WO2024066345A1 - 一种ccfd系统中频谱资源的分配方法及网络设备 - Google Patents

一种ccfd系统中频谱资源的分配方法及网络设备 Download PDF

Info

Publication number
WO2024066345A1
WO2024066345A1 PCT/CN2023/091701 CN2023091701W WO2024066345A1 WO 2024066345 A1 WO2024066345 A1 WO 2024066345A1 CN 2023091701 W CN2023091701 W CN 2023091701W WO 2024066345 A1 WO2024066345 A1 WO 2024066345A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
uplink
downlink
spectrum
network
Prior art date
Application number
PCT/CN2023/091701
Other languages
English (en)
French (fr)
Inventor
杨波靖
吴俊�
李刚
王斌
王许旭
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2024066345A1 publication Critical patent/WO2024066345A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W72/00—Local resource management
    • H04W72/04—Wireless resource allocation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W72/00—Local resource management
    • H04W72/04—Wireless resource allocation
    • H04W72/044—Wireless resource allocation based on the type of the allocated resource
    • H04W72/0453—Resources in frequency domain, e.g. a carrier in FDMA
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W72/00—Local resource management
    • H04W72/50—Allocation or scheduling criteria for wireless resources
    • H04W72/53—Allocation or scheduling criteria for wireless resources based on regulatory allocation policies

