Calculation device and calculation method

Abstract

A calculation device according to one aspect calculates a setting value of a parameter related to handover in a cellular network, the calculation device comprising: a first acquisition unit that uses a simulated environment of the cellular network to acquire simulated environment search information represented by a set of a setting value at which an objective function configured by an index representing a predetermined communication quality is optimized and an objective function value corresponding to the setting value; a second acquisition unit that uses the simulated environment search information and real environment search information represented by a set of the setting value of the parameter and the objective function value when the setting value is set to acquire, as new real environment search information, a set of a setting value at which the communication quality does not become a predetermined allowable value or less and an objective function value when the setting value is set; and a calculation unit that calculates, as the setting value to be set in the cellular network, a setting value at which the objective function value when the setting value is set is evaluated to be the best among setting values included in the real environment search information.

Description

Calculation Device and Calculation Method

[0001] The present disclosure relates to a calculation device and a calculation method.

[0002] Base stations constituting cellular networks such as 4G and 5G have handover-related parameters such as HOM (Handover Margin) and TTT (Time To Trigger) (hereinafter, also referred to as "handover-related parameters").

[0003] As a prior art for calculating handover-related parameters, the technique disclosed in Non-Patent Document 1 is known. Non-Patent Document 1 discloses a technique for calculating handover-related parameters that optimize the value of an objective function composed of indicators representing communication quality using a cellular network simulation device that simulates a cellular network. In the technique disclosed in Non-Patent Document 1, RSRP (Reference Signal Received Power), SINR (Signal to Interference plus Noise Ration), and HOSR (Handover Success Rate) are used as indicators representing communication quality.

[0004] However, in actual operation, it is often the case that handover-related parameters are calculated based on the operator's know-how using an actual cellular network.

[0005] M. Umar Bin Farooq et al., "A Data-Driven Self-Optimization Solution for Inter-Frequency Mobility Parameters in Emerging Networks," in IEEE Transactions on Cognitive Communications and Networking, vol. 8, no. 2, pp. 570-583, June 2022.

[0006] However, since cellular network simulators cannot perfectly simulate actual cellular networks, communication quality may not improve as expected if there is a large discrepancy between the two. Furthermore, when using an actual cellular network, the communication quality of the actual cellular network may fall below a predetermined acceptable level during the process of exploring handover-related parameters.

[0007] This disclosure is made in view of the above points, and aims to provide a technology that can calculate handover-related parameters that improve the communication quality of a cellular network while reducing the risk of the communication quality of the cellular network falling below acceptable levels.

[0008] A calculation device according to one aspect of the present disclosure is a calculation device for calculating setting values ​​of parameters related to handover in a cellular network, comprising: a first acquisition unit that uses an environment that simulates the cellular network to acquire simulated environment search information represented by a set of setting values ​​that optimize an objective function composed of an index representing a predetermined communication quality and an objective function value corresponding to the setting values; a second acquisition unit that uses the simulated environment search information and real environment search information represented by a set of the parameter setting values ​​and an objective function value when the setting values ​​are set in the cellular network to acquire a set of setting values ​​that ensure the communication quality of the cellular network does not fall below a predetermined allowable value and an objective function value when the setting values ​​are set in the cellular network as new real environment search information; and a calculation unit that calculates, among the setting values ​​included in the real environment search information, the setting value that is evaluated as having the best objective function value when the setting values ​​are set in the cellular network, as a setting value to be set in the cellular network.

[0009] It is possible to calculate handover-related parameters that improve cellular network communication quality while reducing the risk of cellular network communication quality falling below acceptable levels.

[0010] This figure shows an example of the overall configuration of the parameter setting system according to this embodiment. This figure shows an example of the hardware configuration of the parameter setting device according to this embodiment. This figure shows an example of the functional configuration of the parameter setting device according to this embodiment. This flowchart shows an example of the parameter setting process according to this embodiment.

