CN113556789A

CN113556789A - Network switching method, network equipment, terminal equipment and chip

Info

Publication number: CN113556789A
Application number: CN202010328501.8A
Authority: CN
Inventors: 赵培尧; 王昭诚; 刘伟; 吴燕鸣; 余小勇; 缑城; 袁文方
Original assignee: Tsinghua University; Huawei Technologies Co Ltd
Current assignee: Tsinghua University; Huawei Technologies Co Ltd
Priority date: 2020-04-23
Filing date: 2020-04-23
Publication date: 2021-10-26
Anticipated expiration: 2040-04-23
Also published as: CN113556789B

Abstract

The application is applicable to the technical field of communication, and provides a network switching method, network equipment, terminal equipment and a chip, wherein the network switching method comprises the following steps: acquiring channel characteristic information of a first network, wherein the channel characteristic information comprises gain variance information, gain amplitude information, time domain information, frequency domain information and/or angle domain information; and inputting the channel characteristic information into the trained SVM model for processing, outputting to obtain switching indication information, and sending the switching indication information to the terminal equipment, wherein the switching indication information is first information used for indicating that the second network is allowed to be accessed or second information used for indicating that the second network is not allowed to be accessed. In the network switching method, the network equipment directly judges whether the terminal equipment is allowed to access the second network according to the channel characteristic information of the first network. The terminal equipment does not need to use a built-in device corresponding to the second network for channel detection, so that the terminal equipment is prevented from using the built-in device for channel detection for a long time, and the power consumption of the terminal equipment is reduced.

Description

Network switching method, network equipment, terminal equipment and chip

Technical Field

The present application belongs to the field of communications technologies, and in particular, to a network switching method, a network device, a terminal device, and a chip.

Background

With the rapid development of communication technology, more and more spectrum resources are available. In order to reduce the construction cost, a base station of a high frequency network (e.g., a millimeter wave network) and a base station of a low frequency network (e.g., a Long Term Evolution (LTE) network) are often deployed in a co-sited manner, that is, the high frequency network and the low frequency network are co-sited. The low frequency network has a relatively long signal wavelength and a relatively large cell coverage area for providing a wide range of communication coverage, and the high frequency network has a relatively short signal wavelength and a relatively small cell coverage area for providing high speed access and capacity enhancement in the area of interest.

When the terminal device needs to be switched from the low-frequency network to the high-frequency network, the terminal device often needs to perform channel detection for a long time by using a device of the high-frequency network, and whether a channel between the terminal device and the network device supports the high-frequency network is judged. The power consumption of the devices of the high-frequency network is often high, so that the terminal equipment maintains the devices of the high-frequency network for a long time to perform channel detection, which causes huge power consumption overhead and resource waste.

Disclosure of Invention

The embodiment of the application provides a network switching method, network equipment, terminal equipment and a chip, and can solve the problems that the terminal equipment maintains a device of a high-frequency network for a long time to perform channel detection, which causes huge power consumption overhead and resource waste.

In a first aspect, the present application provides a network switching method, applied to a network device, including: acquiring channel characteristic information of a first network, wherein the channel characteristic information comprises gain variance information, gain amplitude information, time domain information, frequency domain information and/or angle domain information; inputting the channel characteristic information into a trained Support Vector Machine (SVM) model for processing, and outputting to obtain switching indication information, wherein the switching indication information is first information used for indicating that the second network is allowed to be accessed or second information used for indicating that the second network is not allowed to be accessed, and the frequency band of the first network is lower than that of the second network; and sending the switching indication information to the terminal equipment.

By adopting the network switching method provided by the application, the network equipment can judge whether the terminal equipment is allowed to access the second network or not according to the channel characteristic information of the first network by utilizing the SVM model, and indicate whether the terminal equipment is allowed to access the second network or not through the switching indication information. The terminal device is not required to use a built-in device corresponding to the second network to perform channel detection to detect whether the second network can be accessed. Therefore, the power consumption problem caused by that the terminal equipment uses the device corresponding to the second network for channel detection for a long time is avoided, and the power consumption of the terminal equipment is reduced.

Optionally, the obtaining of the channel characteristic information of the first network includes: receiving a request message sent by a terminal device, wherein the request message requests to be switched from a first network to a second network, and the request message carries Channel characteristic Information, and the Channel characteristic Information is obtained by preprocessing measured Channel State Information (CSI) of the first network by the terminal device.

Optionally, the obtaining of the channel characteristic information of the first network includes:

receiving a request message sent by terminal equipment, wherein the request message requests to be switched from a first network to a second network, and the request message carries CSI (channel state information) of the first network measured by the terminal equipment; and preprocessing the CSI to obtain channel characteristic information.

Based on the two optional modes, the network device may directly obtain the channel characteristic information extracted from the measured CSI by the terminal device, or may extract the channel characteristic information based on the CSI sent by the terminal device.

Optionally, the SVM model is obtained by training based on a training sample set, where the training sample set includes a plurality of channel feature information samples of the first network and switching indication information corresponding to each channel feature information sample, and when the switching indication information output by the SVM model is the first information, the method further includes: if the terminal equipment is successfully switched from the first network to the second network, adding the channel characteristic information and the first information into a training sample set; if the terminal equipment is not successfully switched from the first network to the second network, adding the channel characteristic information and the second information into a training sample set; the SVM model is periodically updated according to the training sample set.

Based on the optional mode, the network equipment periodically updates the SVM model by adopting the training sample set which is continuously updated based on the actual switching condition, so that the SVM model can be further optimized, the error of the SVM model is reduced, and the accuracy of the SVM model is improved.

Optionally, before inputting the channel feature information into the trained SVM model for processing and outputting the obtained switching indication information, the method further includes: detecting whether the first network and the second network are co-sited; and when the first network and the second network are co-located, inputting the channel characteristic information into the trained SVM model for processing, and outputting the obtained switching indication information.

In a second aspect, the present disclosure provides a network switching method applied to a terminal device, including: sending a request message to a network device, wherein the request message is used for requesting to switch from a first network to a second network, the frequency band of the first network is smaller than that of the second network, the request message carries channel characteristic information of the first network, and the channel characteristic information comprises gain variance information, gain amplitude information, time domain information, frequency domain information and/or angle domain information; receiving switching indication information sent by network equipment, wherein the switching indication information is determined by the network equipment according to the channel characteristic information, and the switching indication information is first information used for indicating that the second network is allowed to be accessed or second information used for indicating that the second network is not allowed to be accessed; and when the switching indication information is the first information, switching operation from the first network to the second network is executed.

By adopting the network switching method provided by the application, the terminal equipment sends the channel characteristic information of the first network to the network equipment, so that the network equipment directly judges whether the terminal equipment is allowed to access the second network according to the channel characteristic information of the first network. And then whether to actually execute the switching operation from the first network to the second network according to the switching indication information sent by the network equipment. The terminal device is not required to use a built-in device corresponding to the second network to perform channel detection to detect whether the second network can be accessed. Therefore, the power consumption problem caused by that the terminal equipment uses the device corresponding to the second network for channel detection for a long time is avoided, and the power consumption of the terminal equipment is reduced.

Optionally, before sending the request message to the network device, the method further includes: measuring a channel of a first network to obtain Channel State Information (CSI) of the first network; and preprocessing the CSI to obtain channel characteristic information.

In a third aspect, the present application provides a network switching method, applied to a network device, the method including: receiving a request message sent by a terminal device, wherein the request message requests to be switched from a first network to a second network, the frequency band of the first network is smaller than that of the second network, and the request message carries Channel State Information (CSI) of the first network; inputting the CSI into a trained Convolutional Neural Network (CNN) model for processing, and outputting switching indication information, wherein the switching indication information is first information used for indicating that the second Network is allowed to be accessed or second information used for indicating that the second Network is not allowed to be accessed, and the frequency band of the first Network is lower than that of the second Network; and sending the switching indication information to the terminal equipment.

By adopting the network switching method provided by the application, the network equipment can judge whether the terminal equipment is allowed to access the second network according to the CSI of the first network by utilizing the CNN model, and indicate whether the terminal equipment is allowed to access the second network through the switching indication information. The terminal device is not required to use a built-in device corresponding to the second network to perform channel detection to detect whether the second network can be accessed. Therefore, the power consumption problem caused by that the terminal equipment uses the device corresponding to the second network for channel detection for a long time is avoided, and the power consumption of the terminal equipment is reduced.