Definitions

  • the present application relates to the field of communications, and in particular to a method for allocating spectrum resources in a CCFD system and a network device.
  • CCFD Co-time Co-frequency Full Duplex
  • 5G Fifth Generation Mobile Communication Technology
  • TDD Time Division Duplex
  • FDD Frequency Division Duplex
  • the 5G system defines a variety of frame structure combinations, but the frame structures are all fixed and cannot support dynamic allocation and switching. Therefore, the uplink and downlink spectrum efficiency is basically fixed in the application scenario, and there is little room for improvement.
  • CCFD technology no longer requires the spectrum resource allocation method of sites in the region to be consistent. It can be intelligently allocated and dynamically switched according to actual business needs, and can achieve simultaneous uplink and downlink transmission in the time domain and flexibly allocate resources according to business characteristics. Therefore, how to adjust the allocation of spectrum resources in the CCFD system has become a new challenge.
  • the purpose of the embodiments of the present application is to provide a method and network device for allocating spectrum resources in a CCFD system.
  • a method for allocating spectrum resources in a CCFD system which is applied to a network device and includes:
  • target spectrum resources allocated to each user in the service cell are obtained.
  • a network device including:
  • An extraction module configured to extract characteristic data of at least two network indicators related to spectrum resources according to current spectrum resources allocated to each user in the serving cell;
  • An adjustment module configured to determine an optimization target corresponding to the network indicator according to the characteristic data of each network indicator, and adjust the spectrum resources allocated to each user in the serving cell based on the optimization target;
  • the determination module is used to obtain the target spectrum resources allocated to each user in the service cell when at least two of the network indicators meet the corresponding optimization objectives after adjustment and at least two of the network indicators meet the multi-indicator gradient descent conditions.
  • a network device comprising a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the method described in the first aspect.
  • FIG1 is a schematic diagram showing a frame structure comparison between a conventional 5G system and a CCFD system provided by an embodiment of the present application;
  • FIG2 is a schematic diagram of a processing flow of a method for allocating spectrum resources in a CCFD system provided by an embodiment of the present application;
  • FIG3 is a schematic diagram of a processing flow of a method for allocating spectrum resources in a 5G CCFD system provided by an embodiment of the present application;
  • FIG4 is a schematic diagram of the structure of a network device provided by an embodiment of the present application.
  • FIG5 is a schematic diagram of a multi-scenario technical effect in a CCFD system provided by an embodiment of the present application.
  • FIG6 is a schematic diagram of the structure of a network device provided by an embodiment of the present application.
  • the flexible frame structure (Flexible Frame) has the necessary and sufficient conditions for application deployment.
  • the duration of an air interface frame (also called a wireless frame) is fixed at 10s and consists of 10 subframes.
  • a subframe contains several time slots.
  • the number of time slots is related to the subcarrier spacing. Under normal cyclic prefix, the number of time slots corresponding to the subcarrier spacing of 15kHz, 30kHz, 60kHz, 120kHz, and 240kHz is 1, 2, 4, 8, and 16.
  • a time slot is fixed to contain 14 OFDM symbols.
  • the 5G system defines a variety of frame structure combinations in 38.213Table11.1.
  • the frame structure consists of full downlink time slot D, full uplink time slot U, and special time slot S.
  • the ratio of downlink symbols, GP (Guard Period) and uplink symbols in special time slots is flexible and adjustable, and GP can occupy 2 to 4 symbol lengths.
  • the combination of frame structures is fixed and cannot support dynamic switching and allocation. Therefore, the uplink and downlink spectrum efficiency is basically solidified in the application scenario, and there is little room for improvement.
  • the CCFD system no longer requires the spectrum resource allocation method of sites in the region to be consistent.
  • the embodiment of the present application provides a spectrum resource allocation method in a CCFD system, which can achieve optimal spectrum resource allocation and effectively improve the system spectrum efficiency and user perception.
  • the main technical concepts of the embodiment of the present application include: setting at least two network indicators related to spectrum resources, which can be uplink and downlink spectrum efficiency, uplink and downlink user perception, etc.; based on the set network indicators, defining corresponding spectrum efficiency models and user perception models, and extracting non-convex optimization problems of the models;
  • the cooperative game method achieves non-cooperative game equilibrium (also known as Nash equilibrium) in at least two dimensions, such as spectrum efficiency and user perception, to ensure optimal allocation.
  • the non-cooperative game method refers to a type of game in which it is impossible for participants to reach a binding agreement.
  • CCFD systems communication systems based on CCFD technology
  • 5G CCFD systems communication systems based on CCFD technology
  • CCFD evolution technology 5G CCFD systems
  • the technical solution provided in the embodiment of the present application can be executed by a network device in the CCFD system or by software installed in the network device.
  • the network device is a device deployed in a radio access network (RAN) to provide wireless communication functions to terminal devices.
  • the network device can be a base station, and the base station can include various forms of macro base stations, micro base stations, relay stations, access points, etc.
  • the name of the device with base station function may be different.
  • gNB 5G base station
  • the technical solution provided in the embodiment of the present application can be deployed in the baseband unit or other control unit of the network device.
  • the baseband unit (Base Band Unit, BBU) function of a 5G base station (gNB) is reconstructed into two functional entities, a central unit (Centralized Unit, CU) and a distributed unit (Distribute Unit, DU), and accordingly, it can be deployed in the CU or DU of a 5G base station (gNode B, gNB).
  • BBU Base Band Unit
  • gNode B gNode B
  • FIG. 2 is a spectrum resource allocation method in a CCFD system provided by an embodiment of the present application, and is applied to a network device (which may be a base station), and includes the following steps:
  • the service cell refers to the area covered by the base station
  • the user in the service cell refers to the in-service user accessing the cell
  • the base station allocates spectrum resources to each user in the service cell.
  • the user using the uplink spectrum resources is called the uplink user
  • the user using the downlink spectrum resources is called the downlink user.
  • the users in the service cell generally refer to terminal devices, which may include but are not limited to mobile stations (Mobile Station, MS), mobile terminals (Mobile Terminal), mobile phones (Mobile Telephone), user equipment (User Equipment, UE), handsets and portable equipment (portable equipment), vehicles, etc.
  • the terminal equipment can communicate with one or more core networks via a wireless access network.
  • the terminal equipment can be a mobile phone (or so-called "cellular" phone), a computer with wireless communication function, etc.
  • the terminal equipment can also be a portable, pocket-sized, handheld, computer-built-in or vehicle-mounted mobile device.
  • the network index may be uplink and downlink spectrum efficiency, and the spectrum efficiency or the mean square error of the spectrum efficiency of each uplink user in the service cell may be determined as characteristic data of the uplink spectrum efficiency, and the spectrum efficiency or the mean square error of the spectrum efficiency of each downlink user in the service cell may be determined as characteristic data of the downlink spectrum efficiency.
  • the characteristic data may be in the form of an eigenvalue matrix or other data forms.
  • the spectrum efficiency of the uplink user may be determined according to the uplink transmit power of the uplink user in the serving cell, the uplink composite channel matrix allocated by the serving cell to the uplink user, the beamforming factor of the uplink user in the serving cell, the uplink composite channel matrix allocated by the neighboring cell to the uplink user, the beamforming factor of the uplink user in the neighboring cell, the beamforming factors of other users in the neighboring cell, and the transmit power of other users in the neighboring cell;
  • the spectral efficiency of the downlink user can be determined based on the downlink composite channel matrix allocated by the serving cell to the downlink user, the beamforming factor of the downlink user in the serving cell, the downlink composite channel matrix allocated by the neighboring cell to the downlink user, the beamforming factor of the downlink user in the neighboring cell, and the transmission power of other users in the neighboring cell.
  • the network indicator may also be uplink and downlink user perception, and the actual transmission rate of each uplink user in the service cell may be determined as characteristic data of uplink user perception, and the actual transmission rate of each downlink user in the service cell may be determined as characteristic data of downlink user perception.
  • the actual transmission rate of the uplink user can be determined according to the spectrum resources available to the uplink user and the signal to interference plus noise ratio (SINR) on the serving cell spectrum; the actual transmission rate of the downlink user can be determined according to the spectrum resources available to the downlink user and the signal to interference plus noise ratio (SINR) on the serving cell spectrum.
  • SINR signal to interference plus noise ratio
  • S202 Determine an optimization target corresponding to the network indicator according to the characteristic data of each network indicator, and adjust the spectrum resources allocated to each user in the service cell based on the optimization target.
  • the optimization target corresponding to the network indicator may include an optimization range corresponding to the network indicator.
  • the optimization target corresponding to the uplink and downlink spectrum efficiency can be called the first optimization target
  • the first optimization target can include the following: a first optimization range corresponding to the uplink and downlink spectrum efficiency, and the first optimization range can be determined based on a first function used to represent the uplink spectrum efficiency and a second function used to represent the downlink spectrum efficiency, requiring that the first function and the second function have an intersection in the same monotonic direction.
  • the uplink spectrum efficiency is the spectrum efficiency of all uplink users in the serving cell.
  • the downlink spectrum efficiency is the sum of the spectrum efficiency or the mean square error of the spectrum efficiency of all downlink users in the serving cell; accordingly, the first function is a function for summing the spectrum efficiency or the mean square error of the spectrum efficiency of all uplink users in the serving cell; the second function is a function for summing the spectrum efficiency or the mean square error of the spectrum efficiency of all downlink users in the serving cell.
  • the derivatives can be taken in the time domain by calculus methods to obtain the slopes of the two functions respectively within the sampling period; if the two functions (usually linear functions) intersect in the same monotonic direction, an effective variation interval can be determined as the first optimization range.
  • the optimization target corresponding to the uplink and downlink user perceptions may be called a second optimization target, and the second optimization target may include the following: a second optimization range corresponding to the uplink and downlink user perceptions, and the second optimization range may be determined based on a third function used to represent the uplink user perception and a fourth function used to represent the downlink user perception, requiring that the third function and the fourth function have an intersection in the same monotonic direction.
  • the uplink user perception is the sum of the actual transmission rates of all uplink users in the service cell
  • the downlink user perception is the sum of the actual transmission rates of all downlink users in the service cell
  • the third function is a function that sums the actual transmission rates of all uplink users in the service cell
  • the fourth function is a function that sums the actual transmission rates of all downlink users in the service cell.
  • the derivative can be performed in the time domain by calculus method to obtain the slopes corresponding to the two functions within the sampling period; if the two functions (usually linear functions) have an intersection in the same monotonic direction, an effective variation interval can be determined as the second optimization range.
  • the proportion of spectrum resources in the service cell in the process of adjusting the spectrum resources allocated to each user in the service cell based on the optimization target, can be adjusted by adjusting the spectrum resources allocated to each user in the service cell in the direction of approaching the optimization target (the first optimization target, the second optimization target).
