results are unavailable or invalid;

[0014] Discontinue the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit.

[0015] Secondly, a method for switching artificial intelligence (AI) models is provided, including:

[0016] The second communication device sends switching gap configuration information to the first communication device, the switching gap configuration information including at least one of the following:

[0017] The first operation is the operation of the first communication device during the switching gap between the first AI model and the second AI model.

[0018] The second operation involves the operation of the first communication device when there is no gap switching or after the gap switching is completed.

[0019] Switch the start time of the gap;

[0020] Switch the end time of the gap;

[0021] Toggle the duration of the gap;

[0022] The cycle for switching gaps;

[0023] Is there a gap switching issue?

[0024] Switch the communication process applicable to the Gap.

[0025] Thirdly, an AI model switching processing device is provided, including:

[0026] The receiving module is used to receive switching indication information sent by the second communication device, the switching indication information being used to instruct the first communication device to switch AI models;

[0027] The first execution module is configured to perform a first operation during the switching gap when there is a switching interval between the first AI model and the second AI model; the first operation includes any one of the following:

[0028] The calculations for the first AI model and the second AI model are not performed;

[0029] Stop using the AI ​​computing unit;

[0030] Perform the calculation of the first AI model, and the calculation result is available or valid;

[0031] The calculation of the first AI model and / or the second AI model is performed, and the calculation results are unavailable or invalid;

[0032] Discontinue the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit.

[0033] Fourthly, a processing device for switching AI models is provided, including:

[0034] The sending module is configured to send switching gap configuration information to the first communication device, wherein the switching gap configuration information includes at least one of the following:

[0035] The first operation is the operation of the first communication device during a gap switching process.

[0036] The second operation involves the operation of the first communication device when there is no gap switching or after the gap switching is completed.

[0037] Switch the start time of the gap;

[0038] Switch the end time of the gap;

[0039] Toggle the duration of the gap;

[0040] The cycle for switching gaps;

[0041] Is there a gap switching issue?

[0042] Switch the communication process applicable to the Gap.

[0043] Fifthly, a communication device is provided, the terminal including a processor and a memory, the memory storing a program or instructions executable on the processor, the program or instructions, when executed by the processor, implementing the steps of the method as described in the first aspect, or implementing the steps of the method as described in the second aspect.

[0044] In a sixth aspect, a communication device is provided, including a processor and a communication interface, wherein the communication interface is used to receive switching indication information sent by a second communication device, the switching indication information being used to instruct the first communication device to switch AI models; the processor is used to perform a first operation during the switching gap when there is a switching interval gap between the first AI model and the second AI model; the first operation includes any one of the following:

[0045] The calculations for the first AI model and the second AI model are not performed;

[0046] Stop using the AI ​​computing unit;

[0047] Perform the calculation of the first AI model, and the calculation result is available or valid;

[0048] The calculation of the first AI model and / or the second AI model is performed, and the calculation results are unavailable or invalid;

[0049] Discontinue the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit.

[0050] In a seventh aspect, a communication device is provided, including a processor and a communication interface, wherein the communication interface is used to send switching gap configuration information to a first communication device, the switching gap configuration information including at least one of the following:

[0051] The first operation is the operation of the first communication device during the switching gap between the first AI model and the second AI model.

[0052] The second operation involves the operation of the first communication device when there is no gap switching or after the gap switching is completed.

[0053] Switch the start time of the gap;

[0054] Switch the end time of the gap;

[0055] Toggle the duration of the gap;

[0056] The cycle for switching gaps;

[0057] Is there a gap switching issue?

[0058] Switch the communication process applicable to the Gap.

[0059] Eighthly, a wireless communication system is provided, comprising: a first communication device and a second communication device, wherein the first communication device is configured to perform the steps of the method described in the first aspect, and the second communication device is configured to perform the steps of the method described in the second aspect.

[0060] A ninth aspect provides a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect, or implement the steps of the method described in the second aspect.

[0061] In a tenth aspect, a chip is provided, the chip including a processor and a communication interface coupled to the processor, the processor being configured to run programs or instructions to implement the method as described in the first aspect, or to implement the method as described in the second aspect.

[0062] Eleventhly, a computer program / program product is provided, the computer program / program product being stored in a storage medium, the computer program / program product being executed by at least one processor to implement the steps of the method as described in the first aspect, or to implement the steps of the method as described in the second aspect.

[0063] In this embodiment, the second communication device instructs the first communication device to switch AI models. When there is a switching gap in the AI ​​model switching, the operation of the first communication device in the switching gap is restricted to reduce the impact of AI model switching on the communication system, improve the performance of the AI ​​model in the communication system, and thus improve system performance. Attached Figure Description

[0064] Figure 1 A block diagram illustrating a wireless communication system to which embodiments of this application may be applied;

[0065] Figure 2This is a flowchart illustrating one of the steps of the AI ​​model switching processing method provided in an embodiment of this application;

[0066] Figure 3 This is the second flowchart illustrating the steps of the AI ​​model switching processing method provided in the embodiments of this application;

[0067] Figure 4 A schematic diagram illustrating the structure of the AI ​​model switching processing device provided in an embodiment of this application;

[0068] Figure 5 This is another schematic diagram of the processing device for switching AI models provided in the embodiments of this application;

[0069] Figure 6 This is a schematic diagram of the structure of the communication device provided in the embodiments of this application;

[0070] Figure 7 This is a schematic diagram of the structure of the terminal provided in the embodiments of this application;

[0071] Figure 8 This is a schematic diagram illustrating the structure of the network-side device provided in the embodiments of this application. Detailed Implementation

[0072] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0073] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, not limited in number; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0074] It is worth noting that the technologies described in this application are not limited to Long Term Evolution (LTE) / LTE-Advanced (LTE-A) systems, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency Division Multiple Access (SC-FDMA), and other systems. The terms "system" and "network" in this application are often used interchangeably, and the described technologies can be used with the systems and radio technologies mentioned above, as well as with other systems and radio technologies. The following description describes New Radio (NR) systems for illustrative purposes, and NR terminology is used in most of the following description; however, these technologies can also be applied to applications beyond NR systems, such as 6th generation (6G) radio systems. th Generation 6G communication system.

[0075] Figure 1This diagram illustrates a block diagram of a wireless communication system applicable to embodiments of this application. The wireless communication system includes a terminal 11 and a network-side device 12. Terminal 11 can be a mobile phone, tablet computer, laptop computer, personal digital assistant (PDA), handheld computer, netbook, ultra-mobile personal computer (UMPC), mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, vehicle-mounted device (VUE), pedestrian terminal (PUE), smart home (home devices with wireless communication capabilities, such as refrigerators, televisions, washing machines, or furniture), game console, personal computer (PC), ATM, or self-service machine, etc. Wearable devices include: smartwatches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart chains, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, etc. It should be noted that the specific type of terminal 11 is not limited in this embodiment. Network-side equipment 12 may include access network equipment or core network equipment. Access network equipment 12 may also be referred to as radio access network equipment, radio access network (RAN), radio access network function, or radio access network unit. Access network equipment 12 may include base stations, WLAN access points, or WiFi nodes, etc. Base stations may be referred to as Node B, evolved Node B (eNB), access point, base transceiver station (BTS), radio base station, radio transceiver, Basic Service Set (BSS), Extended Service Set (ESS), home B node, home evolved B node, Transmitting Receiving Point (TRP), or any other suitable term in the field, as long as the same technical effect is achieved. The base station is not limited to specific technical terms. It should be noted that in this application embodiment, only a base station in an NR system is used as an example for description, and the specific type of base station is not limited.

[0076] The AI ​​model switching processing method, apparatus, and communication device provided in this application will be described in detail below with reference to the accompanying drawings and through some embodiments and application scenarios.

[0077] Please see Figure 2 , Figure 2 A flowchart illustrating the steps of an AI model switching processing method provided for at least one embodiment of this application, the AI ​​model switching processing method comprising:

[0078] Step 201: The first communication device receives a switching instruction information sent by the second communication device, the switching instruction information being used to instruct the first communication device to switch AI models;

[0079] Optionally, the switching instruction information includes: information about the second AI model (indirectly instructing the first communication device to switch from the first AI model to the second AI model); and / or, the switching instruction information includes: instruction information 1 instructing the first communication device to use the second AI model; and / or, the switching instruction information includes: instruction information 2 instructing the first communication device that the first AI model is unavailable or that the first AI model should be replaced; and / or, the switching instruction information includes: instruction information 3 instructing the first communication device to replace the first AI model with the second AI model.

[0080] Step 202: When there is a switching interval gap between the first AI model and the second AI model, the first communication device performs a first operation during the switching gap; the first operation includes any one of the following:

[0081] The calculations for the first AI model and the second AI model are not performed;

[0082] The AI ​​computing unit is stopped from being used; that is, not only the first AI model and the second AI model cannot be calculated, but also the AI ​​models of other modules cannot be calculated. Optionally, when the first communication device has multiple AI computing units, the stopping of the use of the AI ​​computing unit refers to stopping the use of the AI ​​computing unit to which the first AI model belongs, and / or stopping the use of the AI ​​computing unit to which the second AI model belongs. Optionally, when the first communication device has multiple AI computing units, the inability to calculate the AI ​​models of other modules means that the AI ​​models of other modules contained in the AI ​​computing unit to which the first AI model belongs cannot be calculated, and / or the AI ​​models of other modules contained in the AI ​​computing unit to which the second AI model belongs cannot be calculated.

[0083] Perform calculations on the first AI model, and the calculation result (which is the output of the first AI model) is available or valid;

[0084] The calculation is performed using the first AI model and / or the second AI model, but the calculation results are unavailable or invalid; that is, the results output by the first AI model and / or the results output by the second AI model are both invalid.

[0085] Stop using the AI ​​computing unit and the communication module associated with the AI ​​computing unit; optionally, the communication module associated with the AI ​​computing unit can be understood as a communication module that requires intermediate information or output information of the AI ​​computing unit, or a communication module whose output information is input information or one of the input information of the AI ​​computing unit; optionally, when the first communication device has multiple AI computing units, stopping the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit means stopping the use of the AI ​​computing unit to which the first AI model belongs and the communication module associated with the AI ​​computing unit, and / or stopping the use of the AI ​​computing unit to which the second AI model belongs and the communication module associated with the AI ​​computing unit.