Optionally, the CSI carried in the request message may be CSI preprocessed by the terminal device.

Optionally, inputting the CSI into the trained CNN model for processing, and outputting to obtain switching indication information, where the processing includes: and preprocessing the CSI, inputting the preprocessed CSI into the CNN model for processing, and outputting to obtain the switching indication information.

Based on the two optional modes, the network device may determine whether to allow the terminal device to access the second network by using the preprocessed CSI. The operation of preprocessing the CSI may be performed by the terminal device, and may be performed by the network device.

Optionally, the preprocessing includes batch normalization processing, conversion processing from a three-dimensional matrix to a two-dimensional matrix, and separation processing of real parts and imaginary parts.

Optionally, the CNN model is obtained by training based on a training sample set, where the training sample set includes CSI samples of a plurality of first networks and switching indication information corresponding to each CSI sample, and when the switching indication information output by the CNN model is the first information, the method further includes: if the terminal equipment is successfully switched from the first network to the second network, adding the CSI and the first information into a training sample set; if the terminal equipment is not successfully switched from the first network to the second network, adding the CSI and the second information into a training sample set; the CNN model is periodically updated from the training sample set.

Based on the optional mode, the network device periodically updates the CNN model by adopting the training sample set which is continuously updated based on the actual switching condition, so that the CNN model can be further optimized, the error of the CNN model is reduced, and the accuracy of the CNN model is improved.

Optionally, before inputting the channel characteristic information into the trained CNN model for processing and outputting the switching indication information, the method further includes: detecting whether the first network and the second network are co-sited; and when the first network and the second network are co-located, inputting the CSI into the trained CNN model for processing, and outputting the obtained switching indication information.

In a fourth aspect, the present application provides a network switching method, applied to a terminal device, including: sending a request message to a network device, wherein the request message is used for requesting to switch from a first network to a second network, the frequency band of the first network is smaller than that of the second network, the request message carries the preprocessed CSI of the first network, and the preprocessing comprises batch normalization processing, conversion processing from a three-dimensional matrix to a two-dimensional matrix and separation processing of a real part and an imaginary part; receiving switching indication information sent by network equipment, wherein the switching indication information is determined by the network equipment according to the preprocessed CSI, and the switching indication information is first information used for indicating that the second network is allowed to be accessed or second information used for indicating that the second network is not allowed to be accessed; and when the switching indication information is the first information, switching operation from the first network to the second network is executed.

By adopting the network switching method provided by the application, the terminal equipment sends the CSI of the first network to the network equipment, so that the network equipment directly judges whether the terminal equipment is allowed to access the second network according to the CSI of the first network. And then whether to actually execute the switching operation from the first network to the second network according to the switching indication information sent by the network equipment. The terminal device is not required to use a built-in device corresponding to the second network to perform channel detection to detect whether the second network can be accessed. Therefore, the power consumption problem caused by that the terminal equipment uses the device corresponding to the second network for channel detection for a long time is avoided, and the power consumption of the terminal equipment is reduced.

In a fourth aspect, the present application provides a network switching apparatus, which may be a network device, a chip in the network device, a terminal device, or a chip in the terminal device. The network switching apparatus has a function of implementing the method of the first aspect, the second aspect, the third aspect, or the fourth aspect. The function can be realized by hardware, and can also be realized by executing corresponding software by hardware. The hardware or software includes one or more modules corresponding to the functions described above. The network switching device comprises a processing unit and a communication unit, and optionally, the network switching device may further comprise a storage unit. The processing unit may complete the receiving or sending of the information through the communication unit, and the processing unit may process the information so that the network switching apparatus implements the functions of the methods of the first aspect, the second aspect, the third aspect, or the fourth aspect.

When the network switching apparatus is a network device or a chip in a network device, the network switching apparatus implements the method of the first aspect or the third aspect.

In one possible design, when the network switching device is a network device, the processing unit may be, for example, a processor, and the communication unit may include, for example, an antenna, a transceiver, and a communication interface. Optionally, the network device further comprises a storage unit, which may be, for example, a memory. When the network device includes a storage unit, the storage unit is configured to store computer-executable instructions, the processing unit is connected to the storage unit, and the processing unit executes the computer-executable instructions stored in the storage unit, so that the network device performs the method according to the first aspect or any optional manner of the first aspect.

In another possible design, when the apparatus is a chip within a network device, the processing unit may be, for example, a processor, and the communication unit may be, for example, a communication interface. The communication interface may include, among other things, an input/output interface, pins or circuitry, etc. The processing unit may execute the computer executable instructions stored by the storage unit to enable the chip to execute the network switching method according to the first aspect or any optional manner of the first aspect. Optionally, the storage unit is a storage unit in the chip, such as a register, a cache, and the like, and the storage unit may also be a storage unit located outside the chip in the terminal, such as a read-only memory (ROM) or another type of static storage device that can store static information and instructions, a Random Access Memory (RAM), and the like.

When the network switching device is a network device or a chip in the network device, the technical effect of the network switching device may refer to the technical effect of the first aspect or the third aspect, and is not described herein again.

When the network switching device is a terminal device or a chip in the terminal device, the network switching device implements the method of the second aspect or the fourth aspect.

In one possible design, when the network switching device is a terminal device, the processing unit may be, for example, a processor, and the communication unit may include, for example, an antenna, a transceiver, and a communication interface. Optionally, the terminal device further comprises a storage unit, which may be a memory, for example. When the terminal device includes a storage unit, the storage unit is configured to store a computer executable instruction, the processing unit is connected to the storage unit, and the processing unit executes the computer executable instruction stored in the storage unit, so that the terminal device performs the method according to the second aspect or any optional manner of the second aspect.

In another possible design, when the apparatus is a chip in a terminal device, the processing unit may be a processor, for example, and the communication unit may be a communication interface, for example. The communication interface may include, among other things, an input/output interface, pins or circuitry, etc. The processing unit may execute the computer executable instructions stored in the storage unit to enable the chip to execute the network switching method according to the second aspect or any optional manner of the second aspect. Alternatively, the storage unit is a storage unit in the chip, such as a register, a cache, and the like, and the storage unit may also be a storage unit located outside the chip in the terminal, such as a ROM or other types of static storage devices that can store static information and instructions, a RAM, and the like.

When the network switching device is a terminal device or a chip in the terminal device, the technical effect of the network switching device may refer to the technical effect of the second aspect or the fourth aspect, and is not described herein again.

The processor referred to in any above may be a general purpose central processing unit, a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of programs of the method of the first or second aspect.

In a fifth aspect, the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, implements the method of the first, second, third or fourth aspect.

In a sixth aspect, embodiments of the present application provide a computer program product, which, when run on a network device, causes the network device to perform the method of the first aspect or the third aspect.

In a seventh aspect, the present application provides a computer program product, which when run on a terminal device, causes the terminal device to execute the method of the second or fourth aspect.

Drawings

Fig. 1 is a schematic diagram of a communication system provided by an embodiment of the present application;

fig. 2 is a schematic structural diagram of a network device according to an embodiment of the present application;

fig. 3 is a schematic structural diagram of a terminal device according to an embodiment of the present application;

fig. 4 is a flowchart illustrating an embodiment of a network handover method provided in the present application;

fig. 5 is a schematic diagram illustrating a processing flow of CSI provided in an embodiment of the present application;

fig. 6 is a schematic view of a processing flow of the CNN model for input information according to an embodiment of the present application;

fig. 7 is a flowchart illustrating a network handover method according to another embodiment of the present application;

fig. 8 is a schematic structural diagram of a network switching device according to an embodiment of the present application.

Detailed Description

First, the terms "first", "second", and the like in the description and claims of the present application and in the above-described drawings are used for distinguishing different objects, not for limiting a specific order. The terms "exemplary" or "such as" are used herein to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "e.g.," is not necessarily to be construed as preferred or advantageous over other embodiments or designs. Rather, use of the word "exemplary" or "such as" is intended to present concepts related in a concrete fashion.