[0011] Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings. In the following embodiment, a parameter setting system 1 will be described that can calculate and set handover-related parameters that improve communication quality while reducing the risk that the actual communication quality of the cellular network will fall below a predetermined tolerance value.

[0012] <Example of Overall Configuration of Parameter Setting System 1> Figure 1 is a diagram showing an example of the overall configuration of the parameter setting system 1 according to this embodiment. As shown in Figure 1, the parameter setting system 1 according to this embodiment includes a parameter setting device 10, a cellular network simulation device 20, and a cellular network 30. The parameter setting device 10 and the cellular network simulation device 20 are connected in a way that allows communication via any communication network. Similarly, the parameter setting device 10 and the cellular network 30 (more precisely, the base stations included in the cellular network 30) are connected in a way that allows communication via any communication network.

[0013] The parameter setting device 10 calculates the setting values ​​(hereinafter also referred to as "parameter setting values") for handover-related parameters (e.g., HOM, TTT, etc.), and then sets these parameter setting values ​​in the cellular network 30 (more precisely, in the base stations included in the cellular network 30).

[0014] More specifically, the parameter setting device 10 acquires simulated environment search information, which is a pair of parameter setting values ​​found by the cellular network simulation device 20 and an objective function value representing the communication quality when those parameter setting values ​​are set. Then, the parameter setting device 10 uses this simulated environment search information and the actual environment search information, which is a pair of parameter setting values ​​of the cellular network 30 and their corresponding objective function values, to repeatedly search for new actual environment search information while considering the risk of the communication quality falling below an acceptable level. This makes it possible to calculate parameter setting values ​​that improve communication quality while reducing the risk of the communication quality falling below an acceptable level. Here, the objective function is a function composed of one or more indicators representing communication quality (e.g., RSRP, SINR, HOSR, etc.). For example, the objective function can be composed of a weighted linear sum of one or more indicators representing communication quality.

[0015] The cellular network simulation device 20 provides an environment that simulates a cellular network 30 (hereinafter also referred to as the "simulation environment"). The cellular network simulation device 20 may provide the simulation environment using a known network simulator (e.g., ns-3, Exata, etc.), an arbitrary machine learning model (e.g., XGBoost, RandomForest, etc.), or other programs, etc.

[0016] The cellular network 30 is an actual cellular network composed of multiple base stations. Hereafter, the cellular network 30 will also be referred to as the "actual environment."

[0017] Note that the overall configuration of the parameter setting system 1 shown in Figure 1 is just one example and is not limited to this. For example, a cellular network simulation device 20 may be incorporated into the parameter setting device 10, and the parameter setting device 10 and the cellular network simulation device 20 may be configured as a single unit.

[0018] <Example of Hardware Configuration of Parameter Setting Device 10> Figure 2 is a diagram showing an example of the hardware configuration of the parameter setting device 10 according to this embodiment. As shown in Figure 2, the parameter setting device 10 according to this embodiment includes an input device 101, a display device 102, an external I / F 103, a communication I / F 104, a RAM (Random Access Memory) 105, a ROM (Read Only Memory) 106, an auxiliary storage device 107, and a processor 108. Each of these hardware components is connected to communicate via a bus 109.

[0019] The input device 101 is, for example, a keyboard, mouse, touch panel, or physical button. The display device 102 is, for example, a display or display panel. The parameter setting device 10 does not necessarily have to have at least one of the input device 101 and the display device 102.

[0020] The external I / F 103 is an interface with external devices such as the recording medium 103a. Examples of recording media 103a include CDs (Compact Discs), DVDs (Digital Versatile Disks), SD memory cards (Secure Digital memory cards), and USB (Universal Serial Bus) memory cards.

[0021] The communication interface 104 is an interface for communicating with the cellular network simulation device 20 and with base stations included in the cellular network 30. The RAM 105 is a volatile semiconductor memory (storage device) for temporarily holding programs and data. The ROM 106 is a non-volatile semiconductor memory (storage device) that can retain programs and data even when the power is turned off. The auxiliary storage device 107 is a non-volatile storage device such as an HDD (Hard Disk Drive), SSD (Solid State Drive), or flash memory. The processor 108 is a type of arithmetic unit such as a CPU (Central Processing Unit) or GPU (Graphic Processing Unit).