[0086] In the embodiments of this application, the first communication device can be a network-side device or a terminal, and the second communication device can be a network-side device or a terminal. In one typical scenario, the first communication device is a terminal and the second communication device is a network-side device; in another typical scenario, the first communication device is a network-side device and the second communication device is a terminal; in yet another typical scenario, the first communication device and the second communication device are different nodes on the network side; and in yet another typical scenario, the first communication device and the second communication device are different terminal devices.

[0087] Optionally, the first AI model is the AI ​​model originally used by the first communication device or the AI ​​model currently in use. The second AI model is the AI ​​model that is prepared to replace the "first AI model".

[0088] As an optional embodiment, the second AI model can be issued by the second communication device, or it can be pre-defined by the protocol, or it can be owned or customized by the first communication device, without any specific limitations.

[0089] It should be noted that the AI ​​computing unit is used for the operation, reasoning, running or calculation of the AI ​​model. It can be a unit, module, chip or hardware, and no specific limitation is made here.

[0090] In at least one embodiment of this application, if the first operation is to not perform calculations on the first AI model and the second AI model, or to perform calculations on the first AI model and / or the second AI model, and the calculation results are unavailable or invalid, the first operation further includes at least one of the following:

[0091] The calculation is performed using the non-AI algorithm corresponding to the first AI model;

[0092] The calculation is performed using the non-AI algorithm corresponding to the second AI model.

[0093] Here, "the non-AI algorithm corresponding to the first AI model" can be understood as a non-AI algorithm that can achieve the function of the first AI model; "the non-AI algorithm corresponding to the second AI model" can be understood as a non-AI algorithm that can achieve the function of the second AI model.

[0094] In another optional embodiment of this application, when the first operation is to not perform calculations on the first AI model and the second AI model, other resources of the AI ​​computing unit of the first communication device can be used, in addition to the first AI model and the second AI model. Optionally, when the first communication device has multiple AI computing units, the other resources of the AI ​​computing unit of the first communication device can be used, meaning that other resources of the AI ​​computing unit to which the first AI model belongs can be used, and / or other resources of the AI ​​computing unit to which the second AI model belongs can be used.

[0095] In at least one embodiment of this application, the method further includes:

[0096] When the switching between the first AI model and the second AI model does not have the switching gap (or the switching gap is 0) or the switching gap ends, the first communication device performs a second operation; the second operation includes any one of the following:

[0097] The second AI model is used for calculation; optionally, the first AI model is no longer used and can be saved to the first communication device or deleted.

[0098] Send a first indication message that the second AI model can be used to the second communication device, and use the second AI model to perform calculations after X1 time units;

[0099] Send a first indication message that the second AI model can be used to the second communication device, receive a confirmation message for the first indication message, and use the second AI model to perform calculations after receiving the confirmation message for X2 time units;

[0100] Wherein, X1 and X2 are both greater than or equal to 0.

[0101] In at least one embodiment of this application, the method further includes:

[0102] The first communication device determines switching gap configuration information, which includes at least one of the following:

[0103] The content of the first operation;

[0104] The second operation involves:

[0105] Switch the start time of the gap;

[0106] Switch the end time of the gap;

[0107] Toggle the duration of the gap;

[0108] The cycle for switching gaps;

[0109] Is there a gap switching issue?

[0110] Switch the communication process applicable to the Gap.

[0111] Optionally, the first communication device determines the switching Gap configuration information, including:

[0112] The first communication device receives switching gap configuration information indicated by the second communication device; wherein, the second communication device determines the switching gap configuration information based on the first information.

[0113] Alternatively, the first communication device determines the switching of Gap configuration information, including:

[0114] The first communication device sends a Gap request message to the second communication device based on the function requests of other communication devices, and receives the switching Gap configuration information returned by the second communication device. Optionally, the first communication device stores multiple AI models and requests a Gap from the second communication device based on the function requests of other communication devices besides the second communication device. For example, positioning requires switching between different AI models under the LMF positioning request of the positioning management function.

[0115] Alternatively, the first communication device determines the switching of Gap configuration information, including:

[0116] The first communication device sends switching gap configuration information to the second communication device; optionally, the switching gap configuration information sent by the first communication device to the second communication device is the switching gap configuration information expected by the first communication device.

[0117] Alternatively, the first communication device determines the switching of Gap configuration information, including:

[0118] The first communication device determines the switching gap configuration information to be used according to a pre-agreed agreement.

[0119] Alternatively, the first communication device determines the switching of Gap configuration information, including:

[0120] The first communication device determines multiple sets of switching gap configuration information according to a pre-agreed agreement, and receives an instruction from the first communication device to determine the switching gap configuration information to use.

[0121] For example, in actual use, the second communication device may instruct the first communication device which set of gap configuration information to use as needed. Optionally, some sets of gap configuration information correspond to short latency, while others correspond to low power consumption.

[0122] For example, the Network Data Analysis Function (NWDAF) pre-defines a pool of handover gap configuration information (including multiple sets of handover gap configuration information), and then the base station instructs which set of handover gap configuration information to use.

[0123] In at least one embodiment of this application, the method further includes:

[0124] The first communication device determines, based on the first information, whether there is a switching gap and / or the duration of the switching gap when switching between the first AI model and the second AI model.

[0125] The first information includes at least one of the following:

[0126] Information on the type differences between the first AI model and the second AI model;

[0127] Structural differences between the first AI model and the second AI model;

[0128] The parameter changes between the first AI model and the second AI model;

[0129] The total capacity or resources of the AI ​​computing unit of the first communication device;

[0130] The remaining capacity or resources of the AI ​​computing unit of the first communication device.

[0131] Optionally, the switching gap for the same type of AI model is smaller than the switching gap for different types of AI models.

[0132] Optionally, AI models of the same type or with type differences less than the first threshold do not need to switch gaps.

[0133] For example, AI model types include: fully connected neural networks, convolutional neural networks, recurrent neural networks, long short-term memory networks, decision trees, Bayesian algorithms, support vector machines, random forests, Gaussian stochastic processes, etc., which will not be listed here.

[0134] Optionally, the switching gap for AI models with the same structure is smaller than the switching gap for AI models with different structures.

[0135] Optionally, the switching gap for AI models with small structural differences is smaller than that for AI models with large structural differences.

[0136] Optionally, AI models with the same structure or with structural differences less than the second threshold do not need to switch gaps.

[0137] For example, taking neural networks as an example, the structure of an AI model includes the number of layers, the number of neurons in each layer, whether the layers are fully connected or convolutional, the type of activation function, and whether special modules (such as attention modules, memory units, etc.) are used.

[0138] Optionally, the switching gap for AI models with small parameter changes is smaller than the switching gap for AI models with large parameter changes.

[0139] Optionally, if the change in the parameters of the AI ​​model is less than a certain threshold, then there is no need to switch gaps.

[0140] Optionally, when the type and / or structure are the same, if the parameter change is less than the third threshold, there is no need to switch the gap.

[0141] Optionally, when the type difference is less than the first threshold and / or the structural difference is less than the second threshold, the parameter change is less than the third threshold, and there is no need to switch the gap.

[0142] Optionally, the greater the total capacity or resources of the AI ​​computing unit of the first communication device, the smaller the switching gap. The first, second, and third thresholds mentioned above are related to the total capacity or resources.

[0143] The phrase "remaining capacity or resources of the AI ​​computing unit of the first communication device" can be understood as the remaining capacity or resources after other functional modules (communication modules and non-communication modules) have occupied the capacity / resources of the AI ​​computing unit. Optionally, the stronger the remaining capacity / more remaining resources, the smaller the switching gap. The aforementioned first threshold, second threshold, and third threshold are related to the remaining capacity or remaining resources.

[0144] As an optional embodiment, the capabilities or resources of the AI ​​computing unit include at least one of the following:

[0145] Computational skills (multiplication, addition, etc.);

[0146] Computing power or computing ability;

[0147] Floating-point computing power; such as FLOPS (floating-point operations per second): the number of floating-point operations performed per second. Floating-point operations include all operations involving decimals and are more time-consuming than integer operations. TFLOPS (Tops) is commonly used as a metric to measure NPU / GPU performance / computing power. For example, a chip with a computing power of 1.7 Tops represents its neural network computing performance.

[0148] Number of processes;

[0149] Number of threads;

[0150] Memory;

[0151] Storage space;

[0152] cache;

[0153] Input channel information, such as the bandwidth, rate, and number of channels / bits / bytes that can be read / written / transmitted simultaneously;

[0154] Output channel information, such as the bandwidth, rate, and number of channels / bits / bytes that can be read / written / transmitted simultaneously.

[0155] In at least one embodiment of this application, the method further includes:

[0156] When the first information meets a preset condition, the first communication device sends the first information to the second communication device; for example, the preset condition is: whether at least one of the contents included in the first information exceeds a threshold, or whether at least one of the contents has changed, or whether the amount of change of at least one of the contents exceeds a threshold.

[0157] or,

[0158] When the target trigger is met, the first communication device sends the first information to the second communication device; for example, as instructed by the second communication device and / or agreed upon by the protocol, the specific content, period, triggering event, and other information of the first information. If the remaining capacity / resources of the AI ​​computing unit reported by the first communication device change, and the previously configured switching gap length is no longer suitable and needs to be adjusted, then the first communication device reports the first information.

[0159] or,

[0160] The first communication device periodically sends the first information to the second communication device; for example, the specific content, period, and other information of the first information are specified by the second communication device and / or agreed upon by the protocol.

[0161] As an optional embodiment, the communication process applicable to the switching gap includes at least one of the following:

[0162] Cell handover;

[0163] Cell redirection;

[0164] Radio Resource Control (RRC) reconfiguration;

[0165] RRC release;

[0166] RRC recovery;

[0167] RRC reconstruction;

[0168] RRC established;

[0169] Measure the gap;

[0170] Measure the gap configuration.

[0171] Accordingly, during the above communication process, the duration of the gap switching is less than or equal to any of the following:

[0172] Cell handover latency;

[0173] Cell redirection delay;

[0174] RRC reconfiguration delay;

[0175] RRC reconstruction delay;

[0176] RRC setup delay;

[0177] Measure the length of the gap.

[0178] Accordingly, during the aforementioned communication process, the first communication device determines the switching Gap configuration information, including:

[0179] The first communication device determines the switching Gap configuration information based on the first message; that is, the first message carries the switching Gap configuration information or indicates the switching Gap configuration information; wherein, the first message includes at least one of the following:

[0180] RRC reconfiguration message;

[0181] RRC release message;

[0182] RRC recovery message;

[0183] RRC reconstruction message;

[0184] RRC establishes a message.