Unless otherwise indicated, "/" herein generally indicates that the former and latter associated objects are in an "or" relationship, e.g., a/B may represent a or B. The term "and/or" is merely an associative relationship that describes an associated object, meaning that three relationships may exist, e.g., a and/or B, may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the present application, "a plurality" means two or more.

Next, the technical solution provided by the present application is applied to the communication system shown in fig. 1. Referring to fig. 1, the communication system includes at least one network device and at least one terminal device. Wherein the network device supports at least one first network and at least one second network simultaneously.

The frequency band of the first network is lower than that of the second network. For example, the first network may be a network with a frequency band lower than 6GHz, such as a communication network supporting a fourth generation (4G) access technology or a communication network supporting a third generation (3G) access technology, such as a Long Term Evolution (LTE) network, an LTE-a network, and the like. The second network may be a network with a frequency band higher than 10GHz, for example, a millimeter wave network supporting a fifth generation (5G) access technology, a New Radio (NR) network, and the like.

In an embodiment of the present application, the network device may include a first device supporting a first network and a second device supporting a second network. Wherein the first device and the second device are co-sited, i.e. the first network and the second network are co-sited. By co-sited, it may be meant that the first device and the second device are network devices that are connected to each other and independent of each other, the first device and the second device being deployed at the same site. Alternatively, the co-site may also mean that a part of physical structures of the first device and the second device are shared, and another part of physical structures are deployed independently and at the same site. For example, a set of baseband processing units (BBUs) is shared by the first device and the second device, and corresponding Radio Remote Units (RRUs) and antennas are respectively deployed, where an antenna of the first device and an antenna of the second device are deployed on the same antenna tower. Alternatively, co-sited may also mean that the first device and the second device share the same physical structure, that is, the first device and the second device are the same network device, and the functional module of the first device and the functional module of the second device are integrated on the first network device.

In the communication system shown in fig. 1, the network device may be a base station, a Transmission Reception Point (TRP), a relay node (relay node), an Access Point (AP), or the like.

Fig. 2 is a schematic structural diagram of a network device provided in the present application. Referring to fig. 2, the network device includes at least one processor 201, at least one memory 202, at least one transceiver 203, one or more antennas 204, and a communication interface 205. The processor 201, the memory 202, the transceiver 203, the communication interface 205 are connected, for example by a bus. An antenna 204 is connected to the transceiver 203.

The processor 201 in the embodiment of the present application may include at least one of the following types: a general-purpose Central Processing Unit (CPU), a Digital Signal Processor (DSP), a microprocessor, an Application-Specific Integrated Circuit (ASIC), a Microcontroller (MCU), a Field Programmable Gate Array (FPGA), or an Integrated Circuit for implementing logic operations. For example, the processor 201 may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. The at least one processor 201 may be integrated in one chip or located on a plurality of different chips.

The memory 202 in the embodiment of the present application may include at least one of the following types: read-only memory (ROM) or other types of static memory devices that may store static information and instructions, Random Access Memory (RAM) or other types of dynamic memory devices that may store information and instructions, and Electrically erasable programmable read-only memory (EEPROM). In some scenarios, the memory may also be, but is not limited to, a compact disk-read-only memory (CD-ROM) or other optical disk storage, optical disk storage (including compact disk, laser disk, optical disk, digital versatile disk, blu-ray disk, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.

The memory 202 may be separate and coupled to the processor 201. Optionally, the memory 202 may also be integrated with the processor 201, for example, within one chip. The memory 202 can store a program for executing the technical solution of the embodiment of the present application, and is controlled by the processor 201 to execute, and various executed computer program codes can also be regarded as a driver of the processor 201. For example, the processor 201 is configured to execute the computer program code stored in the memory 202, so as to implement the technical solution in the embodiment of the present application.

The transceiver 203 may be configured to support receiving or transmitting air interface signals between the network device and the terminal device, and the transceiver 203 may be connected to the antenna 204. The transceiver 203 includes a transmitter Tx and a receiver Rx. Specifically, one or more antennas 204 may receive an air interface signal, and the receiver Rx of the transceiver 203 is configured to receive the air interface signal from the antennas, convert the air interface signal into a digital baseband signal or a digital intermediate frequency signal, and provide the digital baseband signal or the digital intermediate frequency signal to the processor 201, so that the processor 201 performs further processing on the digital baseband signal or the digital intermediate frequency signal, such as demodulation processing and decoding processing. In addition, the transmitter Tx in the transceiver 203 is further configured to receive a modulated digital baseband signal or a digital intermediate frequency signal from the processor 201, convert the modulated digital baseband signal or the modulated digital intermediate frequency signal into an air interface signal, and transmit the air interface signal through the one or more antennas 204.

In the communication system shown in fig. 1, the terminal device may be a device that supports both the first network and the second network and the mobile communication device. For example, the terminal device may be a mobile phone, a tablet computer, a wearable device, an in-vehicle device, an Augmented Reality (AR)/Virtual Reality (VR) device, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a Personal Digital Assistant (PDA), or other terminal devices. The embodiment of the present application does not set any limit to the specific type of the terminal device.

Fig. 3 is a schematic structural diagram of a terminal device provided in the present application. Referring to fig. 3, the terminal device includes at least one processor 301, at least one memory 302, at least one transceiver 303, one or more antennas 304, a communication interface 305. The processor 301, the memory 302, the transceiver 303, the communication interface 305 are connected, for example by a bus. An antenna 304 is connected to the transceiver 303.

The processor 301 in the embodiment of the present application may include at least one of the following types: a CPU, DSP, microprocessor, ASIC, MCU, FPGA, or integrated circuit for implementing logical operations. For example, processor 301 may be a single core processor or a multi-core processor. The at least one processor 301 may be integrated in one chip or located on multiple different chips.

The memory 302 in the embodiment of the present application may include at least one of the following types: ROM, or other types of static storage devices that may store static information and instructions, and RAM, or other types of dynamic storage devices that may store information and instructions, may be EEPROM. In some scenarios, but is not limited to, a CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.

The memory 302 may be separate and coupled to the processor 301. Alternatively, the memory 302 may be integrated with the processor 301, for example, within a chip. The memory 302 can store a program for executing the technical solution of the embodiment of the present application, and is controlled by the processor 301 to execute, and various executed computer program codes can also be regarded as drivers of the processor 301. For example, the processor 301 is configured to execute the computer program code stored in the memory 302, so as to implement the technical solution in the embodiment of the present application.

The transceiver 303 may be configured to support receiving or transmitting air interface signals between the network device and the terminal device, and the transceiver 303 may be connected to the antenna 304. The transceiver 303 includes a transmitter Tx and a receiver Rx. Specifically, one or more antennas 304 may receive an air interface signal, and the receiver Rx of the transceiver 303 is configured to receive the air interface signal from the antennas, convert the air interface signal into a digital baseband signal or a digital intermediate frequency signal, and provide the digital baseband signal or the digital intermediate frequency signal to the processor 301, so that the processor 301 performs further processing on the digital baseband signal or the digital intermediate frequency signal, such as demodulation processing and decoding processing. In addition, the transmitter Tx in the transceiver 303 is further configured to receive a modulated digital baseband signal or a digital intermediate frequency signal from the processor 301, convert the modulated digital baseband signal or the modulated digital intermediate frequency signal into an air interface signal, and transmit the air interface signal through the one or more antennas 304.

The second network has a relatively shorter signal wavelength and a relatively smaller cell coverage than the first network, but provides higher speed data transmission. Therefore, the terminal device usually uses a device corresponding to the second network to perform channel detection continuously, and when a channel of the second network is detected, the terminal device switches from the first network to the second network to obtain higher-speed data transmission. And under the condition that the channel of the second network cannot be detected or the detection result does not meet the switching condition, the terminal equipment maintains the connection state of the first network.

However, the terminal device uses the device corresponding to the second network for channel detection for a long time, which may result in increasing power consumption of the terminal device. Currently, there is a method for determining whether to switch to a second network by determining whether a Line of Sight (LOS) channel exists in a first network. Namely, the real part and the imaginary part of the Channel State Information (CSI) of the first network are combined, or the amplitude of the CSI is directly extracted and input into the SVM model, so as to obtain the existence label of the LoS path of the low frequency Channel (i.e. the label indicating whether the LoS Channel exists in the first network). An indicator of the availability of the high frequency channel (i.e. an indicator indicating whether the second network is available) is then generated based on the presence tag.