[0022] Note that the hardware configuration shown in Figure 2 is just one example, and the hardware configuration of the parameter setting device 10 is not limited to this. For example, the parameter setting device 10 may have multiple auxiliary storage devices 107 and multiple processors 108, it may not have some of the hardware shown, or it may have various other hardware besides the hardware shown.

[0023] <Example of Functional Configuration of Parameter Setting Device 10> Figure 3 is a diagram showing an example of the functional configuration of the parameter setting device 10 according to this embodiment. As shown in Figure 3, the parameter setting device 10 according to this embodiment has a simulated environment search unit 110, a search information management unit 111, and a real environment search unit 112. Each of these units is realized, for example, by a process in which one or more programs installed in the parameter setting device 10 are executed by a processor 108 or the like. The parameter setting device 10 according to this embodiment also has an environment search information storage unit 113. The environment search information storage unit 113 is realized, for example, by a storage area such as an auxiliary storage device 107. The environment search information storage unit 113 may also be realized, for example, by a storage area of ​​a storage device (e.g., a storage device provided by a database server) that is communicably connected to the parameter setting device 10.

[0024] The simulated environment search unit 110 acquires the best parameter setting values ​​and their corresponding objective function values ​​as simulated environment search information in the simulated environment provided by the cellular network simulation device 20, using known optimization methods.

[0025] The exploration information management unit 111 manages the simulated environment exploration information acquired by the simulated environment exploration unit 110 and the real environment exploration information acquired by the real environment exploration unit 112. In other words, the exploration information management unit 111 stores the simulated environment exploration information and real environment exploration information in the environment exploration information storage unit 113, and retrieves the simulated environment exploration information and real environment exploration information from the environment exploration information storage unit 113.

[0026] Furthermore, the exploration information management unit 111 uses the simulated environment exploration information and the actual environment exploration information acquired from the environment exploration information storage unit 113 to create dual environment exploration information that includes parameter setting values ​​and objective function values ​​included in the simulated environment exploration information or the actual environment exploration information, as well as identification information of the environment from which they were acquired.

[0027] The real-world exploration unit 112 uses the exploration information created by the exploration information management unit 111 to acquire new real-world exploration information while considering the risk that the communication quality of the cellular network 30 will fall below an acceptable level. That is, the real-world exploration unit 112 uses the exploration information to search for parameter setting values ​​while considering the risk that the communication quality of the cellular network 30 will fall below an acceptable level. Subsequently, the real-world exploration unit 112 acquires the objective function value by setting the parameter setting value in the cellular network 30 (more precisely, in the base stations included in the cellular network 30). As a result, a pair of parameter setting values ​​and their corresponding objective function values ​​is obtained as real-world exploration information.

[0028] The environment exploration information storage unit 113 stores simulated environment exploration information and actual environment exploration information. The environment exploration information storage unit 113 may also store both types of environment exploration information.

[0029] Hereinafter, simulated environment exploration information and real-world environment exploration information will be represented as (x, y), where x is the parameter setting value and y is the objective function value. For example, if HOM is h and TTT is t, the parameter setting value x will be represented as x = (h, t). Also, to distinguish between simulated environment exploration information and real-world environment exploration information, simulated environment exploration information will have an "s" added to the upper right corner (x ｓ , y ｓ ) is represented as, and real-world exploration information is given an "r" in the upper right corner (x ｒ , y ｒ ) shall be used as a notation.

[0030] <Example of parameter setting process> Figure 4 is a flowchart showing an example of the parameter setting process according to this embodiment. Hereinafter, the environment search information storage unit 113 stores the current parameter setting value x of the cellular network 30. １ ｒand the corresponding objective function value y １ ｒ The actual environment search information (x １ ｒ , y １ ｒ ) representing the pair is assumed to be stored.