  • the adjustment of the proportion of spectrum resources in the service cell may include adjustment in the frequency domain, and may also include adjustment in the time domain and adjustment in the frequency domain.
  • the proportion of uplink and downlink spectrum resources in the service cell may be adjusted in the time domain in units of a scheduling period (TTI) or a larger time granularity, such as an air interface frame.
  • TTI scheduling period
  • a larger time granularity such as an air interface frame.
  • the network indicator after adjustment satisfies the corresponding optimization target may mean that: after the spectrum resources in the serving cell are adjusted, the latest network indicator obtained is within the corresponding optimization range.
  • the uplink and downlink spectrum efficiencies after adjustment satisfying the corresponding first optimization target may mean that the latest uplink spectrum efficiency and the latest downlink spectrum efficiency obtained after the spectrum resources in the serving cell are adjusted are within the corresponding first optimization range;
  • the uplink and downlink user perceptions after adjustment satisfying the corresponding second optimization target may mean that the latest uplink user perception and the latest downlink user perception obtained after the spectrum resources in the serving cell are adjusted are within the corresponding second optimization range.
  • step S201 extracting characteristic data of at least two network indicators related to the spectrum resources based on the current spectrum resources allocated to each user in the service cell.
  • the multi-index gradient descent condition includes: the gradient descent of each model corresponding to at least two of the network indicators converges monotonically or satisfies the Nash equilibrium point.
  • the corresponding model can be called a spectrum efficiency model.
  • the CCFD system is different from the spectrum resource arrangement method of the traditional 5G system.
  • the allocation of uplink and downlink spectrum resources of each base station is no longer required to be strictly consistent, which means that the spectrum efficiency of each base station needs to be converted into the spectrum efficiency set of users in the service cell.
  • the spectrum efficiency of each downlink user i in serving cell n can be determined by the following formula [1]:
  • ⁇ 2 represents the additive white Gaussian noise.
  • the spectrum efficiency of uplink user i in serving cell n can be determined by the following formula [2]:
  • ⁇ 2 represents additive white Gaussian noise.
  • the uplink and downlink spectrum efficiencies are normalized by the mean square error (MSE).
  • MSE mean square error
  • gD (i) represents the downlink spectrum efficiency
  • gU (j) represents the uplink spectrum efficiency
  • i represents the mean square error of spectrum efficiency of downlink user i
  • j represents the mean square error of the spectrum efficiency of uplink user j
  • Kd represents the number of downlink users
  • KU represents the number of uplink users
  • i and j are the i-th downlink user and the j-th uplink user, respectively.
  • the total system resources in serving cell n can be determined by the following formula [5]:
  • Rn total represents the total system resources in the serving cell n. It can be seen that the allocation of uplink and downlink spectrum resources is adjusted in opposite directions in the CCFD system. Therefore, the uplink and downlink spectrum efficiency problem in the serving cell can be normalized to a discrete non-convex optimization problem (problem P0).
  • the uplink and downlink spectrum efficiency in the serving cell in the actual network and time domain is expressed by the following formulas [6] and [7]:
  • ⁇ D (i, t) (Formula [6]) is the second function used to represent the downlink spectrum efficiency
  • ⁇ U (j, t) (Formula [7]) is the first function used to represent the uplink spectrum efficiency
  • t n+1 , t n-1 represent the mean eigenvalues of two sampling periods in the time domain.
  • a linear fitting method is used here, and a matrix inner product method can also be used.
  • the optimization target corresponding to the spectrum efficiency model can be referred to as the first optimization target for the sake of distinction, specifically: the first optimization range corresponding to the uplink and downlink spectrum efficiencies, the first optimization range is determined according to the first function used to represent the uplink spectrum efficiency and the second function used to represent the downlink spectrum efficiency, and the first function and the second function have an intersection in the same monotonic direction. It can be understood that since the spectrum efficiency model is used to solve the uplink and downlink spectrum efficiency problems in the cell, the spectrum efficiency model corresponds to The optimization goal is also the optimization goal corresponding to the uplink and downlink spectrum efficiency (network indicator).
  • the derivative can be performed in the time domain by calculus method to obtain the slopes corresponding to the two functions within the sampling period; if the two functions have an intersection in the same monotonic direction, an effective variation interval can be determined as the first optimization range.
  • the corresponding model may be called a user perception model.
  • the characteristic data of user perception may be represented by the actual transmission rate of the user in the serving cell.
  • Rn ,i represents the actual transmission rate of user i in serving cell n
  • Bn ,i represents the spectrum resources available to user i in serving cell n
  • Xn represents the signal to interference plus noise ratio (SINR) on the spectrum of serving cell n.
  • the signal to interference plus noise ratio (SINR) on the spectrum of serving cell n can be determined by the following formula [9]:
  • Xn represents the signal-to-interference-plus-noise ratio (SINR) on the spectrum of serving cell n
  • pn represents the transmit power of the user in serving cell n
  • pm represents the transmit power of the user in the adjacent cell
  • hn,n represents the channel gain between the serving cell and the user
  • hms represents the channel gain between the user in the adjacent cell and serving cell n
  • ⁇ i ,n represents additive white Gaussian noise.
  • Bn ,i represents the spectrum resources available to user i in service cell n
  • Mmax represents the maximum time-frequency domain resources of the service cell
  • Mcsn ,i represents the modulation mode of user i in service cell n
  • BD represents the resource ratio of the service cell
  • Flown ,i represents the number of transmission layers of user i in service cell n.
  • formulas [8], [9], and [10] are applicable to both uplink and downlink users.
  • User i can be an uplink user in serving cell n or a downlink user in serving cell n.
  • the calculation methods for uplink and downlink user perception are the same, only the user sets are different.
  • the user set corresponding to the uplink user perception is the uplink user
  • the user set corresponding to the downlink user perception is the downlink user.
  • ⁇ (D,i,t) (Formula [11]) is the fourth function used to represent the downlink user perception
  • ⁇ (U,j,t) (Formula [12]) is the third function used to represent the uplink user perception
  • tn+1 , tn-1 represents the mean eigenvalue of two sampling periods in the time domain
  • the linear fitting method is used here, and the matrix inner product method can also be used.
  • problem P1 has non-convex characteristics
  • the optimization target corresponding to the user perception model can be extracted.
  • the second optimization target specifically: the second optimization range corresponding to the uplink and downlink user perceptions
  • the second optimization range is determined according to the third function used to represent the uplink user perception and the fourth function used to represent the downlink user perception
  • the third function and the fourth function have an intersection in the same monotonic direction.
  • the optimization target corresponding to the user perception model is also the optimization target corresponding to the uplink and downlink user perceptions (network indicators).
  • the derivative can be performed in the time domain by calculus method to obtain the slopes corresponding to the two functions within the sampling period; if the two functions have an intersection in the same monotonic direction, an effective variation interval can be determined as the second optimization range.
  • a multi-indicator gradient descent method can be used to determine whether the two network indicators meet the multi-indicator gradient descent conditions.
  • the multi-indicator gradient descent method derives the two problem functions in continuous sampling periods after time domain sampling. If the gradient can converge to 0, the algorithm operation result is confirmed to be optimal, otherwise it needs to be re-iterated.
  • Nash equilibrium also known as non-cooperative game equilibrium, is an important term in game theory, that is, any party unilaterally changes its own strategy under this strategy combination (the strategies of other parties remain unchanged) and will not increase its own benefits. In the embodiment of the present application, it specifically refers to the non-cooperative game equilibrium obtained in the two dimensions of spectrum efficiency and user perception.
  • the initial spectrum resource scheduling state in the initial spectrum resource scheduling state, key indicators such as SINR (Signal to Interference plus Noise Ratio) and MCS (Modulation and Coding Scheme) of each cell’s serving users can be measured during the scheduling period, and fed back to the corresponding serving cell.
  • the initial spectrum resources within each user’s scheduling period are allocated in combination with the service characteristics.
  • the spectrum resource allocation method provided in the embodiment of the present application is adopted to achieve the optimal allocation of spectrum resources.
  • the initial spectrum resource scheduling state refers to a spectrum resource scheduling state in which the uplink and downlink spectrum resource ratio is fixed. For example, in a 5G system, the uplink and downlink spectrum resource ratio can be fixed at 74.28%. That is, before executing step S201, the method further includes:
  • S200 allocating initial spectrum resources to each user in the serving cell according to a signal to interference plus noise ratio (SINR), a modulation and coding strategy (MCS), and service characteristics of each user in the serving cell within a scheduling period TTI.
  • SINR signal to interference plus noise ratio
  • MCS modulation and coding strategy
  • the types of service characteristics may include: uplink and downlink services, unidirectional large-capacity services, low-latency services, high-speed mobile services, etc.
  • the process can return to step S200, that is, allocating initial spectrum resources to each user in the service cell based on the signal-to-interference and noise ratio (SINR), modulation and coding strategy (MCS), and service characteristics of each user in the service cell within the scheduling period TTI.
  • SINR signal-to-interference and noise ratio
  • MCS modulation and coding strategy
  • the spectrum resource allocation method in the CCFD system provided in the embodiment of the present application is applied to network equipment, and can extract characteristic data based on at least two network indicators of each user according to the current spectrum resource allocation of each user in the service cell; extract the optimization target corresponding to each network indicator for the above characteristic data, and adjust the proportion of spectrum resources in the service cell based on the optimization target; if at least two network indicators meet the corresponding optimization target after adjustment, and at least two network indicators meet the multi-indicator gradient descent condition, the target spectrum resources allocated to each user in the service cell can be obtained.
  • the spectrum resources in the service cell are adjusted by the optimization targets corresponding to at least two network indicators, and the multi-indicator gradient descent method is used to select the best and act in the system, so as to achieve the optimal allocation of spectrum resources, thereby effectively improving the system spectrum efficiency and user perception.
  • the following takes the 5G CCFD system as an example to explain in detail the spectrum resource allocation method provided in the embodiment of the present application.
  • the method can be applied to gNB. Assuming that two network indicators, uplink and downlink spectrum efficiency and uplink and downlink user perception, are used, as shown in FIG3, the method includes the following steps:
  • the proportion of uplink and downlink spectrum resources in the initial spectrum resource arrangement state is usually a fixed value, for example, 74.28%. It should be noted that S301 is a system initialization process and is not a necessary step of the spectrum resource allocation method provided in the embodiment of the present application.
  • S302 In the initial spectrum resource scheduling state, measure key indicators of users in the serving cell within a scheduling period (TTI), wherein the key indicators may include SINR, MCS, etc., and feed back to the corresponding serving cell.
  • TTI scheduling period
  • S303 Allocate initial spectrum resources to each user within a scheduling period (TTI) according to the SINR, MCS, and service characteristics of each user in the serving cell.
  • TTI scheduling period
  • the spectrum resources in the scheduling period allocated to users according to SINR, MCS, and service characteristics in the initial spectrum resource arrangement state are referred to as initial spectrum resources.
  • the types of service characteristics may include: uplink and downlink services, unidirectional large-capacity services, low-latency services, high-speed mobile services, etc.