The LOS channel is a direct path between the network device and the terminal device, and no shelter (e.g., human body, vehicle, building, etc.) exists on the direct path, so that the LOS channel has short transmission distance, small reflection LOSs, good channel quality, and can maintain stable and high-speed data transmission. On the contrary, if a Non-direct path exists between the network device and the terminal device, it indicates that the channel between the network device and the terminal device is a Non Line of Sight (NLOS) channel. On the NLOS channel, the signal is blocked by a blocking object, and diffraction and refraction are required, which results in reduced signal intensity and longer transmission distance. The path LOSs of the NLOS channel is on average 20dB greater than the LOS channel over a transmission distance of 100 m.

Compared with the signal wavelength of the first network, the signal wavelength of the second network is short, and the signal has poor diffraction and refraction capabilities and is easily influenced by the shielding object. Since the first network and the second network are co-sited, the channel state of the first network can often directly reflect the channel state of the second network communication. Therefore, when it is detected that the LOS channel exists in the first network, it is indicated at a high probability that the LOS channel also exists in the second network, and at this time, the terminal device may determine that it is possible to switch to the second network.

However, the existing SVM model judges whether the first network is an LOS channel through CSI, and has no robustness to the signal-to-noise ratio of the second network. For example, although the LOS channel exists in the first network, the LOS channel of the second network has poor signal quality and may not support the terminal device to handover to the second network. At this time, the network device still generates a high-frequency channel availability index based on the existence tag of the low-frequency channel LoS path, and instructs the terminal device to switch, thereby causing the terminal device to fail to switch. Or, although the LOS channel does not exist in the first network, the NLOS channel of the second network has better quality (e.g., fewer obstructions), and the terminal device can still be supported to switch to the second network. At this time, the network device still generates a high-frequency channel availability index based on the existence tag of the low-frequency channel LoS path, and indicates the terminal device not to switch, so that the terminal device cannot perform high-speed data transmission.

Therefore, the application provides a network switching method, and the network device directly judges whether the terminal device is allowed to access the second network according to the channel information of the first network. The terminal device is not required to use a built-in device corresponding to the second network to perform channel detection to detect whether the second network can be accessed. Therefore, the power consumption problem caused by that the terminal equipment uses the device corresponding to the second network for channel detection for a long time is avoided, and the power consumption of the terminal equipment is reduced.

The following describes an exemplary network handover method provided by the present application with reference to specific embodiments.

Based on the communication system shown in fig. 1, as shown in fig. 4, a schematic flow chart of an embodiment of a network handover method provided by the present application is shown, where the method includes:

s401, the terminal device sends a request message to the network device, where the request message carries channel information, and the request message is used to request to switch from the first network to the second network.

Wherein the channel information is used to describe a channel status of the first network.

Illustratively, the Channel Information may be Channel State Information (CSI) of the first network. The terminal device may directly obtain Channel State Information (CSI) of the first network by measuring a Channel of the first network.

The CSI of the first network includes a channel gain matrix H, where the channel gain matrix H is a complex matrix with a size of M × N × Q, where M is the number of antennas (i.e., transmit antennas) of the network device, N is the number of antennas (i.e., receive antennas) of the terminal device, and Q is the number of subcarriers. The channel gain matrix H describes the complex gain on each subcarrier for each pair of transceiving antennas of the terminal device and the network device (comprising one antenna of the network device and one antenna of the terminal device) in the first network.

Optionally, the channel information may also be the CSI after being preprocessed. Namely, after measuring the CSI of the first network, the terminal device preprocesses the CSI to obtain channel information.

Illustratively, the pre-processing of the CSI by the terminal device may include extracting channel characteristic information from the CSI. The channel characteristic information may include, but is not limited to, gain variance information, gain magnitude information, time domain information, frequency domain information, and/or angle domain information, among others.

In one example, after measuring the CSI of the first network, the terminal device may extract L (L ≧ 1) eigenvalues f from the CSI₁、f₂、……、f_LThen, the L characteristic values are normalized to obtain L pieces of channel characteristic information.

Exemplarily, the characteristic value f₁May be the coefficient of variance of the channel between the terminal device and the network device. The coefficient of variance is used to reflect the randomness of the channel, and in general, the smaller the coefficient of variance, the more the channel between the terminal device and the network device is prone to the LOS channel. In the embodiment of the present application, when the terminal device calculates the variance coefficient based on the CSI, the terminal device may first calculate an average gain amplitude of each pair of transceiving antennas on Q subcarriers by using the following formula (1):

wherein, mu_mnWhich represents the average gain amplitude of the mth antenna of the network device and the nth antenna of the terminal device over Q subcarriers. Q denotes a subcarrier index, Q is 1, 2, … …, Q. N denotes an antenna index of the terminal device, and N is 1, 2, … …, N. m denotes an antenna of a network deviceIndex, M ═ 1, 2, … …, M.

Then, the terminal device calculates the average variance of each pair of transceiving antennas on Q subcarriers using the following formula (2):

wherein the content of the first and second substances,

representing the mean variance over Q subcarriers of the mth antenna of the network device and the nth antenna of the terminal device.

Finally, the terminal device calculates the variance coefficient f using the following formula (3)₁：

Characteristic value f₂May be the average gain magnitude of the channel between the terminal device and the network device. The average gain magnitude may reflect the degree of attenuation of the channel, and in general, the greater the average gain magnitude, the more the channel between the terminal device and the network device is prone to the LOS channel. For example, after the terminal device calculates the average gain amplitude of each pair of transceiving antennas on Q subcarriers based on the above formula (1), the average gain amplitude f of the channel between the terminal device and the network device may be calculated based on the following formula (4)₂：

Characteristic value f₃May be the time domain peak of the channel between the terminal device and the network device. The time domain peak represents the sharpness value of the peak of the time domain impulse response distribution of the channel, and in general, the larger the average time domain peak, the more the channel between the terminal device and the network device tends to be an LOS channel.

The terminal equipment can add from the channelAnd extracting the subcarrier complex gain matrix corresponding to each pair of transceiving antennas from the gain matrix. And then carrying out inverse Fourier transform on the subcarrier complex gain matrix to obtain time domain channel impulse response between each pair of transceiving antennas. And obtaining a fourth-order central moment after taking an absolute value of each time-domain channel impulse response. For example, the terminal device may first calculate a fourth-order central moment μ obtained by taking an absolute value of a time-domain channel impulse response between the mth antenna of the network device and the nth antenna of the terminal device based on the following formula (5)_4,mn：

Then, the square of the variance calculated after taking the absolute value of the time domain channel impulse response between the m-th antenna of the network equipment and the n-th antenna of the terminal equipment is calculated by using the following formula (6)

Furthermore, the terminal device can calculate the time domain peak value γ between the m-th antenna of the network device and the n-th antenna of the terminal device by using the following formula (7)_mn：

Finally, the terminal device may bring the time domain peak between each pair of transceiving antennas into the following formula (8), and calculate to obtain an average time domain peak f₃：

Characteristic value f₄May be between a terminal device and a network deviceThe average K factor (factor) of the channel of (1). The K factor is a ratio of the energy of the path with the strongest energy in the time domain channel to the sum of the energies of the other paths, and the representation of the K factor in the frequency domain can be shown as the following equation (9). In general, the greater the K factor, the more likely there is a less attenuated LOS path for the channel, i.e. the more the channel between the terminal device and the network device tends towards an LOS channel, the better the channel quality. The terminal device may first calculate the K-factor for each pair of transceiving antennas using the following equation (9):

where ρ is_mnRepresenting the K-factor between the mth antenna of the network device and the nth antenna of the terminal device. Omega_4,mnFourth-order central moment, omega, representing the channel gain between the m-th antenna of the network device and the n-th antenna of the terminal device extracted from the CSI_4,mnCan be calculated by the following equation (10):

ω_2,mnsecond-order central moment, ω, representing the channel gain between the mth antenna of the network device and the nth antenna of the terminal device extracted from the CSI_2,mnCan be calculated by the following formula (11):

after obtaining the K-factors of each pair of transmit-receive antennas, the terminal device can calculate the average K-factor f by using the following formula (12)₄：

Characteristic value f₅May be an average of the channels between the terminal device and the network devicePeak in the angular domain. Generally speaking, if the channel has an angle domain peak value, and the larger the angle domain peak value is, the more the channel between the terminal device and the network device tends to be an LOS channel, the better the channel quality is, indicating that the channel quality is better.