[0031] The simulation environment search unit 110 obtains a plurality of pieces of simulation environment search information respectively by a known optimization method in the simulation environment provided by the cellular network simulation device 20 (step S101). Note that the optimization method is not limited to a specific method, but for example, particle swarm optimization (Reference 1) or the like can be used.

[0032] Hereinafter, a set of a plurality of pieces of simulation environment search information obtained in step S101 above (hereinafter, also referred to as a "simulation environment search information set") is S = {(x ｉ ｓ , y ｉ ｓ [[ID=二十一]])|i=1,・・・,N}とする。ここで、x ｉ ｓ is the i-th parameter setting value in the simulation environment, and y ｉ ｓ is the objective function value corresponding to the i-th parameter setting value in the simulation environment. N is the total number of pieces of simulation environment search information.

[0033] The search information management unit 111 stores the simulation environment search information set S obtained in step S101 above in the environment search information storage unit 113 (step S102).

[0034] Here, the following steps S103 to step S107 are repeatedly executed until it is determined in step S107 below that a predetermined condition is satisfied. Hereinafter, the repetition of steps S103 to step S107 is referred to as an "iteration", and steps S103 to step S107 in a certain iteration τ will be described. Note that the iteration τ starts from τ = 1 and is updated as τ ← τ + 1 every time steps S103 to step S107 are executed.

[0035] Furthermore, below, the set of real-world exploration information at the start of iteration τ (hereinafter also referred to as the "real-world exploration information set") is given by R = {(x ｉ ｒ , y ｉ ｒ Let i = 1, ..., τ. Here, x ｉ ｒ is the i-th parameter setting value in the actual environment, y ｉ ｒ This is the objective function value corresponding to the i-th parameter setting in the real environment.

[0036] The search information management unit 111 obtains the simulated environment search information set S and the actual environment search information set R from the environment search information storage unit 113, and then creates a set of both environment search information (hereinafter also referred to as the "both environment search information set") from these simulated environment search information set S and the actual environment search information set R (step S103). The search information management unit 111 may create the both environment search information set B, for example, as follows.

[0037] B = {(x ｉ , y ｉ , z ｉ ) | i = 1, ..., N, N+1, ..., N+τ} Here, when i ∈ {1, ..., N}, x ｉ = x ｉ ｓ , y ｉ = y ｉ ｓ , z ｉ = 1, and when i ∈ {N+1, ..., N+τ}, x ｉ = x ｉ－Ｎ ｒ , y ｉ = y ｉ－Ｎ ｒ , z ｉ = 0. Note that z ｉ This is the search information for both environments (x ｉ , y ｉ , z ｉ ) Parameter setting value x ｉ and objective function value y ｉ This is identification information representing the environment from which the data was obtained, z ｉ = 1 when the source environment is a simulated environment, z ｉWhen the value is 0, it indicates that the source environment is the actual environment.

[0038] The real-world environment search unit 112 uses the set of both environment search information B created in step S103 above to set a new parameter setting value x, while considering the risk that the communication quality of the cellular network 30 (real environment) will fall below an acceptable value. τ＋１ ｒ The parameter setting value x is calculated (step S104). The real environment search unit 112 takes the set of both environment search information B and the tolerance value C of the communication quality of the real environment as input and calculates x such that the communication quality of the real environment does not fall below the tolerance value C (i.e., the objective function value does not fall below the tolerance value C) and the objective function value is the best parameter setting value. τ＋１ ｒ You can calculate it as follows.