[0185] In an optional embodiment of this application, the start time of the gap switching includes at least one of the following:

[0186] The first moment to instruct the first communication device to switch AI models;

[0187] The X3rd time unit after the first time;

[0188] The X4th time unit prior to the first time;

[0189] The second time of receiving information from the second AI model;

[0190] The X5th time unit after the second time;

[0191] The X6th time unit preceding the second time;

[0192] The third time after the second AI model receives the information and sends a confirmation message;

[0193] The X7th time unit following the third time;

[0194] The X8th time unit preceding the third time;

[0195] The fourth time when the second communication device receives the confirmation information of the second AI model information;

[0196] The X9th time unit following the fourth time;

[0197] The X10th time unit preceding the fourth time;

[0198] The fifth time when the first AI model fails;

[0199] The X11th time unit following the fifth time;

[0200] The X12th time unit preceding the fifth time;

[0201] The second AI model is expected to take effect in the sixth time.

[0202] The X13th time unit following the sixth time;

[0203] The X14th time unit preceding the sixth time;

[0204] Among them, X3 to X14 are all greater than or equal to 1.

[0205] In another optional embodiment of this application, the end time of the switching gap includes at least one of the following:

[0206] The seventh time is obtained by adding the start time of the gap switching to the duration of the gap switching;

[0207] The X15th time unit following the seventh time;

[0208] The X16th time unit prior to the seventh time;

[0209] The eighth time is obtained by adding the start time of the gap switching to the duration of the AI ​​model switching;

[0210] The X17th time unit following the eighth time;

[0211] The X18th time unit preceding the eighth time;

[0212] The ninth time the AI ​​model completes the switch;

[0213] The X19th time unit following the ninth time;

[0214] The X20th time unit prior to the ninth time;

[0215] The tenth time when the first communication device sends the first indication information that the second AI model can use;

[0216] The X21st time unit following the tenth time;

[0217] The X22nd time unit prior to the tenth time;

[0218] The eleventh time when the second communication device receives the first indication information;

[0219] The 23rd time unit following the eleventh time;

[0220] The X24th time unit preceding the eleventh time;

[0221] The twelfth time after the first communication device sends the first instruction information and the second communication device sends back the confirmation information;

[0222] The X25th time unit following the twelfth time;

[0223] The X26th time unit preceding the twelfth time;

[0224] The thirteenth time after the first communication device sends the first instruction information and receives the confirmation information from the second communication device;

[0225] The X27th time unit following the thirteenth time;

[0226] The X28th time unit preceding the thirteenth time;

[0227] Among them, X15 to X28 are all greater than or equal to 1.

[0228] In summary, in this embodiment of the application, the second communication device instructs the first communication device to switch AI models. When there is a switching gap in the AI ​​model switching, the operation of the first communication device in the switching gap is restricted, thereby reducing the impact of AI model switching on the communication system, improving the performance of the AI ​​model in the communication system, and thus improving system performance.

[0229] Please see Figure 3 , Figure 3Another flowchart of the AI ​​model switching processing method provided in this application embodiment is shown. The AI ​​model switching processing method includes:

[0230] Step 301: The second communication device sends switching gap configuration information to the first communication device, the switching gap configuration information including at least one of the following:

[0231] The first operation is the operation of the first communication device during the switching gap between the first AI model and the second AI model.

[0232] The second operation involves the operation of the first communication device when there is no gap switching or after the gap switching is completed.

[0233] Switch the start time of the gap;

[0234] Switch the end time of the gap;

[0235] Toggle the duration of the gap;

[0236] The cycle for switching gaps;

[0237] Is there a gap switching issue?

[0238] Switch the communication process applicable to the Gap.

[0239] Optionally, the first operation includes any of the following:

[0240] The first communication device does not perform calculations on the first AI model and the second AI model;

[0241] The first communication device stops using the AI ​​computing unit; that is, not only the first AI model and the second AI model cannot be calculated, but also the AI ​​models of other modules cannot be calculated. Optionally, when the first communication device has multiple AI computing units, stopping the use of the AI ​​computing unit means stopping the use of the AI ​​computing unit to which the first AI model belongs, and / or stopping the use of the AI ​​computing unit to which the second AI model belongs. Optionally, when the first communication device has multiple AI computing units, the inability to calculate the AI ​​models of other modules means that the AI ​​models of other modules contained in the AI ​​computing unit to which the first AI model belongs cannot be calculated, and / or the AI ​​models of other modules contained in the AI ​​computing unit to which the second AI model belongs cannot be calculated.

[0242] The first communication device performs calculations on the first AI model, and the calculation result (which is the output of the first AI model) is available or valid;

[0243] The first communication device performs calculations on the first AI model and / or the second AI model, and the calculation results are unusable or invalid; that is, the results output by the first AI model and / or the results output by the second AI model are both invalid.

[0244] The first communication device stops using the AI ​​computing unit and the communication module associated with the AI ​​computing unit; optionally, the communication module associated with the AI ​​computing unit can be understood as a communication module that requires intermediate information or output information of the AI ​​computing unit, or a communication module whose output information is input information or one of the input information of the AI ​​computing unit; optionally, when the first communication device has multiple AI computing units, stopping the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit means stopping the use of the AI ​​computing unit to which the first AI model belongs and the communication module associated with the AI ​​computing unit, and / or stopping the use of the AI ​​computing unit to which the second AI model belongs and the communication module associated with the AI ​​computing unit.

[0245] Optionally, the first AI model is the AI ​​model originally used by the first communication device or the AI ​​model currently in use. The second AI model is the AI ​​model that is prepared to replace the "first AI model".

[0246] As an optional embodiment, the second AI model can be issued by the second communication device, or it can be pre-defined by the protocol, or it can be owned or customized by the first communication device, without any specific limitations.

[0247] In at least one embodiment of this application, if the first operation is to not perform calculations on the first AI model and the second AI model, or to perform calculations on the first AI model and / or the second AI model, and the calculation results are unavailable or invalid, the first operation further includes at least one of the following:

[0248] The first communication device uses a non-AI algorithm corresponding to the first AI model to perform calculations;

[0249] The first communication device uses a non-AI algorithm corresponding to the second AI model to perform calculations.

[0250] Here, "the non-AI algorithm corresponding to the first AI model" can be understood as a non-AI algorithm that can achieve the function of the first AI model; "the non-AI algorithm corresponding to the second AI model" can be understood as a non-AI algorithm that can achieve the function of the second AI model.

[0251] In another optional embodiment of this application, when the first operation is to not perform calculations on the first AI model and the second AI model, other resources of the AI ​​computing unit of the first communication device can be used, in addition to the first AI model and the second AI model. Optionally, when the first communication device has multiple AI computing units, the other resources of the AI ​​computing unit of the first communication device can be used, meaning that other resources of the AI ​​computing unit to which the first AI model belongs can be used, and / or other resources of the AI ​​computing unit to which the second AI model belongs can be used.

[0252] In at least one embodiment of this application, the second operation includes any one of the following:

[0253] The first communication device uses the second AI model for calculation; optionally, if the first AI model is no longer in use, it can be saved on the first communication device side, or it can be deleted.

[0254] The first communication device sends a first indication information to the second communication device indicating that the second AI model can be used, and performs calculations using the second AI model after X1 time units;

[0255] The first communication device sends a first indication information to the second communication device that the second AI model can be used, receives confirmation information of the first indication information, and uses the second AI model to perform calculations after receiving the confirmation information for X2 time units.

[0256] Wherein, X1 and X2 are both greater than or equal to 0.

[0257] In at least one embodiment of this application, the method further includes:

[0258] The second communication device determines the switching gap configuration information used by the first communication device based on the first information; wherein the first information includes at least one of the following:

[0259] Information on the type differences between the first AI model and the second AI model;

[0260] Structural differences between the first AI model and the second AI model;

[0261] The parameter changes between the first AI model and the second AI model;

[0262] The total capacity or resources of the AI ​​computing unit of the first communication device;

[0263] The remaining capacity or resources of the AI ​​computing unit of the first communication device.

[0264] Optionally, the switching gap for the same type of AI model is smaller than the switching gap for different types of AI models.

[0265] Optionally, AI models of the same type or with type differences less than the first threshold do not need to switch gaps.

[0266] For example, AI model types include: fully connected neural networks, convolutional neural networks, recurrent neural networks, long short-term memory networks, decision trees, Bayesian algorithms, support vector machines, random forests, Gaussian stochastic processes, etc., which will not be listed here.

[0267] Optionally, the switching gap for AI models with the same structure is smaller than the switching gap for AI models with different structures.

[0268] Optionally, the switching gap for AI models with small structural differences is smaller than that for AI models with large structural differences.

[0269] Optionally, AI models with the same structure or with structural differences less than the second threshold do not need to switch gaps.

[0270] For example, taking neural networks as an example, the structure of an AI model includes the number of layers, the number of neurons in each layer, whether the layers are fully connected or convolutional, the type of activation function, and whether special modules (such as attention modules, memory units, etc.) are used.

[0271] Optionally, the switching gap for AI models with small parameter changes is smaller than the switching gap for AI models with large parameter changes.

[0272] Optionally, if the change in the parameters of the AI ​​model is less than a certain threshold, then there is no need to switch gaps.

[0273] Optionally, when the type and / or structure are the same, if the parameter change is less than the third threshold, there is no need to switch the gap.

[0274] Optionally, when the type difference is less than the first threshold and / or the structural difference is less than the second threshold, the parameter change is less than the third threshold, and there is no need to switch the gap.

[0275] Optionally, the greater the total capacity or resources of the AI ​​computing unit of the first communication device, the smaller the switching gap. The first, second, and third thresholds mentioned above are related to the total capacity or resources.

[0276] The phrase "remaining capacity or resources of the AI ​​computing unit of the first communication device" can be understood as the remaining capacity or resources after other functional modules (communication modules and non-communication modules) have occupied the capacity / resources of the AI ​​computing unit. Optionally, the stronger the remaining capacity / more remaining resources, the smaller the switching gap. The aforementioned first threshold, second threshold, and third threshold are related to the remaining capacity or remaining resources.

[0277] Furthermore, the method also includes:

[0278] The second communication device receives the first information sent by the first communication device.

[0279] For example, when the first information meets a preset condition, the first communication device sends the first information to the second communication device; for example, the preset condition is: whether at least one of the contents included in the first information exceeds a threshold, or whether at least one of the contents has changed, or whether the amount of change of at least one of the contents exceeds a threshold.