For example, in the channel gain matrix H, for each antenna (i.e., receiving antenna) of the terminal device, the fourier transform is performed on the dimension of the antennas (i.e., transmitting antennas) of the M network devices for the channel coefficient of each subcarrier, so as to obtain an angle domain channel matrix. For example, the channel coefficients for the nth receive antenna form an M × 1 vector [ H (1, n, q), … …, H (M, n, q)]^TAnd carrying out Fourier transform on the vector to obtain an angle domain channel matrix of the nth receiving antenna on the q-th subcarrier. And obtaining the NK angle domain channel matrix. And after taking an absolute value of each angle domain channel matrix, calculating to obtain the fourth-order central moment and the square of the variance of each angle domain channel matrix. And the square ratio of the fourth central moment to the variance is the angular domain peak value of the angular domain channel matrix.

After the fourth-order central moment and the square of the variance of each angle domain channel matrix are obtained, the average angle domain peak value f can be calculated by using the following formula (13)₅：

Wherein, mu_4,nqRepresenting the fourth-order central moment of an angle domain channel matrix of the nth receiving antenna on the qth subcarrier;

representing the square of the variance of the angle domain channel matrix for the nth receive antenna on the qth subcarrier. And (3) calculating to obtain an average value of the peak values of the NK angle domains through a formula (13), namely the peak value of the average angle domain of the channel between the end equipment and the network equipment.

When the terminal device is extracting L eigenvalues from the CSI (for example, including f above)₁、f₂、f₃、f₄、f₅Etc.) then, the first and second substrates are,namely, the following formula (14) is used to normalize the obtained L characteristic values to obtain each normalized characteristic value g_i：

Where i denotes a feature value index, i ═ 1, 2, … …, L. Then, the variance coefficient f₁Normalized value g₁Gain variance information extracted from the CSI for the terminal device. Average gain amplitude f₂Normalized value g₂And gain amplitude information extracted from the CSI for the terminal equipment. Average time domain peak f₃Normalized value g₃And extracting time domain information from the CSI for the terminal equipment. Mean K factor f₄Normalized value g₄And extracting frequency domain information from the CSI for the terminal equipment. Mean angle domain peak value f₅Normalized value g₅And extracting angle domain information from the CSI for the terminal equipment.

Optionally, if the algorithm adopted by the network device cannot support the complex number operation and the operation of the three-dimensional matrix, in order to adapt to the algorithm adopted by the network device, the preprocessing of the terminal device on the CSI may further include Batch Normalization (BN) processing, conversion processing from the three-dimensional matrix to the two-dimensional matrix, separation processing of a real part and an imaginary part, and the like.

The BN processing may be normalization processing performed on the complex gain data in the CSI in b (b > 1) blocks, where a mean value of each block of the complex gain data after normalization is 0 and a variance is 1. The conversion process from the three-dimensional matrix to the two-dimensional matrix may refer to dividing C adjacent receiving antennas of the N receiving antennas into a group based on the number of the receiving antennas, so as to divide the CSI from a three-dimensional complex matrix with a size of M × N × Q into N/C two-dimensional complex matrices with a size of M × Q and a number of channels C (C ≧ 1). That is, the CSI is antenna separated. The real and imaginary part separation processing may mean that the real and imaginary parts of the complex matrix are respectively taken as parameters.

It should be noted that the above listed CSI and the preprocessed CSI are examples of channel information provided in the present application, and the channel information may also be other descriptive channel state information based on actual requirements (for example, limitations of an algorithm actually adopted by a network device). Similarly, the preprocessing of the CSI may also include other processing procedures, which is not limited in this application.

In the embodiment of the application, when the terminal device needs to be switched from the first network to the second network, the channel information of the first network is carried in the request message and sent to the network device to request the network device to determine whether to allow the terminal device to be switched to the second network. And the channel detection of the second network is not needed to detect whether the second network can be accessed, thereby reducing the power consumption of the terminal equipment.

S402, the network equipment determines switching indication information according to the channel information.

The switching indication information is divided into first information and second information, wherein the first information is used for indicating that the second network is allowed to be accessed, and the second information is used for indicating that the second network is not allowed to be accessed.

In one example, the network device may analyze a channel state between the terminal device and the network device according to the channel information sent by the terminal device, and determine whether a current channel state supports the terminal device to access the second network.

For example, when the network device determines that the channel between the terminal device and the network device is an LOS channel or an NLOS channel with higher quality based on the channel information, since the LOS channel and the NLOS channel with higher quality support the terminal device to access the second network, the network device determines that the corresponding handover indication information is the first information. When the network device determines that the channel between the terminal device and the network device is the NLOS channel based on the channel information and the signal quality of the NLOS channel is poor, the network device determines that the corresponding switching indication information is the second information because the NLOS channel with the poor signal quality does not support the terminal device to access the second network.

Or, the channel information sent by the terminal device is channel characteristic information, and the network device may determine, based on a preset channel characteristic value threshold, whether a channel state between the terminal device and the network device meets a requirement for accessing the second network, so as to determine corresponding handover indication information.

For example, the channel information transmitted by the terminal device is the channel characteristic information g₁、g₂And g₅. Then, when the network device determines g₁Less than a predetermined variance coefficient threshold, g₂Greater than a predetermined amplitude threshold and g₅And when the peak value is larger than the preset angle domain peak value threshold, the network equipment determines that the channel state between the terminal equipment and the network equipment meets the requirement of accessing a second network, and determines that the switching indication information is the first information. Otherwise, determining that the channel state between the terminal equipment and the network equipment does not meet the requirement of accessing the second network, and determining that the switching indication information is the second information.

In another example, a trained classifier is preset in the network device, and the network device may process the channel information by using the classifier to obtain the handover indication information.

In the embodiment of the present application, the classifier is trained based on a training sample set, where the training sample set includes a plurality of channel information samples of the first network and switching indication information (which is the first information or the second information) corresponding to each channel information sample. Classifiers include, but are not limited to, Support Vector Machines (SVMs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), bayesian classifiers, decision trees, and the like. The training process of the classifier will be described later, and will not be described here for the moment.

It is understood that the classifiers used by the network devices are different, and the channel information sent by the terminal device through the request information is also different. And because of the different classifiers, after receiving the channel information sent by the terminal device, the network device may directly input the channel information into the classifier for processing, or may pre-process the channel information to obtain input information matching with the input format of the classifier, and then input the input information into the classifier for processing to obtain corresponding switching indication information. Of course, the classifiers used by the network devices are different, and the preprocessing methods used by the network devices are also different.

An exemplary description is given below of a manner in which the SVM model and the CNN model are taken as examples, and the network device processes the channel information by using the classifier to obtain the handover indication information.

For example, the network device employs an SVM model. The SVM model provided in the embodiment of the present application is obtained by training based on a training sample set, where the training sample set includes a plurality of channel characteristic information samples of a first network and switching indication information corresponding to each channel characteristic information sample. Thus, in this example, the input information of the SVM model is channel feature information, including gain variance information, gain magnitude information, time domain information, frequency domain information, and/or angle domain information, etc.

Input vector g formed by five pieces of channel characteristic information as input information₁,g₂,g₃,g₄,g₅Take the example. The channel information sent by the terminal device may be CSI of the first network detected by the terminal device, or may be five pieces of channel characteristic information extracted from the CSI.

If the channel information sent by the terminal device is CSI, the network device first needs to pre-process the CSI when processing the CSI by using the SVM model. Corresponding to the input format of the SVM model, the preprocessing includes extracting channel characteristic information from the CSI, and five channel characteristic information g₁、g₂、g₃、g₄、g₅To obtain an input vector { g }₁,g₂,g₃,g₄,g₅}. The network device then obtains the input vector g₁,g₂,g₃,g₄,g₅And inputting the input vector into the SVM model for processing (namely performing classification calculation on the input vector) to obtain switching indication information.