  • S304 extracting an eigenvalue matrix of uplink and downlink spectrum efficiency and uplink and downlink user perception based on each user according to the allocation of spectrum resources of each user in the serving cell.
  • the allocation of spectrum resources for each user in the service cell is the current spectrum resources allocated to each user in the service cell.
  • normalization processing can be performed after multi-point sampling in the time domain to fit the eigenvalue matrix, which can improve targeting and accuracy compared to the extraction method of the traditional 5G system.
  • Non-convex optimization problem refers to a method that directly optimizes non-convex formulas without using relaxation processing.
  • adjusting the proportion of spectrum resources in the service cell is to adjust the spectrum resources allocated to each user in the service cell. Because the CCFD system can support simultaneous co-frequency transmission, there is no need to consider the interference factors of neighboring cells, and the uplink and downlink spectrum resources can be flexibly adjusted according to the business needs in the service cell.
  • S307 Determine whether the uplink and downlink spectrum efficiency and uplink and downlink user perception in the adjusted serving cell meet the corresponding optimization targets. If not, return to S304. If yes, continue. Execute S308.
  • S309 Determine whether the two network indicators meet the multi-indicator gradient descent conditions, that is, whether the multi-indicator gradient descent converges monotonically or meets the Nash equilibrium point. If so, it is considered that the optimal allocation is achieved.
  • the spectrum resources allocated to each user in the service cell at this time are called target spectrum resources. At this point, the spectrum resource allocation process ends. If not, return to execute S302.
  • the spectrum resource allocation method provided in the embodiment of the present application is only a one-time spectrum resource allocation process. In specific implementation, it is necessary to repeat the above-mentioned spectrum resource allocation process. After the spectrum resource allocation process in S309 is completed, it can return to S304 or S302 to continue the next spectrum resource allocation process.
  • the spectrum resource allocation method in the 5G CCFD system measures the key indicators such as SINR and MCS of the serving users of each cell in the scheduling period under the initial spectrum resource arrangement state, and feeds back to the corresponding service cell, so as to allocate the spectrum resources in the scheduling period of each user in combination with the service characteristics.
  • the feature data based on the uplink and downlink spectrum efficiency and uplink and downlink user perception of each user are extracted.
  • multi-point sampling in the time domain can be performed to improve the targeting and accuracy.
  • the corresponding optimization target is extracted for the above-mentioned feature data using the characteristics of the non-convex optimization problem, and the proportion of spectrum resources in the service cell is adjusted in the direction of the target approach. If the uplink and downlink spectrum efficiency and the uplink and downlink user perception meet the corresponding optimization target, the adjusted multi-indicator gradient descent evaluation is performed, otherwise the proportion of spectrum resources in the service cell is adjusted again. Subsequently, it is determined whether the multi-indicator gradient descent conditions are met. If so, the spectrum resource allocation process is terminated. Through the non-cooperative game method, the non-convex optimization problem and resource optimization problem in the two dimensions of spectrum efficiency and user perception are iterated and cycled. The multi-indicator gradient descent method is used to select the best one and act in the system, which can effectively improve the system spectrum efficiency and ensure the optimal uplink and downlink perception of the users in the system.
  • the embodiment of the present application further provides a network device.
  • the network device 400 provided in the embodiment of the present application, as shown in FIG. 4 includes:
  • An extraction module 401 is used to extract characteristic data of at least two network indicators related to spectrum resources according to current spectrum resources allocated to each user in the serving cell;
  • An adjustment module 402 is configured to determine an optimization target corresponding to the network indicator according to the characteristic data of each network indicator, and adjust the spectrum resources allocated to each user in the serving cell based on the optimization target;
  • the determination module 403 is used to obtain the target spectrum resources allocated to each user in the service cell when at least two of the network indicators meet the corresponding optimization objectives after adjustment and at least two of the network indicators meet the multi-indicator gradient descent conditions.
  • the network indicator includes uplink and downlink spectrum efficiency
  • the extraction module 401 may include:
  • a first extraction submodule configured to determine the spectrum efficiency or the mean square error of the spectrum efficiency of each uplink user in the serving cell as characteristic data of uplink spectrum efficiency, wherein the uplink user is a user using uplink spectrum resources;
  • the second extraction submodule is used to determine the spectrum efficiency or the mean square error of the spectrum efficiency of each downlink user in the serving cell as the characteristic data of the downlink spectrum efficiency, wherein the downlink user is a user using the downlink spectrum resources.
  • the first extraction submodule is specifically used to determine the spectrum efficiency of the uplink user according to the uplink transmission power of the uplink user in the serving cell, the uplink composite channel matrix allocated by the serving cell to the uplink user, the beamforming factor of the uplink user in the serving cell, the uplink composite channel matrix allocated by the neighboring cell to the uplink user, the beamforming factor of the uplink user in the neighboring cell, the beamforming factors of other users in the neighboring cell, and the transmission power of other users in the neighboring cell;
  • the second extraction submodule is specifically used to determine the spectral efficiency of the downlink user based on the downlink composite channel matrix allocated by the serving cell to the downlink user, the beamforming factor of the downlink user in the serving cell, the downlink composite channel matrix allocated by the neighboring cell to the downlink user, the beamforming factor of the downlink user in the neighboring cell, and the transmission power of other users in the neighboring cell.
  • the adjustment module 402 may include:
  • the first adjustment submodule is used to determine the first optimization target corresponding to the uplink and downlink spectrum efficiency, including: a first optimization range corresponding to the uplink and downlink spectrum efficiency, the first optimization range is determined according to a first function used to represent the uplink spectrum efficiency and a second function used to represent the downlink spectrum efficiency, the first function and the second function have an intersection in the same monotonic direction, the uplink spectrum efficiency is the sum of the spectrum efficiency or the mean square error of the spectrum efficiency of all uplink users in the service cell, and the downlink spectrum efficiency is the sum of the spectrum efficiency or the mean square error of the spectrum efficiency of all downlink users in the service cell.
  • the network indicator includes uplink and downlink user perception
  • the extraction module 401 may include:
  • a third extraction submodule configured to determine the actual transmission rate of each uplink user in the serving cell as characteristic data of uplink user perception, wherein the uplink user is a user using uplink spectrum resources;
  • the fourth extraction submodule is used to determine the actual transmission rate of each downlink user in the serving cell as characteristic data of downlink user perception, wherein the downlink user is a user using downlink spectrum resources.
  • the third extraction submodule is specifically used to determine the actual transmission rate of the uplink user according to the spectrum resources available to the uplink user and the signal to interference and noise ratio (SINR) on the serving cell spectrum;
  • SINR signal to interference and noise ratio
  • the fourth extraction submodule is specifically configured to determine the actual transmission rate of the downlink user according to the spectrum resources available to the downlink user and the signal to interference plus noise ratio (SINR) on the serving cell spectrum.
  • SINR signal to interference plus noise ratio
  • the adjustment module 402 may include:
  • the second adjustment submodule is used to determine the second optimization target corresponding to the uplink and downlink user perceptions, including: a second optimization range corresponding to the uplink and downlink user perceptions, the second optimization range is determined according to a third function used to represent the uplink user perception and a fourth function used to represent the downlink user perception, the third function and the fourth function have an intersection in the same monotonic direction, the uplink user perception is the sum of the actual transmission rates of all uplink users in the serving cell, and the downlink user perception is the sum of the actual transmission rates of all downlink users in the serving cell.
  • the network device 400 may further include:
  • the control module is used to return to trigger the execution of the extraction module 401 when any of the network indicators after adjustment does not meet the corresponding optimization target.
  • the network device 400 may further include:
  • the initial allocation module is used to allocate initial spectrum resources to each user in the service cell according to the signal to interference plus noise ratio (SINR), modulation and coding strategy (MCS), and service characteristics of each user in the service cell within a scheduling period (TTI) before the extraction module 401 is executed.
  • SINR signal to interference plus noise ratio
  • MCS modulation and coding strategy
  • TTI scheduling period
  • the types of service characteristics include: uplink and downlink services, unidirectional large-capacity services, low-latency services, and high-speed mobile services.
  • control module is also used to return to trigger the execution of the initial allocation module when at least two of the network indicators do not meet the multi-indicator gradient descent conditions; wherein the multi-indicator gradient descent conditions include: the gradient descent of each model corresponding to at least two of the network indicators converges monotonically or meets the Nash equilibrium point.
  • the extraction module 401 is specifically used to sample characteristic data sampling values of the network indicator at multiple points in the time domain for each network indicator and perform normalization processing to obtain characteristic data of the network indicator.
  • the network device provided in the embodiment of the present application can be used as the execution subject of the spectrum resource allocation method in the CCTD system shown in Figure 2, so it can realize the function of the spectrum resource allocation method in the CCTD system implemented in Figure 2. Since the principle is the same, it will not be repeated here.
  • the embodiment of the present application further provides a network device 600, including: a processor 601, a transceiver 602, a memory 603 and a bus interface, wherein:
  • the memory 603 stores programs or instructions that can be run on the processor 601.
  • the various steps of the embodiment of the spectrum resource allocation method in the above-mentioned CCTD system are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
  • the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 601 and memory represented by memory 603.
  • the bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein.
  • the bus interface provides an interface.
  • the transceiver 602 may be a plurality of components, namely, a transmitter and a receiver, providing a unit for communicating with various other devices over a transmission medium.
  • the processor 601 is responsible for managing the bus architecture and general processing, and the memory 603 can store data used by the processor 601 when performing operations.
  • An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored.
  • a program or instruction is stored.
  • each process of the embodiment of the spectrum resource allocation method in the above-mentioned CCTD system is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
  • the processor is the processor in the network device described in the above embodiment.
  • the readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.
  • An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the embodiment of the spectrum resource allocation method in the above-mentioned CCTD system, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.
  • the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.
  • An embodiment of the present application provides a computer program product, which is stored in a storage medium.
  • the program product is executed by at least one processor to implement the various processes of the embodiment of the spectrum resource allocation method in the above-mentioned CCTD system, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.
  • the technical solution of the present application can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM/RAM, a disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.
  • a storage medium such as ROM/RAM, a disk, or an optical disk
  • a terminal which can be a mobile phone, a computer, a server, or a network device, etc.