[0280] or,

[0281] When the target trigger is met, the first communication device sends the first information to the second communication device; for example, as instructed by the second communication device and / or agreed upon by the protocol, the specific content, period, triggering event, and other information of the first information. If the remaining capacity / resources of the AI ​​computing unit reported by the first communication device change, and the previously configured switching gap length is no longer suitable and needs to be adjusted, then the first communication device reports the first information.

[0282] or,

[0283] The first communication device periodically sends the first information to the second communication device; for example, the specific content, period, and other information of the first information are specified by the second communication device and / or agreed upon by the protocol.

[0284] As an optional embodiment, the capabilities or resources of the AI ​​computing unit include at least one of the following:

[0285] Computational skills (multiplication, addition, etc.);

[0286] Computing power or computing ability;

[0287] Floating-point computing power; such as FLOPS (floating-point operations per second): the number of floating-point operations performed per second. Floating-point operations include all operations involving decimals and are more time-consuming than integer operations. TFLOPS (Tops) is commonly used as a metric to measure NPU / GPU performance / computing power. For example, a chip with a computing power of 1.7 Tops represents its neural network computing performance.

[0288] Number of processes;

[0289] Number of threads;

[0290] Memory;

[0291] Storage space;

[0292] cache;

[0293] Input channel information, such as the bandwidth, rate, and number of channels / bits / bytes that can be read / written / transmitted simultaneously;

[0294] Output channel information, such as the bandwidth, rate, and number of channels / bits / bytes that can be read / written / transmitted simultaneously.

[0295] As an optional embodiment, the communication process applicable to the switching gap includes at least one of the following:

[0296] Cell handover;

[0297] Cell redirection;

[0298] Radio Resource Control (RRC) reconfiguration;

[0299] RRC release;

[0300] RRC recovery;

[0301] RRC reconstruction;

[0302] RRC established;

[0303] Measure the gap;

[0304] Measure the gap configuration.

[0305] Accordingly, during the above communication process, the duration of the gap switching is less than or equal to any of the following:

[0306] Cell handover latency;

[0307] Cell redirection delay;

[0308] RRC reconfiguration delay;

[0309] RRC reconstruction delay;

[0310] RRC setup delay;

[0311] Measure the length of the gap.

[0312] In an optional embodiment of this application, the start time of the gap switching includes at least one of the following:

[0313] The first moment to instruct the first communication device to switch AI models;

[0314] The X3rd time unit after the first time;

[0315] The X4th time unit prior to the first time;

[0316] The second time of receiving information from the second AI model;

[0317] The X5th time unit after the second time;

[0318] The X6th time unit preceding the second time;

[0319] The third time after the second AI model receives the information and sends a confirmation message;

[0320] The X7th time unit following the third time;

[0321] The X8th time unit preceding the third time;

[0322] The fourth time when the second communication device receives the confirmation information of the second AI model information;

[0323] The X9th time unit following the fourth time;

[0324] The X10th time unit preceding the fourth time;

[0325] The fifth time when the first AI model fails;

[0326] The X11th time unit following the fifth time;

[0327] The X12th time unit preceding the fifth time;

[0328] The second AI model is expected to take effect in the sixth time.

[0329] The X13th time unit following the sixth time;

[0330] The X14th time unit preceding the sixth time;

[0331] Among them, X3 to X14 are all greater than or equal to 1.

[0332] In another optional embodiment of this application, the end time of the switching gap includes at least one of the following: the start time of the switching gap plus the duration of the switching gap, resulting in a seventh time;

[0333] The X15th time unit following the seventh time;

[0334] The X16th time unit prior to the seventh time;

[0335] The eighth time is obtained by adding the start time of the gap switching to the duration of the AI ​​model switching;

[0336] The X17th time unit following the eighth time;

[0337] The X18th time unit preceding the eighth time;

[0338] The ninth time the AI ​​model completes the switch;

[0339] The X19th time unit following the ninth time;

[0340] The X20th time unit prior to the ninth time;

[0341] The tenth time when the first communication device sends the first indication information that the second AI model can use;

[0342] The X21st time unit following the tenth time;

[0343] The X22nd time unit prior to the tenth time;

[0344] The eleventh time when the second communication device receives the first indication information;

[0345] The 23rd time unit following the eleventh time;

[0346] The X24th time unit preceding the eleventh time;

[0347] The twelfth time after the first communication device sends the first indication information and the second communication device sends back confirmation information; the X25th time unit after the twelfth time.

[0348] The X26th time unit preceding the twelfth time;

[0349] The thirteenth time after the first communication device sends the first instruction information and receives the confirmation information from the second communication device;

[0350] The X27th time unit following the thirteenth time;

[0351] The X28th time unit preceding the thirteenth time;

[0352] Among them, X15 to X28 are all greater than or equal to 1.

[0353] As an optional embodiment, step 301 includes:

[0354] The second communication device receives a Gap request message sent by the first communication device;

[0355] The second communication device sends Gap switching configuration information to the first communication device according to the Gap request message.

[0356] Optionally, the first communication device stores multiple AI models and sends a Gap request message to the second communication device based on function requests from other communication devices besides the second communication device. The second communication device then sends a corresponding switchback response to the first communication device based on the Gap request message. For example, positioning may require switching Gap configuration information among different AI models under a Location Management Function (LMF) positioning request.

[0357] In summary, in this embodiment of the application, the second communication device instructs the first communication device to switch AI models. When there is a switching gap in the AI ​​model switching, the operation of the first communication device in the switching gap is restricted, thereby reducing the impact of AI model switching on the communication system, improving the performance of the AI ​​model in the communication system, and thus improving system performance.

[0358] The AI ​​model switching processing method provided in this application can be executed by an AI model switching processing device. This application uses an AI model switching processing device executing the AI ​​model switching processing method as an example to illustrate the AI ​​model switching processing device provided in this application.

[0359] Please see Figure 4 , Figure 4 A schematic diagram of the structure of an AI model switching processing device 400 provided in an embodiment of this application, the AI ​​model switching processing device 400 including:

[0360] The receiving module 401 is used to receive switching indication information sent by the second communication device, the switching indication information being used to instruct the first communication device to switch AI models;

[0361] The first execution module 402 is configured to perform a first operation during the switching gap when there is a switching interval gap between the first AI model and the second AI model; the first operation includes any one of the following:

[0362] The calculations for the first AI model and the second AI model are not performed;

[0363] Stop using the AI ​​computing unit;

[0364] Perform the calculation of the first AI model, and the calculation result is available or valid;

[0365] The calculation of the first AI model and / or the second AI model is performed, and the calculation results are unavailable or invalid;

[0366] Discontinue the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit.

[0367] As an optional embodiment, the first operation further includes at least one of the following:

[0368] The calculation is performed using the non-AI algorithm corresponding to the first AI model;

[0369] The calculation is performed using the non-AI algorithm corresponding to the second AI model.

[0370] As an optional embodiment, the apparatus further includes:

[0371] The second execution module is configured to perform a second operation when there is no switching gap or when the switching gap ends during the switching between the first AI model and the second AI model; the second operation includes any one of the following:

[0372] The second AI model is used for calculation;

[0373] Send a first indication message that the second AI model can be used to the second communication device, and use the second AI model to perform calculations after X1 time units;

[0374] Send a first indication message that the second AI model can be used to the second communication device, receive a confirmation message for the first indication message, and use the second AI model to perform calculations after receiving the confirmation message for X2 time units;

[0375] Wherein, X1 and X2 are both greater than or equal to 0.

[0376] As an optional embodiment, the apparatus further includes:

[0377] The first determining module is used to determine the switching Gap configuration information, which includes at least one of the following:

[0378] The content of the first operation;

[0379] The second operation involves:

[0380] Switch the start time of the gap;

[0381] Switch the end time of the gap;

[0382] Toggle the duration of the gap;

[0383] The cycle for switching gaps;

[0384] Is there a gap switching issue?

[0385] Switch the communication process applicable to the Gap.

[0386] As an optional embodiment, the first determining module includes:

[0387] The first determining submodule is used to receive the switching gap configuration information indicated by the second communication device;

[0388] Alternatively, it can be used to send a Gap request message to the second communication device according to the function request of other communication devices, and to receive the switching Gap configuration information returned by the second communication device;

[0389] Alternatively, it can be used to send Gap switching configuration information to a second communication device;

[0390] Alternatively, it can be used to determine the switching gap configuration information to be used based on a pre-agreed agreement.

[0391] Alternatively, it can be used to determine multiple sets of switching gap configuration information according to a pre-agreed agreement, and receive instructions from the first communication device to determine the switching gap configuration information to be used.

[0392] As an optional embodiment, the apparatus further includes:

[0393] The second determining submodule is used to determine, based on the first information, whether there is a switching gap and / or the duration of the switching gap when switching between the first AI model and the second AI model.

[0394] The first information includes at least one of the following:

[0395] Information on the type differences between the first AI model and the second AI model;

[0396] Structural differences between the first AI model and the second AI model;

[0397] The parameter changes between the first AI model and the second AI model;

[0398] The total capacity or resources of the AI ​​computing unit of the first communication device;

[0399] The remaining capacity or resources of the AI ​​computing unit of the first communication device.

[0400] As an optional embodiment, the capabilities or resources of the AI ​​computing unit include at least one of the following:

[0401] Computational ability;

[0402] Computing power or computing ability;

[0403] Floating-point computing power;

[0404] Number of processes;

[0405] Number of threads;

[0406] Memory;

[0407] Storage space;

[0408] cache;

[0409] Input channel information;

[0410] Output channel information.

[0411] As an optional embodiment, the apparatus further includes:

[0412] An information sending module is used to send the first information to the second communication device when the first information meets preset conditions;

[0413] Alternatively, it can be used to send the first information to the second communication device when the target trigger is met;

[0414] Alternatively, it can be used to periodically send the first information to the second communication device.

[0415] As an optional embodiment, the communication process applicable to the switching gap includes at least one of the following:

[0416] Cell handover;

[0417] Cell redirection;

[0418] Radio Resource Control (RRC) reconfiguration;

[0419] RRC release;

[0420] RRC recovery;

[0421] RRC reconstruction;

[0422] RRC established;

[0423] Measure the gap;

[0424] Measure the gap configuration.

[0425] As an optional embodiment, the duration of the switching gap is less than or equal to any of the following:

[0426] Cell handover latency;

[0427] Cell redirection delay;

[0428] RRC reconfiguration delay;

[0429] RRC reconstruction delay;

[0430] RRC setup delay;

[0431] Measure the length of the gap.