For example, the handover indication information may be a classification value, for example, when the obtained classification value is 1, the handover indication information is the first information, and when the classification value is 0, the handover indication information is the second information.

If it is finishedThe channel information sent by the end equipment is 5 pieces of channel characteristic information g extracted from CSI types₁、g₂、g₃、g₄、g₅Then the network device directly processes g without preprocessing the CSI₁、g₂、g₃、g₄、g₅Formed input vector g₁,g₂,g₃,g₄,g₅Inputting the data into an SVM model for processing to obtain switching indication information.

It can be understood that, in the embodiment of the present application, the channel characteristic information samples of the first network of the SVM model and the handover indication information corresponding to the channel characteristic information samples are obtained through training. The SVM model can directly obtain the switching indication information whether the second network is available or not from the channel characteristic information of the first network, fully utilizes the correlation between the channel state of the first network and the channel state of the second network, and has robustness on the signal-to-noise ratio of the second network. Thereby reducing the probability of false positives.

As another example, the network device employs a CNN model. The CNN model provided in the embodiment of the present application is obtained by training based on a training sample set, where the training sample set includes a plurality of preprocessed CSI samples of the first network and switching indication information corresponding to each CSI sample. Therefore, in this example, the input information of the CNN model is CSI preprocessed by BN processing, conversion processing of a three-dimensional matrix into a two-dimensional matrix, and separation processing of real and imaginary parts. The channel information sent by the terminal device may be CSI or CSI after being preprocessed.

If the channel information sent by the terminal device is CSI, when the network device uses the CNN model to process CSI, based on the input format of the CNN model, the network device needs to perform BN processing, conversion processing from a three-dimensional matrix to a two-dimensional matrix, and separation processing of a real part and an imaginary part, which are equal to preprocessing operations, on the CSI to obtain input information matching the input format of the CNN model.

For the real part and imaginary part separation process, the network device may use the real part and the imaginary part of the complex matrix as the input channel numbers of the CNN model, respectively. On the one hand, since the CNN model cannot support complex operations, the network device enables the CNN model to process CSI by separating the real part and the imaginary part of the complex matrix. On the other hand, the CNN model can learn classification characteristics from the real part and the imaginary part of the CSI respectively, and obtain the combined characteristics of the real part and the imaginary part, so that the characteristic quantity extracted from the CSI by the CNN model is greatly enriched, and the processing accuracy of the CNN model is improved.

For example, as shown in fig. 5, the network device may perform BN processing on the CSI first, and the CSI after the BN processing is still an M × N × Q three-dimensional complex matrix. And then, carrying out conversion processing from a three-dimensional matrix to a two-dimensional matrix on the CSI processed by BN to obtain N/C two-dimensional complex matrices with the size of M multiplied by Q and the number of input channels of C. And finally, performing real part and imaginary part separation processing on each two-dimensional complex matrix with the size of M multiplied by Q and the number of input channels of C to obtain N/C real number matrixes with the size of M multiplied by Q and the number of input channels of 2C. Then, the N/C real number matrices are M × Q, and the real number matrix with the input channel number of 2C is the input information of the CNN model.

After the network equipment inputs the obtained input information into the CNN model, the CNN model performs convolution calculation on N/C real number matrixes with the size of M multiplied by Q and the number of input channels of 2C to obtain N/C classification values. For example, the N/C classification values may be a number between 0 and 1. And the CNN model determines the switching indication information corresponding to the CSI by averaging the N/C classification values. For example, if the average value is less than 0.5, the CNN model may determine that the handover indication information is 0, i.e., the handover indication information is the second information. If the average value is greater than or equal to 0.5, the CNN model may determine that the handover indication information is 1, i.e., the handover indication information is the first information.

For example, assume M-8, N-2, Q-288, and C-2. After the network device preprocesses the CSI with a size of 8 × 2 × 288, the obtained input information is a real matrix with a size of 8 × 288 and a number of input channels of 4. The size of the input information of the real matrix can be represented by three dimensions (input channel, receive antenna, subcarrier), i.e. the size of the input information of the real matrix can be represented as (4, 8, 288).

Taking the 6-Layer CNN model as an example, the CNN model includes 6 convolutional layers, pooling (pooling) layers, and Fully Connected Layer (full Connected Layer). Where the convolution kernel size of each convolution layer is 3 x 3 and the edges are extended to 1 (i.e. one row 0 is supplemented before the first row and one column 0 is supplemented after the last column of the input 8 x 288 matrix). Since the subcarrier data Q is usually 8-32 times the number M of transmitting antennas under normal conditions, the sizes of the input information in the subcarrier dimension and the transmitting antenna dimension can be equalized by setting the sampling step length of each convolutional layer in the subcarrier dimension and the receiving antenna dimension. For example, the sampling step size of each convolutional layer in the subcarrier dimension is set to 2, the sampling step size of the first 5 convolutional layers in the receive antenna dimension is set to 1, and the sampling step size of the 6 th convolutional layer in the receive antenna dimension is set to 2. In addition, to increase the operation speed and improve the accuracy, BN processing can be performed for each convolution layer.

Then, after the input information is input into the CNN model, the processing flow of the input information by the CNN model may be as shown in fig. 6. The convolutional layer 1 performs convolution processing and BN processing on the input information (4, 8, 288) to obtain output information of size (32, 8, 144). The convolutional layer 2 performs convolution processing, ReLU function processing, and BN processing on the output information (32, 8, 144) of the convolutional layer 1, and obtains output information having a size of (64, 8, 72). The convolutional layer 3 performs convolution processing, ReLU function processing, and BN processing on the output information (64, 8, 72) of the convolutional layer 2, and obtains output information of size (128, 8, 36). The convolutional layer 4 performs convolution processing, ReLU function processing, and BN processing on the output information (128, 8, 36) of the convolutional layer 3 to obtain output information of size (256, 8, 18). The convolutional layer 5 performs convolution processing, ReLU function processing, and BN processing on the output information (256, 8, 18) of the convolutional layer 4 to obtain output information of size (256, 8, 9). The convolutional layer 6 performs convolution processing, ReLU function processing, and BN processing on the output information (256, 8, 9) of the convolutional layer 5 to obtain output information of size (256, 4, 4). The pooling layer performs maximum pooling on the output information (256, 4, 4) of the convolutional layer 6 to obtain output information of size (256, 1, 1). Finally, the output of 256 channels of the output information (256, 1, 1) of the pooling layer is linearly and fully connected by the full connection layer to obtain a value between 0 and 1. By making a hard decision on the value, i.e., if the value is greater than or equal to 0.5, the fully-connected layer outputs 1 (i.e., the first information), and if the value is less than 0.5, the fully-connected layer outputs 0 (i.e., the second information).

S403, the network device sends the switching indication information to the terminal device.

S404, if the switching indication information received by the terminal equipment is first information, switching operation from the first network to the second network is executed; otherwise, no switching operation is performed.

Wherein, the switching operation is a conventional network switching operation. For example, a handover request sent by a terminal device to a first device of network devices over a first network. After the first device forwards the switching request to the second device, the second device determines whether to allow the terminal device to access the second network. After the second device responds to the switching request, the second device allocates network resources for the terminal device and issues the network resources to the terminal device through the first device. The terminal device and the second device establish a connection based on the network resource, and the terminal device is successfully switched to the second network.

In this embodiment of the application, when the terminal device needs to be switched from the first network to the second network, the terminal device only needs to send the measured channel information of the first network to the network device, and the network device generates corresponding switching indication information based on the channel information of the first network to indicate whether the terminal device is allowed to access the second network. The terminal equipment is not required to use a built-in device corresponding to the second network to perform channel detection to detect whether the second network can be accessed, and particularly, the problem of power consumption caused by long-time use of the device corresponding to the second network to perform channel detection under the condition that the second network cannot be accessed is solved, so that the power consumption of the terminal equipment is reduced.

In one application scenario, the network device may not support the second network at the time of receiving the request message of the terminal device. For example, the network device may temporarily fail to provide services of the second network due to the second network being in a maintenance, or deployment, phase.