Abstract

本申请实施例公开了一种CCFD系统中频谱资源的分配方法及网络设备，能够实现频谱资源的最优分配，有效提升系统频谱效率和用户感知度。CCFD系统中频谱资源的分配方法，应用于网络设备，包括：根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据；根据每个所述网络指标的特征数据确定与所述网络指标相对应的优化目标，基于所述优化目标调整为所述服务小区中每个用户分配的频谱资源；在调整后至少两个所述网络指标均满足对应的优化目标，且至少两个所述网络指标满足多指标梯度下降条件的情况下，得到为所述服务小区中每个用户分配的目标频谱资源。

Description

一种CCFD系统中频谱资源的分配方法及网络设备
交叉引用
本发明要求在2022年09月26日提交中国专利局、申请号为202211175347.0、发明名称为“一种CCFD系统中频谱资源的分配方法及网络设备”的中国专利申请的优先权，该申请的全部内容通过引用结合在本发明中。
技术领域
本申请涉及通信领域，尤其涉及一种CCFD系统中频谱资源的分配方法及网络设备。
背景技术
为了缓解随着无线设备指数级的增加，无线频谱资源越来越稀缺的问题，通信领域通常采用CCFD(Co-time Co-frequency Full Duplex，同时同频全双工)技术以提高无线通信链路的频谱效率，CCFD是指无线设备使用相同的时间、相同的频率，同时发射和接收无线信号。5G(5th Generation Mobile Communication Technology，第五代移动通信技术)系统已将包括动态TDD(Time Division Duplex，时分双工)和灵活FDD(Frequency Division Duplex，频分双工)的灵活双工技术作为关键功能。CCFD模式相比传统的HD(Half Duplex，半双工)模式，比如FDD模式、TDD模式，能够提高近一倍的频谱效率(Spectral Efficiency)。
5G系统定义了多种帧结构组合方式，但是帧结构方式都为固定形式，无法支持动态分配和切换，因此上下行频谱效率在应用场景中基本固化，提升空间较小。CCFD技术不再要求区域内站点的频谱资源分配方式保持一致，可以根据实际业务需求进行智能分配以及动态切换，且在时域上可以做到上下行同时传输并根据业务特性灵活分配资源。因此，如何在CCFD系统中调整频谱资源的分配出现了新的挑战。
发明内容
本申请实施例的目的是提供一种CCFD系统中频谱资源的分配方法及网络设备。
为了实现上述目的，本申请实施例采用下述技术方案：
第一方面，提供一种CCFD系统中频谱资源的分配方法，应用于网络设备，包括：
根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据；
根据每个所述网络指标的特征数据确定与所述网络指标相对应的优化目标，基于所述优化目标调整为所述服务小区中每个用户分配的频谱资源；
在调整后至少两个所述网络指标均满足对应的优化目标，且至少两个所述网络指标满足多指标梯度下降条件的情况下，得到为所述服务小区中每个用户分配的目标频谱资源。
第二方面，提供一种网络设备，包括：
提取模块，用于根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据；
调整模块，用于根据每个所述网络指标的特征数据确定与所述网络指标相对应的优化目标，基于所述优化目标调整为所述服务小区中每个用户分配的频谱资源；
确定模块，用于在调整后至少两个所述网络指标均满足对应的优化目标，且至少两个所述网络指标满足多指标梯度下降条件的情况下，得到为所述服务小区中每个用户分配的目标频谱资源。
第三方面，提供一种网络设备，包括处理器，存储器及存储在所述存储器上并可在所述处理器上运行的程序或指令，所述程序或指令被所述处理器执行时实现如第一方面所述的方法。
附图说明
此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：
图1为本申请的一个实施例提供的一种传统5G系统与CCFD系统的帧结构对比示意图；
图2为本申请的一个实施例提供的一种CCFD系统中频谱资源的分配方法的处理流程示意图；
图3为本申请的一个实施例提供的一种5G CCFD系统中频谱资源的分配方法的处理流程示意图；
图4为本申请的一个实施例提供的一种网络设备的结构示意图；
图5为本申请的一个实施例提供的CCFD系统中多场景技术效果示意图；
图6为本申请的一个实施例提供的一种网络设备的结构示意图。
具体实施方式
为使本申请的目的、技术方案和优点更加清楚，下面将结合本申请具体实施例及相应的附图对本申请技术方案进行清楚、完整地描述。显然，所描述的实施例仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都应当属于本文件保护的范围。
本申请的说明书和权利要求书中的术语“第一”、“第二”等是用于区别类似的对象，而不用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便本申请的实施例能够以除了在这里图示或描述的那些以外的顺序实施。此外，说明书以及权利要求中“和/或”表示所连接对象的至少其中之一，字符“/”，一般表示前后关联对象是一种“或”的关系。
由于CCFD系统支持时频域上下行双工的特性，使得灵活帧结构(Flexible Frame)具备应用部署的充分必要条件，传统5G系统与CCFD系统的帧结构对比示意图请参考图1。一个空口帧(也可以称为无线帧)的时长固定为10s，由10个子帧(Subframe)组成，一个子帧包含若干时隙(slot)，时隙数量与子载波间隔有关，在正常循环前缀下，子载波间隔为15kHz、30kHz、60kHz、120kHz、240kHz对应的时隙数量为1、2、4、8、16。一个时隙固定包含14个OFDM符号。5G系统在38.213Table11.1中定义了多种帧结构组合方式，帧结构由全下行时隙D、全上行时隙U和特殊时隙S组成。其中，特殊时隙的下行符号、GP(Guard Period，保护间隔)和上行符号的配比灵活可调，GP可占2～4个符号长度。但是帧结构的组合方式都为固定形式，无法支持动态切换和分配，因此上下行频谱效率在应用场景中基本固化，提升空间较小。CCFD系统不再要求区域内站点的频谱资源分配方式保持一致，可以根据实际业务需求进行智能分配以及动态切换，且在时域上可以做到上下行同时传输并根据业务特性灵活分配资源，频谱效率提升空间非常大。如何在CCFD系统中根据业务特性、频谱效率、用户感知度等维度调整频谱资源的分配出现了新的挑战，需要在多维度中达到最优分配。
本申请实施例提供一种CCFD系统中频谱资源的分配方法，能够实现频谱资源的最优分配，有效提升系统频谱效率和用户感知度。本申请实施例的主要技术构思包括：设定与频谱资源相关的至少两个网络指标，所述网络指标可以为上下行频谱效率、上下行用户感知度等；基于设定的网络指标，定义相应的频谱效率模型、用户感知度模型，提取模型非凸优化问题；通过非 合作博弈方法，在频谱效率、用户感知度等至少两个维度达到非合作博弈均衡(也可以称为纳什均衡)，从而保证最优分配，非合作博弈方法是指一种参与者不可能达成具有约束力的协议的博弈类型。
本申请实施例提供的技术方案，可以应用于基于CCFD技术的通信系统，本申请实施例中统称为CCFD系统，例如：5G CCFD系统，或者其他基于CCFD技术或CCFD演进技术的通信系统。
本申请实施例提供的技术方案，可以由CCFD系统中的网络设备执行或安装在网络设备中的软件执行。本申请实施例中，所述的网络设备是一种部署在无线接入网(Radio Access Network，RAN)中用以向终端设备提供无线通信功能的装置。所述网络设备可以为基站，所述基站可以包括各种形式的宏基站，微基站，中继站，接入点等。在采用不同的无线接入技术的系统中，具有基站功能的设备的名称可能会有所不同，例如在5G网络中，称为5G基站(gNB)，本申请实施例中并不限定。具体的，本申请实施例提供的技术方案，可以部署于所述网络设备的基带单元或其他控制单元中，例如5G基站(gNB)的基带单元(Base Band Unit，BBU)功能被重构为中心单元(Centralized Unit，CU)和分布单元(Distribute Unit，DU)两个功能实体，相应的，可以部署于5G基站(gNode B，gNB)的CU或DU中。本申请实施例提供的技术方案，适用于CCFD系统中多基站多小区的应用场景，可以采用集中式或分布式部署的方式。
以下结合附图，详细说明本申请各实施例提供的技术方案。
请参考图2，为本申请的一个实施例提供的一种CCFD系统中频谱资源的分配方法，应用于网络设备(可以为基站)，包括如下步骤：
S201、根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据。
所述服务小区是指基站覆盖的区域，所述服务小区中的用户是指接入该小区的在服用户，基站为服务小区中每个用户分配频谱资源。本申请实施例中，为了便于区分，将使用上行频谱资源的用户称为上行链路用户，将使用下行频谱资源的用户称为下行链路用户。
可以理解，所述服务小区中的用户通常是指终端设备，可以包括但不限于移动台(Mobile Station，MS)、移动终端(Mobile Terminal)、移动电话(Mobile Telephone)、用户设备(User Equipment，UE)、手机(handset)及便携设备(portable equipment)、车辆(vehicle)等，该终端设备可以经无线接入网与一个或多个核心网进行通信，例如，终端设备可以是移动电话(或称为“蜂窝”电话)、具有无线通信功能的计算机等，终端设备还可以是便携式、袖珍式、手持式、计算机内置的或者车载的移动装置。
在一种可选的实现方式中，所述的网络指标可以为上下行频谱效率，可以将所述服务小区中每个上行链路用户的频谱效率或频谱效率均方差确定为上行频谱效率的特征数据，将所述服务小区中每个下行链路用户的频谱效率或频谱效率均方差确定为下行频谱效率的特征数据。所述特征数据可以采用特征值矩阵的方式，也可以采用其他数据方式。
具体的，所述上行链路用户的频谱效率可以根据所述服务小区中所述上行链路用户的上行发射功率、所述服务小区为所述上行链路用户分配的上行复合信道矩阵、所述服务小区中所述上行链路用户的波束赋形因子、相邻小区为所述上行链路用户分配的上行复合信道矩阵、相邻小区中所述上行链路用户的波束赋形因子、相邻小区中其他用户的波束赋形因子、以及相邻小区中其他用户的发射功率确定；
所述下行链路用户的频谱效率可以根据所述服务小区为所述下行链路用户分配的下行复合信道矩阵、所述服务小区中所述下行链路用户的波束赋形因子、相邻小区为所述下行链路用户分配的下行复合信道矩阵、相邻小区中所述下行链路用户的波束赋形因子、以及相邻小区中其他用户的发射功率确定。
在一种可选的实现方式中，所述的网络指标还可以为上下行用户感知度，可以将所述服务小区中每个上行链路用户的实际传输速率确定为上行用户感知度的特征数据，将所述服务小区中每个下行链路用户的实际传输速率确定为下行用户感知度的特征数据。
具体的，所述上行链路用户的实际传输速率可以根据所述上行链路用户可使用的频谱资源、以及所述服务小区频谱上的信干噪比(Signal to Interference plus Noise Ratio，SINR)确定；所述下行链路用户的实际传输速率可以根据所述下行链路用户可使用的频谱资源、以及所述服务小区频谱上的信干噪比(SINR)确定。
S202、根据每个所述网络指标的特征数据确定与所述网络指标相对应的优化目标，基于所述优化目标调整为服务小区中每个用户分配的频谱资源。
具体实施中，网络指标对应的优化目标可以包括网络指标对应的优化范围。
在一种可选的实现方式中，所述上下行频谱效率对应的优化目标可以称为第一优化目标，第一优化目标可以包括如下内容：所述上下行频谱效率对应的第一优化范围，所述第一优化范围可以根据用于表示上行频谱效率的第一函数和用于表示下行频谱效率的第二函数确定，要求所述第一函数和所述第二函数在同一单调方向上存在交集。
其中，所述上行频谱效率为所述服务小区中所有上行链路用户的频谱效 率或频谱效率均方差之和，所述下行频谱效率为所述服务小区中所有下行链路用户的频谱效率或频谱效率均方差之和；相应的，第一函数是针对服务小区中所有上行链路用户的频谱效率或频谱效率均方差进行求和运算的函数；第二函数是针对服务小区中所有下行链路用户的频谱效率或频谱效率均方差进行求和运算的函数。
示例性的，根据用于表示上行频谱效率的第一函数、以及用于表示下行频谱效率的第二函数，可以通过微积分方法在时域上进行求导，得到采样周期内两个函数各自对应的斜率；如果两个函数(通常为线性函数)在同一单调方向上存在交集，则可以确定一个有效的变化区间作为第一优化范围。
在一种可选的实现方式中，所述上下行用户感知度对应的优化目标可以称为第二优化目标，第二优化目标可以包括如下内容：所述上下行用户感知度对应的第二优化范围，所述第二优化范围可以根据用于表示上行用户感知度的第三函数和用于表示下行用户感知度的第四函数确定，要求所述第三函数和所述第四函数在同一单调方向上存在交集。
其中，所述上行用户感知度为所述服务小区中所有上行链路用户的实际传输速率之和，所述下行用户感知度为所述服务小区中所有下行链路用户的实际传输速率之和；相应的，第三函数是对服务小区中所有上行链路用户的实际传输速率进行求和运算的函数；第四函数是对服务小区中所有下行链路用户的实际传输速率进行求和运算的函数。
示例性的，根据用于表示上行用户感知度的第三函数、以及用于表示下行用户感知度的第四函数，可以通过微积分方法在时域上进行求导，得到采样周期内两个函数各自对应的斜率；如果两个函数(通常为线性函数)在同一单调方向上存在交集，则可以确定一个有效的变化区间作为第二优化范围。
具体实施中，在基于所述优化目标调整为服务小区中每个用户分配的频谱资源的过程中，可以在优化目标(第一优化目标、第二优化目标)逼近方向，通过调整为服务小区中每个用户分配的频谱资源，进行服务小区中频谱资源的占比调整，服务小区中频谱资源的占比调整可以包括频域的调整，也可以包括时域的调整和频域的调整，示例性的，可以在时域上以一个调度周期(TTI)为单位或者更大时间粒度为单位，例如一个空口帧为单位，调整服务小区中上下行频谱资源的占比。
S203、在调整后至少两个所述网络指标均满足对应的优化目标，且至少两个所述网络指标满足多指标梯度下降条件的情况下，得到为所述服务小区中每个用户分配的目标频谱资源。
针对任一网络指标，调整后所述网络指标满足对应的优化目标可以是指：在服务小区中的频谱资源调整之后，得到的最新网络指标在对应的优化范围 内。具体的，调整后所述上下行频谱效率满足对应的第一优化目标可以是指：服务小区中频谱资源调整后得到的最新上行频谱效率、最新下行频谱效率在对应的第一优化范围内；调整后所述上下行用户感知度满足对应的第二优化目标可以是指：服务小区中频谱资源调整后得到的最新上行用户感知度、最新下行用户感知度在对应的第二优化范围内。
在一种可选的实现方式中，在调整后任一所述网络指标不满足对应的优化目标的情况下，可以返回执行步骤S201，即根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据。
具体实施中，所述多指标梯度下降条件包括：至少两个所述网络指标对应的各模型梯度下降单调收敛或满足纳什均衡点。下面，对至少两个所述网络指标对应的各模型进行详细说明。
在所述网络指标为上下行频谱效率的情况下，对应的模型可以称为频谱效率模型。CCFD系统有别于传统5G系统的频谱资源编排方式，每个基站上下行频谱资源的分配不再要求必须严格保持一致，这就意味着每个基站频谱效率需要折算到服务小区中用户的频谱效率集合。
服务小区n中每个下行链路用户i的频谱效率可以通过如下公式[1]确定：
其中，表示服务小区n中下行链路用户i的频谱效率，表示服务小区n为下行链路用户i分配的下行复合信道矩阵，wn,i表示服务小区n中下行链路用户i的波束赋形因子；表示相邻小区m为下行链路用户i分配的下行复合信道矩阵，wm,i表示相邻小区m中下行链路用户i的波束赋形因子，Qm,j表示行链第m个小区中第j个用户的发射功率，δ2表示加性高斯白噪声。
服务小区n中上行链路用户i的频谱效率可以通过如下公式[2]确定：
其中，表示服务小区n中上行链路用户i的频谱效率，qn,i表示服务小区n中上行链路用户i的上行发射功率，表示服务小区n为上行链路用户i分配的上行复合信道矩阵，表示服务小区n中上行链路用户i的波束赋形因子；表示相邻小区m为上行链路用户i分配的上行复合信道矩阵，表示相邻小区m中上行链路用户i的波束赋形因子，表示相邻小区m中其他用户j的波束赋形因子，Qm,j表示行链第m个小区中第j个用户的发射功率，δ2表示加性高斯白噪声。
考虑到时域采样统计以及惩罚大尺度错误，将上下行谱效率通过均方误差(MSE)进行归一化处理，下行频谱效率、上行频谱效率分别通过如下公式[3]、[4]确定：

其中，gD(i)表示下行频谱效率，gU(j)表示上行频谱效率，表示下行链路用户i的频谱效率均方差，表示上行链路用户j的频谱效率均方差，Kd表示下行链路用户数，KU表示上行链路用户数，i、j分别为第i个下行链路用户、第j个上行链路用户。
服务小区n中的系统总资源可以通过如下公式[5]确定：
其中，Rn,total表示服务小区n中的系统总资源，可见，上下行频谱资源的分配在CCFD系统中是相反的两个调整方向，因此可将服务小区中上下行频谱效率问题归一化到离散非凸优化问题(问题P0)，服务小区中上下行频谱效率在实际网络及时域连续表达式通过如下公式[6]、[7]表示：