[0432] As an optional embodiment, the first determining module includes:

[0433] The second determining submodule is used to determine the switching Gap configuration information based on the first message; wherein the first message includes at least one of the following:

[0434] RRC reconfiguration message;

[0435] RRC release message;

[0436] RRC recovery message;

[0437] RRC reconstruction message;

[0438] RRC establishes a message.

[0439] As an optional embodiment, the start time of the gap switching includes at least one of the following:

[0440] The first moment to instruct the first communication device to switch AI models;

[0441] The X3rd time unit after the first time;

[0442] The X4th time unit prior to the first time;

[0443] The second time of receiving information from the second AI model;

[0444] The X5th time unit after the second time;

[0445] The X6th time unit preceding the second time;

[0446] The third time after the second AI model receives the information and sends a confirmation message;

[0447] The X7th time unit following the third time;

[0448] The X8th time unit preceding the third time;

[0449] The fourth time when the second communication device receives the confirmation information of the second AI model information;

[0450] The X9th time unit following the fourth time;

[0451] The X10th time unit preceding the fourth time;

[0452] The fifth time when the first AI model fails;

[0453] The X11th time unit following the fifth time;

[0454] The X12th time unit preceding the fifth time;

[0455] The second AI model is expected to take effect in the sixth time.

[0456] The X13th time unit following the sixth time;

[0457] The X14th time unit preceding the sixth time;

[0458] Among them, X3 to X14 are all greater than or equal to 1.

[0459] As an optional embodiment, the end time of the switching gap includes at least one of the following:

[0460] The seventh time is obtained by adding the start time of the gap switching to the duration of the gap switching;

[0461] The X15th time unit following the seventh time;

[0462] The X16th time unit prior to the seventh time;

[0463] The eighth time is obtained by adding the start time of the gap switching to the duration of the AI ​​model switching;

[0464] The X17th time unit following the eighth time;

[0465] The X18th time unit preceding the eighth time;

[0466] The ninth time the AI ​​model completes the switch;

[0467] The X19th time unit following the ninth time;

[0468] The X20th time unit prior to the ninth time;

[0469] The tenth time when the first communication device sends the first indication information that the second AI model can use;

[0470] The X21st time unit following the tenth time;

[0471] The X22nd time unit prior to the tenth time;

[0472] The eleventh time when the second communication device receives the first indication information;

[0473] The 23rd time unit following the eleventh time;

[0474] The X24th time unit preceding the eleventh time;

[0475] The twelfth time after the first communication device sends the first instruction information and the second communication device sends back the confirmation information;

[0476] The X25th time unit following the twelfth time;

[0477] The X26th time unit preceding the twelfth time;

[0478] The thirteenth time after the first communication device sends the first instruction information and receives the confirmation information from the second communication device;

[0479] The X27th time unit following the thirteenth time;

[0480] The X28th time unit preceding the thirteenth time;

[0481] Among them, X15 to X28 are all greater than or equal to 1.

[0482] In this embodiment, the second communication device instructs the first communication device to switch AI models. When there is a switching gap in the AI ​​model switching, the operation of the first communication device in the switching gap is restricted to reduce the impact of AI model switching on the communication system, improve the performance of the AI ​​model in the communication system, and thus improve system performance.

[0483] It should be noted that the AI ​​model switching processing device provided in this application embodiment is a device capable of executing the above-described AI model switching processing method. Therefore, all embodiments of the above-described AI model switching processing method are applicable to the AI ​​model switching processing device and can achieve the same or similar beneficial effects.

[0484] Please see Figure 5 , Figure 5 A schematic diagram of the structure of an AI model switching processing device 500 provided in an embodiment of this application, the AI ​​model switching processing device 500 includes:

[0485] The sending module 501 is configured to send switching gap configuration information to the first communication device, wherein the switching gap configuration information includes at least one of the following:

[0486] The first operation is the operation of the first communication device during the switching gap between the first AI model and the second AI model.

[0487] The second operation involves the operation of the first communication device when there is no gap switching or after the gap switching is completed.

[0488] Switch the start time of the gap;

[0489] Switch the end time of the gap;

[0490] Toggle the duration of the gap;

[0491] The cycle for switching gaps;

[0492] Is there a gap switching issue?

[0493] Switch the communication process applicable to the Gap.

[0494] As an optional embodiment, the first operation includes any of the following:

[0495] The first communication device does not perform calculations on the first AI model and the second AI model;

[0496] The first communication device ceased using the AI ​​computing unit;

[0497] The first communication device performs calculations on the first AI model, and the calculation results are available or valid;

[0498] The first communication device performs calculations on the first AI model and / or the second AI model, and the calculation results are unavailable or invalid;

[0499] The first communication device ceases to use the AI ​​computing unit and the communication module associated with the AI ​​computing unit.

[0500] As an optional embodiment, the first operation further includes at least one of the following:

[0501] The first communication device uses a non-AI algorithm corresponding to the first AI model to perform calculations;

[0502] The first communication device uses a non-AI algorithm corresponding to the second AI model to perform calculations.

[0503] As an optional embodiment, the second operation includes any of the following:

[0504] The first communication device uses the second AI model to perform calculations;

[0505] The first communication device sends a first indication information to the second communication device indicating that the second AI model can be used, and performs calculations using the second AI model after X1 time units;

[0506] The first communication device sends a first indication information to the second communication device that the second AI model can be used, receives confirmation information of the first indication information, and uses the second AI model to perform calculations after receiving the confirmation information for X2 time units.

[0507] Wherein, X1 and X2 are both greater than or equal to 0.

[0508] As an optional embodiment, the apparatus further includes:

[0509] The third determining module is configured to determine the switching gap configuration information used by the first communication device based on the first information; wherein the first information includes at least one of the following:

[0510] Information on the type differences between the first AI model and the second AI model;

[0511] Structural differences between the first AI model and the second AI model;

[0512] The parameter changes between the first AI model and the second AI model;

[0513] The total capacity or resources of the AI ​​computing unit of the first communication device;

[0514] The remaining capacity or resources of the AI ​​computing unit of the first communication device.

[0515] As an optional embodiment, the apparatus further includes:

[0516] The information receiving module is used to receive the first information sent by the first communication device.

[0517] As an optional embodiment, the capabilities or resources of the AI ​​computing unit include at least one of the following:

[0518] Computational ability;

[0519] Computing power or computing ability;

[0520] Floating-point computing power;

[0521] Number of processes;

[0522] Number of threads;

[0523] Memory;

[0524] Storage space;

[0525] cache;

[0526] Input channel information;

[0527] Output channel information.

[0528] As an optional embodiment, the communication process applicable to the handover gap includes at least one of the following: cell handover;

[0529] Cell redirection;

[0530] Radio Resource Control (RRC) reconfiguration;

[0531] RRC release;

[0532] RRC recovery;

[0533] RRC reconstruction;

[0534] RRC established;

[0535] Measure the gap;

[0536] Measure the gap configuration.

[0537] As an optional embodiment, the start time of the switching gap includes at least one of the following: a first time instructing the first communication device to switch the AI ​​model;

[0538] The X3rd time unit after the first time;

[0539] The X4th time unit prior to the first time;

[0540] The second time of receiving information from the second AI model;

[0541] The X5th time unit after the second time;

[0542] The X6th time unit preceding the second time;

[0543] The third time after the second AI model receives the information and sends a confirmation message;

[0544] The X7th time unit following the third time;

[0545] The X8th time unit preceding the third time;

[0546] The fourth time when the second communication device receives the confirmation information of the second AI model information;

[0547] The X9th time unit following the fourth time;

[0548] The X10th time unit preceding the fourth time;

[0549] The fifth time when the first AI model fails;

[0550] The X11th time unit following the fifth time;

[0551] The X12th time unit preceding the fifth time;

[0552] The second AI model is expected to take effect in the sixth time.

[0553] The X13th time unit following the sixth time;

[0554] The X14th time unit preceding the sixth time;

[0555] Among them, X3 to X14 are all greater than or equal to 1.

[0556] As an optional embodiment, the end time of the switching gap includes at least one of the following:

[0557] The seventh time is obtained by adding the start time of the gap switching to the duration of the gap switching;

[0558] The X15th time unit following the seventh time;

[0559] The X16th time unit prior to the seventh time;

[0560] The eighth time is obtained by adding the start time of the gap switching to the duration of the AI ​​model switching;

[0561] The X17th time unit following the eighth time;

[0562] The X18th time unit preceding the eighth time;

[0563] The ninth time the AI ​​model completes the switch;

[0564] The X19th time unit following the ninth time;

[0565] The X20th time unit prior to the ninth time;

[0566] The tenth time when the first communication device sends the first indication information that the second AI model can use;

[0567] The X21st time unit following the tenth time;

[0568] The X22nd time unit prior to the tenth time;

[0569] The eleventh time when the second communication device receives the first indication information;

[0570] The 23rd time unit following the eleventh time;

[0571] The X24th time unit preceding the eleventh time;

[0572] The twelfth time after the first communication device sends the first instruction information and the second communication device sends back the confirmation information;

[0573] The X25th time unit following the twelfth time;

[0574] The X26th time unit preceding the twelfth time;

[0575] The thirteenth time after the first communication device sends the first instruction information and receives the confirmation information from the second communication device;

[0576] The X27th time unit following the thirteenth time;

[0577] The X28th time unit preceding the thirteenth time;

[0578] Among them, X15 to X28 are all greater than or equal to 1.

[0579] As an optional embodiment, the sending module includes:

[0580] The request receiving submodule is used to receive the Gap request message sent by the first communication device;

[0581] The sending submodule is used to send Gap switching configuration information to the first communication device according to the Gap request message.

[0582] In this embodiment, the second communication device instructs the first communication device to switch AI models. When there is a switching gap in the AI ​​model switching, the operation of the first communication device in the switching gap is restricted to reduce the impact of AI model switching on the communication system, improve the performance of the AI ​​model in the communication system, and thus improve system performance.

[0583] It should be noted that the AI ​​model switching processing device provided in this application embodiment is a device capable of executing the above-described AI model switching processing method. Therefore, all embodiments of the above-described AI model switching processing method are applicable to the AI ​​model switching processing device and can achieve the same or similar beneficial effects.

[0584] The AI ​​model switching processing device in this application embodiment 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, or other devices besides a terminal. For example, the terminal can include, but is not limited to, the type of terminal 11 listed above; other devices can be servers, network attached storage (NAS), etc., and this application embodiment does not specifically limit the type.