In this case, as shown in fig. 7, a flowchart of another embodiment of a network handover method provided by the present application may include:

s701, the terminal equipment sends a request message to the network equipment.

S702, the network device detects whether the first network and the second network are co-sited.

For example, the network device may determine the network currently supported by the network device based on the preconfigured network information. The network device may determine that the first network and the second network are not co-sited if the identity of the first network is present but the identity of the second network is not present in the networks currently supported by the network device. The network device may determine that the first network and the second network are co-sited if the identity of the first network and the identity of the second network are present in the currently supported network at the same time.

When the network device determines that the first network and the second network are co-located, the network device may further perform step S703. When the network device determines that the first network and the second network are not co-sited, the network device may further perform step S706.

And S703, determining switching indication information according to the channel information.

S704, the network equipment sends switching indication information to the terminal equipment.

Wherein, when the network device determines that the handover indication information sent to the terminal device is the first information, the terminal device may further perform step S705. When the network device determines and transmits the handover indication information to be the second information, the terminal device may perform step S707.

S705, the terminal device performs a handover operation for handover from the first network to the second network.

S706, the network device sends a failure response message to the terminal device.

When the network device does not support the second network, the network device may directly feed back a failure response message to the terminal device after receiving the request message, and notify that the terminal device cannot respond to the request message or that the network device does not support the second network.

S707, the terminal device does not perform the switching operation.

When the terminal device receives the second information or receives the failure response message, the terminal device may not perform the handover operation. And resends the request message to the network device based on actual needs.

The training process of the classifier is exemplarily described below.

In the implementation of the present application, the training of the classifier includes two stages. The first stage is an initial training stage, i.e. a training stage for training an initial network mode into a classifier with corresponding functions. The second phase is a training phase that periodically updates the training phase, i.e., a training phase that periodically optimizes classifiers with corresponding functions.

For the initial training phase, the network device first needs to obtain a training sample set. The training sample set includes a plurality of channel information samples of the first network and handover indication information (which is the first information or the second information) corresponding to each channel information sample.

The channel information sample of the first network refers to a channel information sample obtained by measuring CSI of a channel between the terminal device and the network device in a case where the terminal device and the network device communicate based on the first network. The switching indication information corresponding to each channel information sample is determined by performing channel sounding on the second network at the position and time of measuring the CSI corresponding to the channel information sample. That is, when the probing result indicates that the second network can be accessed, the handover indication information corresponding to the channel information sample is determined to be the first information. And determining that the switching indication information corresponding to the channel information sample is the second information under the condition that the second network cannot be accessed in the detection result.

In the embodiment of the present application, each channel information sample in the training sample set and the switching indication information corresponding to each channel information sample may be simulation data or measured data. The simulation data may be based on a channel simulation model or channel simulation software to simulate a specific communication environment, and obtain corresponding channel information samples and switching indication information by performing CSI measurement on the first network and channel sounding on the second network. Wherein, the channel simulation model can be WINNER II; the channel simulation software may be a wireless field (WI) software based on a ray tracing (RayTracing) algorithm, and the like, which is not limited in this application.

And after the channel information sample is obtained, inputting the channel information sample into the initial network model to train the initial network model. For example, the network device sequentially inputs each channel information sample in the training sample set into the initial network model for classification calculation, so as to obtain a corresponding classification value (i.e., switching indication information calculated by the initial network model). And evaluating the difference degree between the classification value and the switching indication information corresponding to the channel information sample in the training sample set through a preset loss function. And when the difference does not meet the preset condition, adjusting the model parameters of the initial network model, and inputting the input information of the next channel information sample again for the next round of training. And when the difference degree meets the preset condition, stopping training the initial network model, and determining the initial network model after training as the classifier with the corresponding function.

The loss function may be designed based on actual requirements, for example, as a Binary Cross Entropy (BCE) function, a mean square error function, and the like. The preset conditions may also be designed based on actual requirements, such as a preset loss value threshold. When the difference degree is smaller than or equal to a preset loss value threshold value, the difference degree meets a preset condition; and when the difference degree is greater than the preset loss value threshold value, the difference degree does not meet the preset condition.

It will be appreciated that the training parameters used in the training process may be adaptively configured based on the type of classifier. For example, taking the CNN model as an example, when training the initial network model of the CNN model, the optimization algorithm may adopt an Adaptive momentum Estimation (Adam) gradient descent method. Momentum learning rate beta₁0.9, energy learning rate β₂0.999, initial learning rate alpha0.001, denominator correction term ∈ 10^-8The learning rate attenuation rate is 0.8.

For the update training phase, the network device may periodically update train the classifier.

For example, after the network device determines the switching indication information corresponding to the channel information by using the classifier each time, if the switching indication information is the first information, the network device may monitor a subsequent switching process, and determine whether the terminal device is successfully switched from the first network to the second network. And if the terminal equipment is successfully switched from the first network to the second network, which indicates that the calculation result of the current classifier is correct, the network equipment adds the channel information and the first information as new samples into the training sample set. If the terminal device is not successfully switched from the first network to the second network, which indicates that the calculation result of the current classifier is wrong, the switching indication information corresponding to the channel information should be the second information, and the network device adds the channel information and the second information as new samples into the training sample set.

Based on the updated training sample set, the network device may periodically update the classifier. The process of updating the classifier by the network device is the same as the process of training the initial network model by the network device, and is not described herein again.

It can be understood that the classifier can be further optimized by periodically updating the classifier by adopting the training sample set which is continuously updated based on the actual switching condition, so that the error of the classifier is reduced, and the accuracy of the classifier is improved.

The above-mentioned scheme provided by the present application is mainly introduced from the perspective of interaction between network elements. It is understood that the terminal device and the network device include corresponding hardware structures and/or software modules for performing the respective functions in order to implement the functions. Those of skill in the art would readily appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as hardware or combinations of hardware and computer software. Whether a function is performed as hardware or computer software drives hardware depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

Referring to fig. 8, a network switching apparatus provided by the present application includes a processing unit 801, a communication unit 802, and a storage unit 803. The processing unit 801, the communication unit 802, and the storage unit 803 are connected by a communication bus 804.

The communication unit 802 may be a device having a transceiving function for communicating with other network apparatuses or terminal apparatuses.

The storage unit 803 may include one or more memories, which may be devices in one or more devices or circuits for storing programs or data. The storage unit 803 may stand alone and be connected to the processing unit 801 via a communication bus 804. The storage unit 803 may also be integrated with the processing unit 801.

The network switching apparatus may be used in a network device, a terminal device, a circuit, a hardware component, or a chip.

The network switching device may be a network device or a chip in a network device. A schematic diagram of a network device may be shown in fig. 2. When the network switching apparatus is a network device, the communication unit 802 of the network switching apparatus may include an antenna, a transceiver, and a communication interface of the network device, such as the antenna 203, the transceiver 204, and the communication interface 205 in fig. 2. The processing unit 801 may include a processor of a network device, such as the processor 201 in fig. 2. Optionally, the storage unit 803 includes a memory in the network device, for example, the memory 202 in fig. 2.

When the network switching device is a chip in the network device in the embodiment of the present application. The communication unit 802 may be a communication interface, for example, the communication interface may include an input or output interface, pins or circuits, and the like. Alternatively, the storage unit 803 may store a computer-executable instruction of the method on the network device side, so that the processing unit 801 executes the method executed by the network device in the above-described embodiment.

When the network switching device is a network device or a chip in the network device in the embodiment of the present application, the network switching device may implement the method executed by the network device in the above embodiment.

The communication unit 802 is configured to receive a request message sent by a terminal device, where the request message requests to switch from a first network to a second network, a frequency band of the first network is smaller than a frequency band of the second network, and the request message carries channel information measured by the terminal device, where the channel information is used to describe a channel state of the first network; the processing unit 801 is configured to determine handover indication information according to the channel information, where the handover indication information is first information indicating that access to the second network is allowed or second information indicating that access to the second network is not allowed; the communication unit 802 is further configured to send handover indication information to the terminal device.

Optionally, the processing unit 801 determines the handover indication information according to the channel information, including: processing the channel information by using the trained classifier to obtain switching indication information; the classifier is obtained by training based on a training sample set, wherein the training sample set comprises a plurality of channel information samples of the first network and switching indication information corresponding to each channel information sample.