其中，θD(i,t)(公式[6])是用于表示下行频谱效率的第二函数，θU(j,t)(公式[7])是用于表示上行频谱效率的第一函数，tn+1,tn-1表示时域两个采样周期的均值特征值，此处采用线性拟合方法，也可以采用矩阵内积方式。
频谱效率模型对应的优化目标，为了便于区分，可以称为第一优化目标，具体为：所述上下行频谱效率对应的第一优化范围，所述第一优化范围根据用于表示上行频谱效率的第一函数和用于表示下行频谱效率的第二函数确定，所述第一函数和所述第二函数在同一单调方向上存在交集。可以理解，由于频谱效率模型用于解决小区中上下行频谱效率问题，因此频谱效率模型对应 的优化目标也就是上下行频谱效率(网络指标)对应的优化目标。
示例性的，根据第一函数以及第二函数，可以通过微积分方法在时域上进行求导，得到采样周期内两个函数各自对应的斜率；如果两个函数在同一单调方向上存在交集，则可以确定一个有效的变化区间作为第一优化范围。
在所述网络指标为上下行用户感知度的情况下，对应的模型可以称为用户感知度模型。用户感知度的特征数据可以通过服务小区中用户的实际传输速率表示。
服务小区n中每个用户的实际传输速率可以通过如下公式[8]确定：
Rn,i＝Bn,ilog2(1+Xn)
其中，Rn,i表示服务小区n中用户i的实际传输速率，Bn,i表示服务小区n中用户i可使用的频谱资源，Xn表示服务小区n频谱上的信干噪比(SINR)。
具体的，服务小区n频谱上的信干噪比(SINR)可以通过如下公式[9]确定：
其中，Xn表示服务小区n频谱上的信干噪比(SINR)，pn表示服务小区n用户发射功率、pm表示相邻小区用户发射功率、hn,n表示服务小区与用户信道增益、hms表示相邻小区与服务小区n用户信道增益、ωi,n表示加性高斯白噪声。
考虑到实际传输速率受限于传输质量波动，服务小区n中每个用户可使用的频谱资源可通过如下公式[10]确定，以100M带宽为例：
Bn,i＝Mmax*Mcsn,i*BD*Flown,i
其中，Bn,i表示服务小区n中用户i可使用的频谱资源，Mmax表示服务小区的最大时频域资源，Mcsn,i表示用户i在服务小区n的调制方式，BD表示服务小区的资源占比，Flown,i表示用户i在服务小区n的传输层数。
需要说明的是，公式[8]、[9]、[10]对上下行链路用户均适用，用户i可以为服务小区n中上行链路用户，也可以为服务小区n中下行链路用户。
由此可得，服务小区中下行用户感知度在实际网络及时域连续表达式通过如下公式[11]表示：
类似的，服务小区中上行用户感知度在实际网络及时域连续表达式通过如下公式[12]表示：
需要说明的是，上、下行用户感知度计算方法一致，仅用户集合存在差异，上行用户感知度对应的用户集合为上行链路用户，下行用户感知度对应的用户集合为下行链路用户。τ(D,i,t)(公式[11])是用于表示下行用户感知度的第四函数，ε(U,j,t)(公式[12])是用于表示上行用户感知度的第三函数，tn+1,tn-1表示时域两个采样周期的均值特征值，此处采用线性拟合方法，也可以采用矩阵内积方式。
由于CCFD系统中小区间存在CLI(Cross Link Interference，交叉链路干扰)、SIC(Self Interference Cancellation，自干扰)，SIC是指抵消固定网络节点在相同频率上发射和接收信号时自身所产生的干扰。因此引入干扰约束Q来抑制链路干扰过载，用户感知度模型(问题P1)通过如下公式[13]表示：