[0585] The AI ​​model switching processing device provided in this application embodiment can achieve Figures 1 to 3 The various processes implemented in the method embodiments achieve the same technical effect, and will not be described again here to avoid repetition.

[0586] Optional, such as Figure 6 As shown, this application embodiment also provides a communication device 600, including a processor 601 and a memory 602. The memory 602 stores programs or instructions that can run on the processor 601. For example, when the communication device 600 is a terminal, the program or instructions executed by the processor 601 implement the various steps of the above-described AI model switching processing method embodiment and achieve the same technical effect. When the communication device 600 is a network-side device, the program or instructions executed by the processor 601 implement the various steps of the above-described AI model switching processing method embodiment and achieve the same technical effect. To avoid repetition, further details are omitted here.

[0587] This application embodiment also provides a terminal, including a processor and a communication interface. The communication interface is used to receive switching instruction information sent by a second communication device, the switching instruction information being used to instruct the first communication device to switch AI models. The processor is used to perform a first operation during a switching gap when there is a switching interval gap between the first AI model and the second AI model. The first operation includes any one of the following: not performing calculations for the first AI model and the second AI model; stopping the use of the AI ​​computing unit; performing calculations for the first AI model, and the calculation result is available or valid; performing calculations for the first AI model and / or the second AI model, and the calculation result is unavailable or invalid; stopping the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit. Alternatively, the communication interface is used to send switching gap configuration information to the first communication device, the switching gap configuration information including at least one of the following: the content of the first operation; the first operation being the operation of the first communication device during a switching gap; the content of the second operation; the second operation being the operation of the first communication device when there is no switching gap or after the switching gap has ended; the start time of the switching gap; the end time of the switching gap; the duration of the switching gap; the period of the switching gap; whether a switching gap exists; and the communication process applicable to the switching gap. This terminal embodiment corresponds to the aforementioned terminal-side method embodiment. All implementation processes and methods of the aforementioned method embodiments can be applied to this terminal embodiment and achieve the same technical effect. Specifically, Figure 7 A schematic diagram of the hardware structure of a terminal to implement an embodiment of this application.

[0588] The terminal 700 includes, but is not limited to, at least some of the following components: radio frequency unit 701, network module 702, audio output unit 703, input unit 704, sensor 705, display unit 706, user input unit 707, interface unit 708, memory 709, and processor 710.

[0589] Those skilled in the art will understand that the terminal 700 may also include a power supply (such as a battery) for powering various components. The power supply can be logically connected to the processor 710 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. 7 The terminal structure shown does not constitute a limitation on the terminal. The terminal may include more or fewer components than shown, or combine certain components, or have different component arrangements, which will not be elaborated here.

[0590] It should be understood that, in this embodiment, the input unit 704 may include a graphics processing unit (GPU) 7041 and a microphone 7042. The GPU 7041 processes image data of still images or videos obtained by an image capture device (such as a camera) in video capture mode or image capture mode. The display unit 706 may include a display panel 7061, which may be configured in the form of a liquid crystal display, an organic light-emitting diode, or the like. The user input unit 707 includes at least one of a touch panel 7071 and other input devices 7072. The touch panel 7071 is also called a touch screen. The touch panel 7071 may include a touch detection device and a touch controller. Other input devices 7072 may include, but are not limited to, a physical keyboard, function keys (such as volume control buttons, power buttons, etc.), a trackball, a mouse, and a joystick, which will not be described in detail here.

[0591] In this embodiment, after receiving downlink data from the network-side device, the radio frequency unit 701 can transmit it to the processor 710 for processing; in addition, the radio frequency unit 701 can send uplink data to the network-side device. Typically, the radio frequency unit 701 includes, but is not limited to, an antenna, amplifier, transceiver, coupler, low-noise amplifier, duplexer, etc.

[0592] The memory 709 can be used to store software programs or instructions, as well as various data. The memory 709 may primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, the memory 709 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DRRAM). The memory 709 in the embodiments of this application includes, but is not limited to, these and any other suitable types of memory.

[0593] Processor 710 may include one or more processing units; optionally, processor 710 integrates an application processor and a modem processor, wherein the application processor mainly handles operations involving the operating system, user interface, and applications, and the modem processor mainly handles wireless communication signals, such as a baseband processor. It is understood that the aforementioned modem processor may also not be integrated into processor 710.

[0594] The radio frequency unit 701 is used to receive switching indication information sent by the second communication device, the switching indication information being used to instruct the first communication device to switch AI models.

[0595] The processor 710 is configured to perform a first operation during a switching gap when there is a switching interval gap between the first AI model and the second AI model; the first operation includes any one of the following: not performing calculations on the first AI model and the second AI model; stopping the use of the AI ​​computing unit; performing calculations on the first AI model, and the calculation result is available or valid; performing calculations on the first AI model and / or the second AI model, and the calculation result is unavailable or invalid; stopping the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit.

[0596] Alternatively, radio frequency unit 701 is used to send switching gap configuration information to the first communication device. The switching gap configuration information includes at least one of the following: the content of a first operation; the first operation is the operation of the first communication device during a switching gap; the content of a second operation; the second operation is the operation of the first communication device when there is no switching gap or after the switching gap has ended; the start time of the switching gap; the end time of the switching gap; the duration of the switching gap; the period of the switching gap; whether there is a switching gap; and the communication process applicable to the switching gap.

[0597] In this embodiment, the second communication device instructs the first communication device to switch AI models. When there is a switching gap in the AI ​​model switching, the operation of the first communication device in the switching gap is restricted to reduce the impact of AI model switching on the communication system, improve the performance of the AI ​​model in the communication system, and thus improve system performance.

[0598] It should be noted that the terminal provided in this application embodiment is a terminal capable of executing the above-described AI model switching processing method. Therefore, all embodiments of the above-described AI model switching processing method are applicable to this terminal and can achieve the same or similar beneficial effects.

[0599] This application embodiment also provides a network-side device, including a processor and a communication interface. The communication interface is used to receive switching indication information sent by a second communication device, the switching indication information being used to instruct the first communication device to switch AI models. The processor is used to perform a first operation during the switching gap when there is a switching interval gap between the first AI model and the second AI model. The first operation includes any one of the following: not performing calculations on the first AI model and the second AI model; stopping the use of the AI ​​computing unit; performing calculations on the first AI model, and the calculation result is available or valid; performing calculations on the first AI model and / or the second AI model, and the calculation result is unavailable or invalid; stopping the use of the AI ​​computing unit and the communication module associated with the AI ​​computing unit. Alternatively, the communication interface is used to send switching gap configuration information to the first communication device. The switching gap configuration information includes at least one of the following: the content of a first operation; the first operation being the operation of the first communication device during a switching gap; the content of a second operation; the second operation being the operation of the first communication device when there is no switching gap or after the switching gap has ended; the start time of the switching gap; the end time of the switching gap; the duration of the switching gap; the period of the switching gap; whether a switching gap exists; and the communication process applicable to the switching gap. This network-side device embodiment corresponds to the above-described network-side device method embodiment. All implementation processes and methods of the above method embodiments can be applied to this network-side device embodiment and achieve the same technical effects.

[0600] Specifically, embodiments of this application also provide a network-side device. For example... Figure 8 As shown, the network-side device 800 includes: an antenna 81, a radio frequency (RF) device 82, a baseband device 83, a processor 84, and a memory 85. The antenna 81 is connected to the RF device 82. In the uplink direction, the RF device 82 receives information through the antenna 81 and transmits the received information to the baseband device 83 for processing. In the downlink direction, the baseband device 83 processes the information to be transmitted and sends it to the RF device 82. The RF device 82 processes the received information and transmits it through the antenna 81.

[0601] The method executed by the network-side device in the above embodiments can be implemented in the baseband device 83, which includes a baseband processor.

[0602] Baseband device 83 may include, for example, at least one baseband board on which multiple chips are disposed, such as... Figure 8 As shown, one of the chips is, for example, a baseband processor, which is connected to the memory 85 via a bus interface to call the program in the memory 85 and execute the network device operation shown in the above method embodiment.

[0603] The network-side device may also include a network interface 86, such as a common public radio interface (CPRI).

[0604] Specifically, the network-side device 800 of this embodiment further includes: instructions or programs stored in a memory 85 and executable on a processor 84, wherein the processor 84 calls the instructions or programs in the memory 85 to execute. Figure 4 or Figure 5 The methods executed by each module shown achieve the same technical effect, and to avoid repetition, they will not be described in detail here.

[0605] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described AI model switching processing method embodiment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0606] The processor is the processor in the terminal described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0607] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described AI model switching processing method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0608] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0609] This application also provides a computer program / program product, which is stored in a storage medium and executed by at least one processor to implement the various processes of the above-described AI model switching processing method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0610] This application also provides a wireless communication system, including: a first communication device and a second communication device, wherein the first communication device can be used to execute the steps of the AI ​​model switching processing method described above, and the second communication device can be used to execute the steps of the AI ​​model switching processing method described above.

[0611] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0612] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0613] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for switching artificial intelligence (AI) models, characterized in that, The method comprises: A first communication device receives switching indication information sent by a second communication device, the switching indication information being used to instruct the first communication device to switch AI models; In the case that switching Gap exists in the switching between a first AI model and a second AI model, the first communication device performs a first operation in the switching Gap; the first operation comprises any one of the following: No calculation of the first AI model and the second AI model is performed; The use of an AI operation unit is stopped; Calculation of the first AI model is performed, and the calculation result is available or valid; Calculation of the first AI model and / or the second AI model is performed, and the calculation result is unavailable or invalid; The use of an AI operation unit and a communication module associated with the AI operation unit is stopped; The method further comprises: after the switching Gap between the first AI model and the second AI model ends, the first communication device performs a second operation; the second operation comprises any one of the following: Calculation is performed using the second AI model; First indication information that the second AI model is available is sent to the second communication device, and calculation is performed using the second AI model after X1 time units; First indication information that the second AI model is available is sent to the second communication device, confirmation information of the first indication information is received, and calculation is performed using the second AI model after X2 time units from the receipt of the confirmation information; Wherein, X1 and X2 are both greater than or equal to 0.

2. The method of claim 1, wherein, The first operation further comprises at least one of the following: Calculation is performed using a non-AI algorithm corresponding to the first AI model; Calculation is performed using a non-AI algorithm corresponding to the second AI model.