Optionally, the channel information is channel state information CSI of the first network measured by the terminal device.

Optionally, the processing unit 801 processes the channel information by using a preset classifier to obtain the switching indication information, where the processing information includes: preprocessing the CSI to obtain input information matched with the input format of the classifier; and inputting the input information into a classifier for processing to obtain switching indication information.

Optionally, the channel information is the CSI of the first network after being preprocessed.

Optionally, the preprocessing includes extracting channel characteristic information from the CSI; the channel characteristic information includes gain variance information, gain amplitude information, time domain information, frequency domain information and/or angle domain information.

Optionally, when the switching indication information is the first information, the processing unit 801 is further configured to add the channel information and the first information to the training sample set when the terminal device is successfully switched from the first network to the second network; adding the channel information and the second information to a training sample set when the terminal device is not successfully switched from the first network to the second network; the classifier is periodically updated according to a training sample set.

Optionally, before the processing unit 801 determines the handover indication information according to the channel information, the processing unit 801 is further configured to detect whether the first network and the second network are co-located; the processing unit 801 determines handover indication information according to the channel information, and includes: and when the first network and the second network are co-sited, determining switching indication information according to the channel information.

It should be noted that, when the network switching device is a terminal device or a chip in the terminal device in the embodiment of the present application, for specific implementation manners of each function implemented by the processing unit 801 and the communication unit 802, reference may be made to the relevant description of the steps executed by the network device as described in fig. 4 to 7, which is not described herein again.

The network switching device can be a terminal device or a chip in the terminal device. A schematic diagram of the terminal device may be as shown in fig. 3. When the network switching apparatus is a terminal device, the communication unit 802 of the network switching apparatus may include an antenna, a transceiver and a communication interface of the terminal device, such as the antenna 303 and the transceiver 304 in fig. 3. Processing unit 801 may include a processor of a terminal device, such as processor 301 in fig. 3. Optionally, the storage unit 803 includes a memory in the terminal device, for example, the memory 302 in fig. 3.

When the network switching device is a chip in the terminal device in the embodiment of the present application. The communication unit 802 may be a communication interface, for example, the communication interface may include an input or output interface, pins or circuits, and the like. Alternatively, the storage unit 803 may store computer-executable instructions of the method on the terminal device side, so that the processing unit 801 executes the method executed by the terminal device in the above-described embodiment.

When the network switching device is the terminal device or the chip in the terminal device in the embodiment of the present application, the network switching device may implement the method executed by the terminal device in the above embodiment.

The processing unit 801 is configured to obtain channel information, where the channel information is used to describe a channel state of the first network; the communication unit 802 is configured to send a request message to a network device and receive a handover indication information sent by the network device, where the request message carries channel information, the request information is used to request handover from a first network to a second network, a frequency band of the first network is smaller than a frequency band of the second network, the handover indication information is determined by the network device according to the channel information, and the handover indication information is first information used to indicate that access to the second network is allowed or second information used to indicate that access to the second network is not allowed; the processing unit 801 is further configured to perform a handover operation from the first network to the second network when the handover indication information is the first information.

Optionally, the channel information is channel state information CSI of the first network.

Optionally, the processing unit 801 acquires channel information, including: measuring a channel of a first network to obtain Channel State Information (CSI) of the first network; and preprocessing the CSI to obtain channel information.

It should be noted that, when the network switching device is a terminal device or a chip in the terminal device in the embodiment of the present application, for specific implementation manners of each function implemented by the processing unit 801 and the communication unit 802, reference may be made to the relevant description of steps executed by the terminal device as described in fig. 4 to 7, which is not described herein again.

The embodiment of the application also provides a computer readable storage medium. The methods described in the above embodiments may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. If implemented in software, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media may include computer storage media and communication media, and may include any medium that can communicate a computer program from one place to another. A storage media may be any available media that can be accessed by a computer.

As an alternative design, a computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc, as used herein, includes Compact Disc (CD), laser disc, optical disc, Digital Versatile Disc (DVD), floppy disk and blu-ray disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

The embodiment of the application also provides a computer program product. The methods described in the above embodiments may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. If implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. The procedures or functions described in the above method embodiments are generated in whole or in part when the above computer program instructions are loaded and executed on a computer. The computer may be a general purpose computer, a special purpose computer, a computer network, a network appliance, a user device, or other programmable apparatus.

The above-mentioned embodiments, objects, technical solutions and advantages of the present invention are further described in detail, it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made on the basis of the technical solutions of the present invention should be included in the scope of the present invention.

Claims

1. A network switching method is applied to network equipment and is characterized by comprising the following steps:

acquiring channel characteristic information of a first network, wherein the channel characteristic information comprises gain variance information, gain amplitude information, time domain information, frequency domain information and/or angle domain information;

inputting the channel characteristic information into a trained Support Vector Machine (SVM) model for processing, and outputting to obtain switching indication information, wherein the switching indication information is first information used for indicating that the second network is allowed to be accessed or second information used for indicating that the second network is not allowed to be accessed, and the frequency band of the first network is lower than that of the second network;

and sending the switching indication information to the terminal equipment.

2. The method of claim 1, wherein the obtaining channel characteristic information of the first network comprises:

receiving a request message sent by a terminal device, wherein the request message requests to be switched from a first network to a second network, the request message carries the channel characteristic information, and the channel characteristic information is obtained by preprocessing the measured Channel State Information (CSI) of the first network by the terminal device.

3. The method of claim 1, wherein the obtaining channel characteristic information of the first network comprises:

receiving a request message sent by a terminal device, wherein the request message requests to be switched from a first network to a second network, and the request message carries CSI (channel state information) of the first network measured by the terminal device;

and preprocessing the CSI to obtain the channel characteristic information.

4. The method according to any one of claims 1 to 3, wherein the SVM model is trained based on a training sample set, the training sample set includes a plurality of channel characteristic information samples of the first network and switching indication information corresponding to each channel characteristic information sample, and when the switching indication information output by the SVM model is the first information, the method further includes:

if the terminal equipment is successfully switched from the first network to the second network, adding the channel characteristic information and the first information into the training sample set;

if the terminal device is not successfully switched from the first network to the second network, adding the channel characteristic information and the second information to the training sample set;

periodically updating the SVM model according to the training sample set.

5. The method according to any one of claims 1-4, wherein before inputting the channel characteristic information into the trained SVM model for processing and outputting the obtained switching indication information, the method further comprises:

detecting whether the first network and the second network are co-sited;

the inputting the channel characteristic information into a trained SVM model for processing and outputting to obtain switching indication information comprises:

and when the first network and the second network are co-located, inputting the channel characteristic information into a trained SVM model for processing, and outputting to obtain switching indication information.

6. A network switching method is applied to terminal equipment and is characterized by comprising the following steps:

sending a request message to a network device, where the request message is used to request a handover from the first network to a second network, a frequency band of the first network is smaller than a frequency band of the second network, and the request message carries channel characteristic information of the first network, where the channel characteristic information includes gain variance information, gain amplitude information, time domain information, frequency domain information, and/or angle domain information;

receiving switching indication information sent by the network equipment, wherein the switching indication information is determined by the network equipment according to the channel characteristic information, and the switching indication information is first information used for indicating that the second network is allowed to be accessed or second information used for indicating that the second network is not allowed to be accessed;

and when the switching indication information is the first information, switching operation from the first network to the second network is executed.

7. The method of claim 6, wherein prior to sending the request message to the network device, the method further comprises:

measuring a channel of a first network, and obtaining Channel State Information (CSI) of the first network;

and preprocessing the CSI to obtain the channel characteristic information.

8. A network device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 1 to 5 when executing the computer program.

9. A chip comprising a processor and a communication interface, the processor and the communication interface being coupled, the processor executing a computer program or instructions to implement the network switching method of any of claims 1 to 5, the communication interface being for communicating with a module other than the chip.

10. A terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 6 to 7 when executing the computer program.

11. A chip comprising a processor and a communication interface, the processor and the communication interface being coupled, the processor executing a computer program or instructions to implement the network switching method of any of claims 6 to 7, the communication interface being for communicating with a module other than the chip.