可见，问题P1是具有非凸特性的，可以提取该用户感知度模型对应的优化目标，为了便于区分，可以称为第二优化目标，具体为：所述上下行用户感知度对应的第二优化范围，所述第二优化范围根据用于表示上行用户感知度的第三函数和用于表示下行用户感知度的第四函数确定，所述第三函数和所述第四函数在同一单调方向上存在交集。可以理解，由于用户感知度模型用于解决小区中上下行用户感知度问题，因此用户感知度模型对应的优化目标也就是上下行用户感知度(网络指标)对应的优化目标。
示例性的，根据第三函数以及第四函数，可以通过微积分方法在时域上进行求导，得到采样周期内两个函数各自对应的斜率；如果两个函数在同一单调方向上存在交集，则可以确定一个有效的变化区间作为第二优化范围。
可以理解，本申请实施例中，仅是以上下行频谱效率和上下行用户感知度两个网络指标进行说明，具体实施中，还可以采用其他类似的与频谱资源相关的网络指标，技术方案中使用的网络指标的个数也不限于两个，具体不再赘述。
示例性的，针对两个网络指标各自对应的频谱效率模型和用户感知度模型，可以通过多指标梯度下降方法判断两个网络指标是否满足多指标梯度下降条件。多指标梯度下降方法通过在时域采样后针对该两种问题函数在连续采样周期进行求导，如果梯度可以收敛逼近于0，则确认算法运算结果最优，否则需要重新迭代计算。纳什均衡也可称为非合作博弈均衡，是博弈论的一个重要术语，即任何一方在此策略组合下单方面改变自身策略(其他方策略不变)都不会提高自身的收益。本申请实施例中，具体是指在频谱效率、以及用户感知度两个维度得到非合作博弈均衡。
在一种可选的实现方式中，可以在初始频谱资源编排状态下，在调度周期内测量每小区在服用户的SINR(Signal to Interference plus Noise Ratio，信干噪比)、MCS(Modulation and Coding Scheme，调制与编码策略)等关键指标，反馈至对应的服务小区，结合业务特性分配每用户调度周期内的初始频谱资源，在此基础上，采用本申请实施例提供的频谱资源的分配方法，实现频谱资源的最优分配。初始频谱资源编排状态是指上下行频谱资源占比固定的频谱资源编排状态，例如在5G系统中，上下行频谱资源占比可以固定为74.28％。也就是说，在执行步骤S201之前，所述方法还包括：
S200、根据所述服务小区中每个用户在调度周期TTI内的信干噪比(SINR)、调制与编码策略(MCS)、以及业务特性，为所述服务小区中每个用户分配初始频谱资源。
其中，所述业务特性的类型可以包括：上下行业务、单向大容量业务、低时延业务、高速移动业务等。
在一种可选的实现方式中，在至少两个所述网络指标不满足多指标梯度下降条件的情况下，可以返回执行步骤S200，即根据所述服务小区中每个用户在调度周期TTI内的信干噪比(SINR)、调制与编码策略(MCS)、以及业务特性，为所述服务小区中每个用户分配初始频谱资源。
本申请实施例提供的CCFD系统中频谱资源的分配方法，应用于网络设备，能够根据服务小区中每个用户当前的频谱资源分配情况，提取基于每个用户的至少两个网络指标的特征数据；针对上述特征数据抽取各网络指标对应的优化目标，并基于优化目标进行服务小区中频谱资源的占比调整；如果调整后至少两个网络指标均满足对应的优化目标，并且至少两个网络指标满足多指标梯度下降条件，即可得到为服务小区中每个用户分配的目标频谱资源。由于网络指标是与频谱资源相关的，通过至少两个网络指标对应的优化目标进行服务小区中频谱资源的调整，通过多指标梯度下降方法进行择优并在系统中作用，能够实现频谱资源的最优分配，从而有效提升系统频谱效率和用户感知度。
下面以5G CCFD系统为例，详细说明本申请实施例提供的频谱资源的分配方法，该方法可应用于gNB，假设采用上下行频谱效率、以及上下行用户感知度两个网络指标，如图3所示，包括如下步骤：
S301、针对5G CCFD系统中服务小区进行初始频谱资源编排，确定初始频谱资源编排状态下的上下行频谱资源的占比。
在5G CCFD系统中，初始频谱资源编排状态下的上下行频谱资源的占比通常为固定值，例如占比为74.28％。需要说明的是，S301为系统初始化流程，并非本申请实施例提供的频谱资源分配方法的必要步骤。
S302、在初始频谱资源编排状态下，在调度周期(TTI)内测量服务小区中用户的关键指标，所述的关键指标可以包括SINR、MCS等，反馈至对应的服务小区。
S303、根据服务小区中每个用户的SINR、MCS、以及业务特性，分配调度周期(TTI)内每个用户的初始频谱资源。
为了便于区分，本申请实施例中，将在初始频谱资源编排状态下，根据SINR、MCS、以及业务特性为用户分配的调度周期内的频谱资源，称为初始频谱资源。所述的业务特性的类型可以包括：上下行业务、单向大容量业务、低时延业务、高速移动业务等。
S304、根据服务小区中每个用户频谱资源的分配情况，提取基于每个用户的上下行频谱效率、以及上下行用户感知度的特征值矩阵。
可以理解，服务小区中每个用户频谱资源的分配情况即是为服务小区中每个用户分配的当前频谱资源。考虑到网络时效性以及记忆性，可以在时域多点采样后进行归一化处理，从而拟合获得特征值矩阵，相比传统5G系统的提取方式能够提升靶向性和准确性。
S305、针对上下行频谱效率、以及上下行用户感知度的特征值矩阵利用非凸优化问题特性抽取对应的优化目标。
需要说明的是，考虑到网络实效性，在重复迭代本步骤时，需要重新抽取对应的优化目标。非凸优化问题是指不使用松弛处理而直接优化非凸公式的方法。
S306、在优化目标逼近方向进行服务小区中频谱资源的占比调整。
可以理解，进行服务小区中频谱资源的占比调整即是调整为所述服务小区中每个用户分配的频谱资源。因为CCFD系统可支持同时同频传输，所以不需要考虑邻区干扰因素，上下行频谱资源可根据服务小区中业务需求进行灵活调整。
S307、判断调整后服务小区中上下行频谱效率、上下行用户感知度是否均满足对应的优化目标，如果不满足，则返回执行S304，如果满足，则继续 执行S308。
S308、提取调整后两个网络指标的下降梯度。
S309、判断两个网络指标是否满足多指标梯度下降条件，即多指标梯度下降是否单调收敛或满足纳什均衡点，如果满足，则认为达到最优分配，为了便于区分，将此时为服务小区中每个用户分配的频谱资源称为目标频谱资源，至此本次频谱资源分配流程结束；如果不满足，则返回执行S302。
需要说明的是，由于服务小区中用户是实时变化的，本申请实施例提供的频谱资源的分配方法仅是一次频谱资源分配流程，具体实施中，需要重复执行上述频谱资源分配流程，在S309本次频谱资源分配流程结束之后，可以返回S304或S302继续下一次频谱资源分配流程。
本申请实施例提供的5G CCFD系统中频谱资源的分配方法，在初始频谱资源编排状态下，在调度周期内测量每个小区在服用户的SINR、MCS等关键指标，反馈至对应的服务小区，能够结合业务特性分配每个用户调度周期内的频谱资源。根据每个用户频谱资源的分配情况，提取基于每个用户的上下行频谱效率、上下行用户感知度的特征数据，考虑到网络时效性以及记忆性，可以进行时域多点采样，提升靶向性和准确性。针对上述特征数据利用非凸优化问题特性抽取对应的优化目标，并在目标逼近方向进行服务小区中频谱资源的占比调整。如果上下行频谱效率、上下行用户感知度均满足对应的优化目标，则进行调整后多指标梯度下降评估，反之重新进行服务小区中频谱资源的占比调整。随后，判断是否满足多指标梯度下降条件，如果满足，则结束本次频谱资源分配流程。通过非合作博弈方法，将频谱效率以及用户感知度两个维度非凸优化问题和资源最优问题进行迭代循环，通过多指标梯度下降方法进行择优并在系统中作用，能够有效提升系统频谱效率，并保证系统中在服用户的上下行最优感知度。
此外，与上述图2所示的CCTD系统中频谱资源的分配方法相对应地，本申请实施例还提供一种网络设备。本申请实施例提供的一种网络设备400，如图4所示，包括：
提取模块401，用于根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据；
调整模块402，用于根据每个所述网络指标的特征数据确定与所述网络指标相对应的优化目标，基于所述优化目标调整为所述服务小区中每个用户分配的频谱资源；
确定模块403，用于在调整后至少两个所述网络指标均满足对应的优化目标，且至少两个所述网络指标满足多指标梯度下降条件的情况下，得到为所述服务小区中每个用户分配的目标频谱资源。
在一种可选的实现方式中，所述网络指标包括上下行频谱效率；
相应的，所述提取模块401，可以包括：
第一提取子模块，用于将所述服务小区中每个上行链路用户的频谱效率或频谱效率均方差确定为上行频谱效率的特征数据，所述上行链路用户为使用上行频谱资源的用户；
第二提取子模块，用于将所述服务小区中每个下行链路用户的频谱效率或频谱效率均方差确定为下行频谱效率的特征数据，所述下行链路用户为使用下行频谱资源的用户。
具体实施中，所述第一提取子模块，具体用于根据所述服务小区中所述上行链路用户的上行发射功率、所述服务小区为所述上行链路用户分配的上行复合信道矩阵、所述服务小区中所述上行链路用户的波束赋形因子、相邻小区为所述上行链路用户分配的上行复合信道矩阵、相邻小区中所述上行链路用户的波束赋形因子、相邻小区中其他用户的波束赋形因子、以及相邻小区中其他用户的发射功率，确定所述上行链路用户的频谱效率；
所述第二提取子模块，具体用于根据所述服务小区为所述下行链路用户分配的下行复合信道矩阵、所述服务小区中所述下行链路用户的波束赋形因子、相邻小区为所述下行链路用户分配的下行复合信道矩阵、相邻小区中所述下行链路用户的波束赋形因子、以及相邻小区中其他用户的发射功率，确定所述下行链路用户的频谱效率。
相应的，所述调整模块402，可以包括：
第一调整子模块，用于确定所述上下行频谱效率对应的第一优化目标包括：所述上下行频谱效率对应的第一优化范围，所述第一优化范围根据用于表示上行频谱效率的第一函数和用于表示下行频谱效率的第二函数确定，所述第一函数和所述第二函数在同一单调方向上存在交集，所述上行频谱效率为所述服务小区中所有上行链路用户的频谱效率或频谱效率均方差之和，所述下行频谱效率为所述服务小区中所有下行链路用户的频谱效率或频谱效率均方差之和。
在一种可选的实现方式中，所述网络指标包括上下行用户感知度；
相应的，所述提取模块401，可以包括：
第三提取子模块，用于将所述服务小区中每个上行链路用户的实际传输速率确定为上行用户感知度的特征数据，所述上行链路用户为使用上行频谱资源的用户；
第四提取子模块，用于将所述服务小区中每个下行链路用户的实际传输速率确定为下行用户感知度的特征数据，所述下行链路用户为使用下行频谱资源的用户。
具体实施中，第三提取子模块，具体用于根据所述上行链路用户可使用的频谱资源、以及所述服务小区频谱上的信干噪比(SINR)确定所述上行链路用户的实际传输速率；
第四提取子模块，具体用于根据所述下行链路用户可使用的频谱资源、以及所述服务小区频谱上的信干噪比(SINR)确定所述下行链路用户的实际传输速率。
相应的，所述调整模块402，可以包括：
第二调整子模块，用于确定所述上下行用户感知度对应的第二优化目标包括：所述上下行用户感知度对应的第二优化范围，所述第二优化范围根据用于表示上行用户感知度的第三函数和用于表示下行用户感知度的第四函数确定，所述第三函数和所述第四函数在同一单调方向上存在交集，所述上行用户感知度为所述服务小区中所有上行链路用户的实际传输速率之和，所述下行用户感知度为所述服务小区中所有下行链路用户的实际传输速率之和。
在一种可选的实现方式中，所述网络设备400还可以包括：
控制模块，用于在调整后任一所述网络指标不满足对应的优化目标的情况下，返回触发所述提取模块401的执行。
在一种可选的实现方式中，所述网络设备400还可以包括：
初始分配模块，用于在所述提取模块401执行之前，根据所述服务小区中每个用户在调度周期(TTI)内的信干噪比(SINR)、调制与编码策略(MCS)、以及业务特性，为所述服务小区中每个用户分配初始频谱资源。
其中，所述业务特性的类型包括：上下行业务、单向大容量业务、低时延业务、高速移动业务。
在一种可选的实现方式中，所述控制模块，还用于在至少两个所述网络指标不满足多指标梯度下降条件的情况下，返回触发所述初始分配模块的执行；其中，所述多指标梯度下降条件包括：至少两个所述网络指标对应的各模型梯度下降单调收敛或满足纳什均衡点。
在一种可选的实现方式中，所述提取模块401，具体用于针对每个所述网络指标，在时域多点采样所述网络指标的特征数据采样值并进行归一化处理，得到所述网络指标的特征数据。
显然，本申请实施例提供的所述网络设备可以作为上述图2所示的CCTD系统中频谱资源的分配方法的执行主体，因此能够实现CCTD系统中频谱资源的分配方法在图2所实现的功能。由于原理相同，在此不再赘述。
针对本申请实施例提供的技术方案能够达到的技术效果，进行示例性说明。请参考图5，假设当前CCFD系统中存在三个基站，分别为基站1、基站2和基站3，共激活三个服务小区，当前网络中存在若干在服用户且按照 非相干性分布，CCFD系统中各服务小区对应引入场景A(高速移动业务)、场景B(大上行业务)、场景C(低时延业务)。采用本申请实施例提供的技术方案，针对两个网络指标对应的非凸优化问题，通过非合作博弈方法循环迭代逼近纳什均衡点，从而得到频谱效率与用户感知度两个维度的最优效果，多指标下降梯度条件满足后的各服务小区频谱资源占比得到相应的调整。
可选地，如图6所示，本申请实施例还提供一种网络设备600，包括：处理器601、收发机602、存储器603和总线接口，其中：
存储器603上存储有可在所述处理器601上运行的程序或指令，该程序或指令被处理器601执行时实现上述CCTD系统中频谱资源的分配方法实施例的各个步骤，且能达到相同的技术效果，为避免重复，这里不再赘述。
在图6中，总线架构可以包括任意数量的互联的总线和桥，具体由处理器601代表的一个或多个处理器和存储器603代表的存储器的各种电路链接在一起。总线架构还可以将诸如外围设备、稳压器和功率管理电路等之类的各种其他电路链接在一起，这些都是本领域所公知的，因此，本文不再对其进行进一步描述。总线接口提供接口。收发机602可以是多个元件，即包括发送机和接收机，提供用于在传输介质上与各种其他装置通信的单元。
处理器601负责管理总线架构和通常的处理，存储器603可以存储处理器601在执行操作时所使用的数据。
本申请实施例还提供一种可读存储介质，所述可读存储介质上存储有程序或指令，该程序或指令被处理器执行时实现上述CCTD系统中频谱资源的分配方法实施例的各个过程，且能达到相同的技术效果，为避免重复，这里不再赘述。
其中，所述处理器为上述实施例中所述的网络设备中的处理器。所述可读存储介质，包括计算机可读存储介质，如计算机只读存储器ROM、随机存取存储器RAM、磁碟或者光盘等。
本申请实施例另提供了一种芯片，所述芯片包括处理器和通信接口，所述通信接口和所述处理器耦合，所述处理器用于运行程序或指令，实现上述CCTD系统中频谱资源的分配方法实施例的各个过程，且能达到相同的技术效果，为避免重复，这里不再赘述。
应理解，本申请实施例提到的芯片还可以称为系统级芯片、系统芯片、芯片系统或片上系统芯片等。
本申请实施例提供一种计算机程序产品，该程序产品被存储在存储介质中，该程序产品被至少一个处理器执行以实现如上述CCTD系统中频谱资源的分配方法实施例的各个过程，且能达到相同的技术效果，为避免重复，这里不再赘述。
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者装置不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者装置所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、方法、物品或者装置中还存在另外的相同要素。此外，需要指出的是，本申请实施方式中的方法和装置的范围不限按示出或讨论的顺序来执行功能，还可包括根据所涉及的功能按基本同时的方式或按相反的顺序来执行功能，例如，可以按不同于所描述的次序来执行所描述的方法，并且还可以添加、省去、或组合各种步骤。另外，参照某些示例所描述的特征可在其他示例中被组合。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以计算机软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端(可以是手机，计算机，服务器，或者网络设备等)执行本申请各个实施例所述的方法。
上面结合附图对本申请的实施例进行了描述，但是本申请并不局限于上述的具体实施方式，上述的具体实施方式仅仅是示意性的，而不是限制性的，本领域的普通技术人员在本申请的启示下，在不脱离本申请宗旨和权利要求所保护的范围情况下，还可做出很多形式，均属于本申请的保护之内。