3. The method of claim 1, wherein, The method further comprises: The first communication device determines switching Gap configuration information, the switching Gap configuration information comprising: the content of the first operation; The content of the second operation; The switching Gap configuration information further comprises at least one of the following: The start time of the switching Gap; The end time of the switching Gap; The duration of the switching Gap; The period of the switching Gap; Whether the switching Gap exists; The communication process to which the switching Gap is applicable.

4. The method of claim 3, wherein, The first communication device determines the switching Gap configuration information, comprising: The first communication device receives switching Gap configuration information indicated by the second communication device; Or, The first communication device sends a Gap request message to the second communication device according to the function request of another communication device, and receives switching Gap configuration information returned by the second communication device; Or, The first communication device sends switching Gap configuration information to the second communication device; Or, The first communication device determines the switching Gap configuration information to be used according to a prior agreement; Or, The first communication device determines multiple sets of switching Gap configuration information according to a prior agreement, receives an indication of the first communication device, and determines the switching Gap configuration information to be used.

5. The method of claim 1, wherein, The method further comprises: The first communication device determines whether switching Gap exists and / or the duration of the switching Gap in the switching between the first AI model and the second AI model according to first information; Wherein, the first information comprises at least one of the following: Type difference information of the first AI model and the second AI model; Structure difference information of the first AI model and the second AI model; Parameter variation of the first AI model and the second AI model; Total capability or total resource of the AI operation unit of the first communication device; Remaining capability or remaining resource of the AI operation unit of the first communication device.

6. The method of claim 5, wherein, The capability or resource of the AI operation unit includes at least one of: Operation capability; Computing power or computing capability; Floating point computing capability; Process number; Thread number; Memory; Storage space; Cache; Input channel information; Output channel information.

7. The method of claim 5, wherein, The method further includes: In a case where the first information meets a preset condition, the first communication device sends the first information to the second communication device; Or, In a case where a target trigger is met, the first communication device sends the first information to the second communication device; Or, The first communication device periodically sends the first information to the second communication device.

8. The method of claim 1 or 3, wherein, The communication process to which the switching gap is applicable includes at least one of: Cell switching; Cell redirection; Radio resource control (RRC) reconfiguration; RRC release; RRC recovery; RRC reestablishment; RRC establishment; Measurement gap; Measurement gap configuration.

9. The method of claim 8, wherein, The duration of the switching gap is less than or equal to any of the following: Cell switching latency; Cell redirection latency; RRC reconfiguration latency; RRC reestablishment latency; RRC establishment latency; Time length of the measurement gap.

10. The method of claim 8, wherein, The first communication device determines the switching gap configuration information, including: The first communication device determines the switching gap configuration information according to a first message; wherein the first message includes at least one of: RRC reconfiguration message; RRC release message; RRC recovery message; RRC reestablishment message; RRC establishment message.

11. The method of claim 1 or 3, wherein, The start time of the switching gap includes at least one of: First time indicating the first communication device to perform AI model switching; X3th time unit after the first time; X4th time unit before the first time; Second time of receiving second AI model information; X5th time unit after the second time; X6th time unit before the second time; Third time of sending confirmation information after receiving the second AI model information; X7th time unit after the third time; X8th time unit before the third time; Fourth time of receiving confirmation information of the second AI model information by the second communication device; X9th time unit after the fourth time; X10th time unit before the fourth time; Fifth time of invalidation of the first AI model; X11th time unit after the fifth time; X12th time unit before the fifth time; Sixth time of expected validity of the second AI model; X13th time unit after the sixth time; X14th time unit before the sixth time; Wherein, X3 to X14 are greater than or equal to 1.

12. The method of claim 1 or 3, wherein, The end time of the switching gap includes at least one of: a seventh time obtained by the start time of the switching gap + the duration of the switching gap; an X15th time unit after the seventh time; an X16th time unit before the seventh time; an eighth time obtained by the start time of the switching gap + the duration of the AI model switching; an X17th time unit after the eighth time; an X18th time unit before the eighth time; a ninth time when the AI model switching is completed; an X19th time unit after the ninth time; an X20th time unit before the ninth time; a tenth time when the first communication device sends first indication information that the second AI model is available for use; an X21th time unit after the tenth time; an X22th time unit before the tenth time; an eleventh time when the second communication device receives the first indication information; an X23th time unit after the eleventh time; an X24th time unit before the eleventh time; a twelfth time when the second communication device feeds back confirmation information after the first communication device sends the first indication information; an X25th time unit after the twelfth time; an X26th time unit before the twelfth time; a thirteenth time when the first communication device receives the confirmation information fed back by the second communication device after the first communication device sends the first indication information; an X27th time unit after the thirteenth time; an X28th time unit before the thirteenth time; wherein X15 to X28 are greater than or equal to 1.

13. A method for switching artificial intelligence (AI) models, characterized in that, comprising: the second communication device sends switching gap configuration information to the first communication device, wherein the switching gap configuration information comprises: the content of a first operation; the first operation is an operation of the first communication device in a switching gap between the first AI model and the second AI model; the content of a second operation; the second operation is an operation of the first communication device after the switching gap ends; wherein the switching gap configuration information further comprises at least one of the following: the start time of the switching gap; the end time of the switching gap; the duration of the switching gap; the period of the switching gap; whether there is a switching gap; a communication process to which the switching gap is applicable; the first operation comprises any of the following: the first communication device does not perform calculation of the first AI model and the second AI model; the first communication device stops using the AI operation unit; the first communication device performs calculation of the first AI model, and the calculation result is available or valid; the first communication device performs calculation of the first AI model and / or the second AI model, and the calculation result is unavailable or invalid; the first communication device stops using the AI operation unit and the communication module associated with the AI operation unit; the second operation comprises any of the following: the first communication device performs calculation using the second AI model; the first communication device sends first indication information that the second AI model is available for use to the second communication device, and performs calculation using the second AI model after X1 time units; The first communication device sends first indication information that a second AI model is available to the second communication device, receives confirmation information of the first indication information, and performs calculation using the second AI model after X2 time units from receiving the confirmation information. Both X1 and X2 are greater than or equal to 0.

14. The method of claim 13, wherein, The first operation further includes at least one of the following: The first communication device performs calculation using a non-AI algorithm corresponding to the first AI model. The first communication device performs calculation using a non-AI algorithm corresponding to the second AI model.

15. The method of claim 13, wherein, The method further includes: The second communication device determines, according to first information, switching gap configuration information used by the first communication device; wherein the first information includes at least one of the following: Type difference information of the first AI model and the second AI model; Structure difference information of the first AI model and the second AI model; Parameter variation of the first AI model and the second AI model; Total capability or total resource of an AI operation unit of the first communication device; Remaining capability or remaining resource of the AI operation unit of the first communication device.

16. The method of claim 15, wherein, The method further includes: The second communication device receives the first information sent by the first communication device.

17. The method of claim 15, wherein, The capability or resource of the AI operation unit includes at least one of the following: Operation capability; Computing power or calculation capability; Floating point calculation capability; Process number; Thread number; Memory; Storage space; Cache; Input channel information; Output channel information.

18. The method of claim 13, wherein, The communication process to which the switching gap is applicable includes at least one of the following: Cell switching; Cell redirection; Radio resource control (RRC) reconfiguration; RRC release; RRC recovery; RRC reestablishment; RRC establishment; Measurement gap; Measurement gap configuration.

19. The method of claim 13, wherein, The start time of the switching gap includes at least one of the following: A first time at which the first communication device is instructed to switch AI models; X3 time units after the first time; X4 time units before the first time; A second time at which second AI model information is received; X5 time units after the second time; X6 time units before the second time; A third time at which confirmation information is sent after receiving the second AI model information; X7 time units after the third time; X8 time units before the third time; A fourth time at which the second communication device receives confirmation information of the second AI model information; X9 time units after the fourth time; X10 time units before the fourth time; A fifth time at which the first AI model is disabled; X11 time units after the fifth time; X12 time units before the fifth time; A sixth time at which the second AI model is expected to be enabled; X13 time units after the sixth time; X14 time units before the sixth time; Both X3 and X14 are greater than or equal to 1.

20. The method of claim 13, wherein, The end time of the switching gap includes at least one of the following: A seventh time obtained by adding the start time of the switching gap to the duration of the switching gap; X15 time units after the seventh time; X16th time unit before the seventh time; eighth time obtained by adding the start time of the switching Gap and the duration of the AI model switching; X17th time unit after the eighth time; X18th time unit before the eighth time; ninth time when the AI model switching is completed; X19th time unit after the ninth time; X20th time unit before the ninth time; tenth time when the first communication device sends first indication information that the second AI model is available for use; X21th time unit after the tenth time; X22th time unit before the tenth time; eleventh time when the second communication device receives the first indication information; X23th time unit after the eleventh time; X24th time unit before the eleventh time; twelfth time when the second communication device feeds back confirmation information after the first communication device sends the first indication information; X25th time unit after the twelfth time; X26th time unit before the twelfth time; thirteenth time when the first communication device receives the confirmation information fed back by the second communication device after the first communication device sends the first indication information; X27th time unit after the thirteenth time; X28th time unit before the thirteenth time. Wherein, X15 to X28 are greater than or equal to 1.

21. The method of claim 13, wherein, The second communication device sends switching Gap configuration information to the first communication device, including: The second communication device receives the Gap request message sent by the first communication device; The second communication device sends switching Gap configuration information to the first communication device according to the Gap request message.

22. A processing device for AI model switching, the processing device comprising: Including: The receiving module is configured to receive switching indication information sent by the second communication device, the switching indication information being used to instruct the first communication device to perform AI model switching; The first execution module is configured to perform a first operation in a switching Gap when the switching Gap exists in the switching of the first AI model and the second AI model; the first operation includes any one of the following: not performing calculation of the first AI model and the second AI model; stopping using the AI operation unit; performing calculation of the first AI model, and the calculation result is available or valid; performing calculation of the first AI model and / or the second AI model, and the calculation result is unavailable or invalid; stopping using the AI operation unit and the communication module associated with the AI operation unit; The device further includes a second execution module configured to perform a second operation after the switching Gap of the first AI model and the second AI model ends; the second operation includes any one of the following: performing calculation using the second AI model; sending first indication information that the second AI model is available for use to the second communication device, and performing calculation using the second AI model after X1 time units; sending first indication information that the second AI model is available for use to the second communication device, receiving confirmation information of the first indication information, and performing calculation using the second AI model after X2 time units of receiving the confirmation information; X1, X2 are both greater than or equal to 0.

23. The apparatus of claim 22, wherein, The first operation further includes at least one of the following: Using a non-AI algorithm corresponding to the first AI model to perform calculation; Using a non-AI algorithm corresponding to the second AI model to perform calculation.

24. The apparatus of claim 22, wherein, The apparatus further includes: A first determining module configured to determine switching gap configuration information, wherein the switching gap configuration information includes: The content of the first operation; The content of the second operation; The switching gap configuration information further includes at least one of the following: The start time of the switching gap; The end time of the switching gap; The duration of the switching gap; The period of the switching gap; Whether there is a switching gap; The communication process to which the switching gap is applicable.

25. The apparatus of claim 24, wherein, The first determining module includes: A first determining submodule configured to receive switching gap configuration information indicated by a second communication device; Or, configured to send a gap request message to the second communication device according to a function request of another communication device, and receive switching gap configuration information returned by the second communication device; Or, configured to send switching gap configuration information to the second communication device; Or, configured to determine the switching gap configuration information to be used according to a prior agreement; Or, configured to determine multiple sets of switching gap configuration information according to a prior agreement, and receive an indication of the first communication device to determine the switching gap configuration information to be used.

26. The apparatus of claim 22, wherein, The apparatus further includes: A second determining submodule configured to determine, according to the first information, whether there is a switching gap and / or the duration of the switching gap when switching between the first AI model and the second AI model; The first information includes at least one of the following: Type difference information between the first AI model and the second AI model; Structural difference information between the first AI model and the second AI model; The parameter variation of the first AI model and the second AI model; The total capability or total resource of the AI operation unit of the first communication device; The remaining capability or remaining resource of the AI operation unit of the first communication device.

27. The apparatus of claim 26, wherein, The capability or resource of the AI operation unit includes at least one of the following: Operation capability; Computing power or computing capability; Floating point computing capability; Process number; Thread number; Memory; Storage space; Cache; Input channel information; Output channel information.

28. The apparatus of claim 26, wherein, The apparatus further includes: An information sending module configured to send the first information to the second communication device if the first information meets a preset condition; Or, configured to send the first information to the second communication device if a target trigger is met; Or, configured to periodically send the first information to the second communication device.

29. The apparatus of claim 22 or 24, wherein, The communication process to which the switching gap is applicable includes at least one of the following: Cell switching; Cell redirection; Radio resource control (RRC) reconfiguration; RRC release; RRC recovery; RRC reestablishment; RRC establishment; Measurement gap; Measurement gap configuration.

30. The apparatus of claim 29, wherein, The duration of the switching gap is less than or equal to any of the following: Cell switching delay; Cell redirection delay; RRC reconfiguration delay; RRC reestablishment delay; RRC establishment delay; The time length of the measurement gap.

31. The apparatus of claim 29, wherein, The first determining module includes: The second determining sub-module is configured to determine the switching gap configuration information according to a first message; wherein the first message comprises at least one of the following: An RRC reconfiguration message; An RRC release message; An RRC resume message; An RRC reestablishment message; An RRC setup message.

32. The apparatus of claim 22 or 24, wherein, The start time of the switching gap comprises at least one of the following: A first time indicating that the first communication device performs AI model switching; An X3th time unit after the first time; An X4th time unit before the first time; A second time of receiving second AI model information; An X5th time unit after the second time; An X6th time unit before the second time; A third time of sending confirmation information after receiving the second AI model information; An X7th time unit after the third time; An X8th time unit before the third time; A fourth time of receiving confirmation information of the second AI model information by the second communication device; An X9th time unit after the fourth time; An X10th time unit before the fourth time; A fifth time of invalidation of the first AI model; An X11th time unit after the fifth time; An X12th time unit before the fifth time; A sixth time of expected validity of the second AI model; An X13th time unit after the sixth time; An X14th time unit before the sixth time; Wherein X3 to X14 are greater than or equal to 1.

33. The apparatus of claim 22 or 24, wherein, The end time of the switching gap comprises at least one of the following: A seventh time obtained by adding the start time of the switching gap and the duration of the switching gap; An X15th time unit after the seventh time; An X16th time unit before the seventh time; An eighth time obtained by adding the start time of the switching gap and the duration of AI model switching; An X17th time unit after the eighth time; An X18th time unit before the eighth time; A ninth time of completing AI model switching; An X19th time unit after the ninth time; An X20th time unit before the ninth time; A tenth time of sending first indication information of availability of the second AI model by the first communication device; An X21th time unit after the tenth time; An X22th time unit before the tenth time; An eleventh time of receiving the first indication information by the second communication device; An X23th time unit after the eleventh time; An X24th time unit before the eleventh time; A twelfth time of feeding back confirmation information by the second communication device after the first communication device sends the first indication information; An X25th time unit after the twelfth time; An X26th time unit before the twelfth time; A thirteenth time of receiving the confirmation information fed back by the second communication device after the first communication device sends the first indication information; An X27th time unit after the thirteenth time; An X28th time unit before the thirteenth time; Wherein X15 to X28 are greater than or equal to 1.

34. A processing device for AI model switching, the processing device comprising: The method comprises: The sending module is configured to send, to the first communication device, switching gap configuration information, wherein the switching gap configuration information comprises: content of a first operation, wherein the first operation is an operation of the first communication device in a switching gap between the first AI model and the second AI model; content of a second operation, wherein the second operation is an operation of the first communication device after the switching gap ends; wherein the switching gap configuration information further comprises at least one of: a start time of the switching gap; an end time of the switching gap; a duration of the switching gap; a period of the switching gap; whether the switching gap exists; a communication process to which the switching gap is applicable; the first operation comprises any one of: the first communication device does not perform calculation of the first AI model and the second AI model; the first communication device stops using an AI operation unit; the first communication device performs calculation of the first AI model, and a calculation result is available or valid; the first communication device performs calculation of the first AI model and / or the second AI model, and a calculation result is unavailable or invalid; the first communication device stops using the AI operation unit and a communication module associated with the AI operation unit; the second operation comprises any one of: the first communication device performs calculation using the second AI model; the first communication device sends, to a second communication device, first indication information that the second AI model is available for use, and performs calculation using the second AI model after X1 time units; the first communication device sends, to a second communication device, first indication information that the second AI model is available for use, receives confirmation information of the first indication information, and performs calculation using the second AI model after X2 time units from receiving the confirmation information; wherein the X1 and the X2 are both greater than or equal to 0.

35. The apparatus of claim 34, wherein, the first operation further comprises at least one of: the first communication device performs calculation using a non-AI algorithm corresponding to the first AI model; the first communication device performs calculation using a non-AI algorithm corresponding to the second AI model.

36. The apparatus of claim 34, wherein, The apparatus further comprises: a third determination module configured to determine, according to first information, switching gap configuration information used by the first communication device, wherein the first information comprises at least one of: type difference information of the first AI model and the second AI model; structure difference information of the first AI model and the second AI model; a parameter variation of the first AI model and the second AI model; total capability or total resource of an AI operation unit of the first communication device; remaining capability or remaining resource of the AI operation unit of the first communication device.

37. The device of claim 36, wherein, The apparatus further comprises: an information receiving module configured to receive the first information sent by the first communication device.

38. The device of claim 36, wherein, The capability or resource of the AI operation unit comprises at least one of: operation capability; computing power or calculation capability; floating point calculation capability; process number; thread number; memory; storage space; cache; input channel information; output channel information.

39. The device of claim 34, wherein, The communication process to which the switching gap is applicable comprises at least one of: cell switching; cell redirection; radio resource control (RRC) reconfiguration; RRC release; RRC recovery; RRC reestablishment; RRC establishment; measurement gap; The measurement gap configuration.

40. The apparatus of claim 34, wherein, The start time of the switching gap includes at least one of: A first time indicating the first communication device to perform AI model switching; An X3th time unit after the first time; An X4th time unit before the first time; A second time of receiving second AI model information; An X5th time unit after the second time; An X6th time unit before the second time; A third time of sending confirmation information after receiving the second AI model information; An X7th time unit after the third time; An X8th time unit before the third time; A fourth time of receiving confirmation information of the second communication device receiving the second AI model information; An X9th time unit after the fourth time; An X10th time unit before the fourth time; A fifth time of the first AI model being invalid; An X11th time unit after the fifth time; An X12th time unit before the fifth time; A sixth time of the second AI model being expected to be valid; An X13th time unit after the sixth time; An X14th time unit before the sixth time; Wherein, X3 to X14 are greater than or equal to 1.

41. The apparatus of claim 34, wherein, The end time of the switching gap includes at least one of: A seventh time obtained by adding the start time of the switching gap and the duration of the switching gap; An X15th time unit after the seventh time; An X16th time unit before the seventh time; An eighth time obtained by adding the start time of the switching gap and the duration of the AI model switching; An X17th time unit after the eighth time; An X18th time unit before the eighth time; A ninth time of the AI model completing switching; An X19th time unit after the ninth time; An X20th time unit before the ninth time; A tenth time of the first communication device sending first indication information indicating that the second AI model is available; An X21th time unit after the tenth time; An X22th time unit before the tenth time; An eleventh time of the second communication device receiving the first indication information; An X23th time unit after the eleventh time; An X24th time unit before the eleventh time; A twelfth time of the second communication device feeding back confirmation information after the first communication device sending the first indication information; An X25th time unit after the twelfth time; An X26th time unit before the twelfth time; A thirteenth time of the first communication device receiving the confirmation information fed back by the second communication device after the first communication device sending the first indication information; An X27th time unit after the thirteenth time; An X28th time unit before the thirteenth time; Wherein, X15 to X28 are greater than or equal to 1.

42. The device of claim 34, wherein, The sending module includes: A request receiving submodule for receiving a gap request message sent by the first communication device; A sending submodule for sending switching gap configuration information to the first communication device according to the gap request message.

43. A first communication device, the first communication device comprising: An apparatus comprising a processor and a memory storing a program or instructions executable on the processor, the program or instructions, when executed by the processor, implement the steps of the processing method for AI model switching according to any one of claims 1 to 12.

44. A second communication device, characterized in that, An apparatus comprising a processor and a memory storing a program or instructions executable on the processor, the program or instructions, when executed by the processor, implement the steps of the processing method for AI model switching according to any one of claims 13 to 21.

45. A readable storage medium, characterized by, A program or instructions stored on a readable storage medium, the program or instructions, when executed by a processor, implement the steps of the processing method for AI model switching according to any one of claims 1 to 12, or implement the steps of the processing method for AI model switching according to any one of claims 13 to 21.

Images