Claims (13)

  1. 一种同时同频全双工CCFD系统中频谱资源的分配方法，应用于网络设备，包括：
    根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据；
    根据每个所述网络指标的特征数据确定与所述网络指标相对应的优化目标，基于所述优化目标调整为所述服务小区中每个用户分配的频谱资源；
    在调整后至少两个所述网络指标均满足对应的优化目标，且至少两个所述网络指标满足多指标梯度下降条件的情况下，得到为所述服务小区中每个用户分配的目标频谱资源。
  2. 根据权利要求1所述的方法，其中，所述网络指标包括上下行频谱效率；
    所述根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据，具体包括：
    在所述网络指标为上下行频谱效率的情况下，将所述服务小区中每个上行链路用户的频谱效率或频谱效率均方差确定为上行频谱效率的特征数据，将所述服务小区中每个下行链路用户的频谱效率或频谱效率均方差确定为下行频谱效率的特征数据，所述上行链路用户为使用上行频谱资源的用户，所述下行链路用户为使用下行频谱资源的用户。
  3. 根据权利要求2所述的方法，其中，所述上行链路用户的频谱效率根据所述服务小区中所述上行链路用户的上行发射功率、所述服务小区为所述上行链路用户分配的上行复合信道矩阵、所述服务小区中所述上行链路用户的波束赋形因子、相邻小区为所述上行链路用户分配的上行复合信道矩阵、相邻小区中所述上行链路用户的波束赋形因子、相邻小区中其他用户的波束赋形因子、以及相邻小区中其他用户的发射功率确定；
    所述下行链路用户的频谱效率根据所述服务小区为所述下行链路用户分配的下行复合信道矩阵、所述服务小区中所述下行链路用户的波束赋形因子、相邻小区为所述下行链路用户分配的下行复合信道矩阵、相邻小区中所述下行链路用户的波束赋形因子、以及相邻小区中其他用户的发射功率确定。
  4. 根据权利要求2所述的方法，其中，所述上下行频谱效率对应的第一优化目标包括：所述上下行频谱效率对应的第一优化范围，所述第一优化范围根据用于表示上行频谱效率的第一函数和用于表示下行频谱效率的第二函数确定，所述第一函数和所述第二函数在同一单调方向上存在交集，所述上行频谱效率为所述服务小区中所有上行链路用户的频谱效率或频谱效率均方差之和，所述下行频谱效率为所述服务小区中所有下行链路用户的频谱效率或频谱效率均方差之和。
  5. 根据权利要求1所述的方法，其中，所述网络指标包括上下行用户感知度；
    所述根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据，具体包括：
    在所述网络指标为上下行用户感知度的情况下，将所述服务小区中每个上行链路用户的实际传输速率确定为上行用户感知度的特征数据，将所述服务小区中每个下行链路用户的实际传输速率确定为下行用户感知度的特征数据，所述上行链路用户为使用上行频谱资源的用户，所述下行链路用户为使用下行频谱资源的用户。
  6. 根据权利要求5所述的方法，其中，所述上行链路用户的实际传输速率根据所述上行链路用户可使用的频谱资源、以及所述服务小区频谱上的信干噪比SINR确定；所述下行链路用户的实际传输速率根据所述下行链路用户可使用的频谱资源、以及所述服务小区频谱上的信干噪比SINR确定。
  7. 根据权利要求5所述的方法，其中，所述上下行用户感知度对应的第二优化目标包括：所述上下行用户感知度对应的第二优化范围，所述第二优 化范围根据用于表示上行用户感知度的第三函数和用于表示下行用户感知度的第四函数确定，所述第三函数和所述第四函数在同一单调方向上存在交集，所述上行用户感知度为所述服务小区中所有上行链路用户的实际传输速率之和，所述下行用户感知度为所述服务小区中所有下行链路用户的实际传输速率之和。
  8. 根据权利要求1所述的方法，其中，所述方法还包括：
    在调整后任一所述网络指标不满足对应的优化目标的情况下，返回执行所述根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据的步骤。
  9. 根据权利要求1所述的方法，其中，所述根据为服务小区中每个用户分配的当前频谱资源，提取与频谱资源相关的至少两个网络指标的特征数据之前，所述方法还包括：
    根据所述服务小区中每个用户在调度周期TTI内的信干噪比SINR、调制与编码策略MCS、以及业务特性，为所述服务小区中每个用户分配初始频谱资源。
  10. 根据权利要求9所述的方法，其中，所述业务特性的类型包括：上下行业务、单向大容量业务、低时延业务、高速移动业务。
  11. 根据权利要求9所述的方法，其中，所述方法还包括：
    在至少两个所述网络指标不满足多指标梯度下降条件的情况下，返回执行所述根据所述服务小区中每个用户在调度周期TTI内的信干噪比SINR、调制与编码策略MCS、以及业务特性，为所述服务小区中每个用户分配初始频谱资源的步骤；其中，所述多指标梯度下降条件包括：至少两个所述网络指标对应的各模型梯度下降单调收敛或满足纳什均衡点。
  12. 根据权利要求1所述的方法，其中，所述提取与频谱资源相关的至少两个网络指标的特征数据，具体包括：
    针对每个所述网络指标，在时域多点采样所述网络指标的特征数据采样 值并进行归一化处理，得到所述网络指标的特征数据。
  13. 一种网络设备，包括处理器，存储器及存储在所述存储器上并可在所述处理器上运行的程序或指令，所述程序或指令被所述处理器执行时实现如权利要求1至12中任一项所述的方法。
PCT/CN2023/091701 2022-09-26 2023-04-28 一种ccfd系统中频谱资源的分配方法及网络设备 WO2024066345A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211175347.0 2022-09-26
CN202211175347.0A CN117835421A (zh) 2022-09-26 2022-09-26 一种ccfd系统中频谱资源的分配方法及网络设备

Publications (1)

Publication Number Publication Date
WO2024066345A1 true WO2024066345A1 (zh) 2024-04-04

Family

ID=90475848

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/091701 WO2024066345A1 (zh) 2022-09-26 2023-04-28 一种ccfd系统中频谱资源的分配方法及网络设备

Country Status (2)

Country Link
CN (1) CN117835421A (zh)
WO (1) WO2024066345A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108880734A (zh) * 2018-04-28 2018-11-23 哈尔滨工程大学 量子回溯搜索优化的CCFD-Massive MIMO系统功率分配方法
CN113038616A (zh) * 2021-03-16 2021-06-25 电子科技大学 一种基于联邦学习的频谱资源管理分配方法
WO2022094612A1 (en) * 2020-10-30 2022-05-05 Battelle Energy Alliance, Llc Systems, devices, and methods for scheduling spectrum for spectrum sharing
CN114928385A (zh) * 2022-03-31 2022-08-19 西安电子科技大学 一种基于同时同频全双工d2d通信的频谱效率提升方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108880734A (zh) * 2018-04-28 2018-11-23 哈尔滨工程大学 量子回溯搜索优化的CCFD-Massive MIMO系统功率分配方法
WO2022094612A1 (en) * 2020-10-30 2022-05-05 Battelle Energy Alliance, Llc Systems, devices, and methods for scheduling spectrum for spectrum sharing
CN113038616A (zh) * 2021-03-16 2021-06-25 电子科技大学 一种基于联邦学习的频谱资源管理分配方法
CN114928385A (zh) * 2022-03-31 2022-08-19 西安电子科技大学 一种基于同时同频全双工d2d通信的频谱效率提升方法

Also Published As

Publication number Publication date
CN117835421A (zh) 2024-04-05

Similar Documents

Publication Publication Date Title
Bao et al. Joint rate control and power allocation for non-orthogonal multiple access systems
US8477679B2 (en) Resource allocation method and device for amplify-and-forward relay network
CN103797725A (zh) 在无线系统中利用同调性区域的系统及方法
Zhang et al. Resource allocation in D2D-based V2V communication for maximizing the number of concurrent transmissions
CN110536468A (zh) 无线通信系统、通信单元，以及用于调度的方法
CN103260258A (zh) 一种蜂窝终端直通系统中资源分配和资源复用联合方法
CN101841496B (zh) 多输入多输出系统中用于多小区协作通信的方法及装置
EP3920434A1 (en) Communication method and device
CN107852703A (zh) 配置信息获取的方法和装置
CN103369568B (zh) Lte-a中继系统中基于博弈论的无线资源优化方法
CN104168574B (zh) 一种混合蜂窝系统中基于干扰适变选择的上行传输方法
Wan et al. User pairing strategy: A novel scheme for non-orthogonal multiple access systems
Cheng et al. Heterogeneous statistical QoS provisioning for full-duplex D2D communications over 5G wireless networks
Ding et al. What is the true value of dynamic TDD: A MAC layer perspective
CN102685903B (zh) 一种ofdma 系统中基于部分信道信息的资源分配方法
WO2024066345A1 (zh) 一种ccfd系统中频谱资源的分配方法及网络设备
Naparstek et al. Distributed energy efficient channel allocation
CN103237309A (zh) 用于lte-a中继系统干扰协调的准动态频率资源划分方法
CN107172574B (zh) 一种d2d用户对与蜂窝用户共享频谱的功率分配方法
CN107148078B (zh) 一种混合全双工半双工网络的用户接入控制方法与装置
Ferreira et al. Power and delay optimization based uplink resource allocation for wireless networks with device-to-device communications
CN110944378B (zh) 5g移动通信场景下d2d通信的noma功率分配方法
CN104429131A (zh) 选择无线接入网络的方法和装置
KR102102060B1 (ko) 무허가 다중 접속을 위한 스케줄링 방법 및 이를 위한 사용자 단말
CN106231683A (zh) 基于小区网络的d2d通信中的机会干扰管理方案