[0039] The above parameter setting value x τ＋１ ｒ The method for calculating is not limited to a specific method, but the real-world exploration unit 112 can use, for example, SaMSBO (Safe Multi-Source Bayesian Optimization) (Reference 2), which is a type of multitask Bayesian optimization. In SaMSBO, in order to safely perform the main task, data from auxiliary tasks is used in addition to the data from the main task. Specifically, real-world exploration information (x) is used as the data for the main task. ｉ , y ｉ , z ｉ ) (i = N+1, ..., N+τ), simulated environment exploration information (x) as data for the auxiliary task. ｉ , y ｉ , z ｉ ) (i = 1, ..., N) is used. In addition, an acceptable value C is used as a safety threshold to safely perform the search for the main task. Then, according to Algorithm1 described in Reference 2, x ｏｐｔ Set the parameter value to x τ＋１ ｒ This is calculated.

[0040] The allowable value C may be a constant value, for example, or the objective function value y represents the current communication quality of the cellular network 30.１ ｒ Values corresponding thereto may be used, or values given by a user or the like may be used.

[0041] The actual environment exploration unit 112 sets the parameter setting value x calculated in step S104 above τ＋１ ｒ in the cellular network 30 (more precisely, the base station included in the cellular network 30), and thereby, the objective function value y τ＋１ ｒ corresponding to it τ＋１ ｒ is obtained (step S105). That is, the actual environment exploration unit 112 obtains an index representing the communication quality when the parameter setting value x τ＋１ ｒ is set in the cellular network 30, and then calculates the objective function value y τ＋１ ｒ from those indexes. Note that an index representing the communication quality of the cellular network 30 can be obtained from, for example, a server or the like that monitors and manages the communication quality of the cellular network 30.

[0042] The exploration information management unit 111 stores the pair of the parameter setting value x τ＋１ ｒ calculated in step S104 above and the objective function value y τ＋１ ! ｒ obtained in step S105 above as actual environment exploration information (x τ＋１ ｒ , y τ＋１ ｒ ) in the environment exploration information storage unit 113 (step S106). As a result, the set of actual environment exploration information becomes R = {(x ｉ ｒ [[ID=ID=ID=38]], y ｉ ｒ ) | i = 1,..., τ + 1}.

[0043] The actual environment exploration unit 112 determines whether or not a predetermined end condition is satisfied (step S107). Various conditions can be used as the end condition. For example, "the iteration τ has exceeded a predetermined threshold value", "the objective function value y τ＋１ｒ The objective function value y １ ｒ It is possible to use phrases such as "an improvement over the previous version."

[0044] If the termination condition is not met in step S107 above, τ is updated to τ ← τ + 1, and the process returns to step S103 above.

[0045] On the other hand, if it is determined in step S107 that the termination condition is met, the real-world search unit 112 sets the parameter setting value corresponding to the best objective function value among the parameter setting values ​​of the real-world search information included in the real-world search information set R to the cellular network 30 (more precisely, the base station included in the cellular network 30) (step S108). The best objective function value is the maximum objective function value if a larger value indicates better communication quality, and the minimum objective function value if a smaller value indicates better communication quality. As a result, the parameter setting value that best improves communication quality is set to the cellular network 30.

[0046] <Summary> As described above, the parameter setting device 10 according to this embodiment can reduce the risk that the communication quality of the cellular network to which the setting value of the handover-related parameters is set will fall below a predetermined tolerance when calculating the setting value of the handover-related parameters. Therefore, by using the parameter setting device 10 according to this embodiment, it is possible to reduce the situation in which the actual communication quality of the cellular network falls below the tolerance, and it is possible to calculate the setting value of the handover-related parameters that improves communication quality without affecting actual operation.

[0047] The present invention is not limited to the embodiments specifically disclosed above, and various modifications, changes, and combinations with known technologies are possible without departing from the spirit of the claims.

[0048] [References] Reference 1: Kennedy, James, and Russell Eberhart. "Particle swarm optimization." Proceedings of ICNN '95 - International Conference on Neural Networks. Vol. 4. IEEE, 1995. Reference 2: Lubsen, Jannis, Christian Hespe, and Annika Eichler. "Towards safe multi-task Bayesian optimization." 6th Annual Learning for Dynamics & Control Conference. PMLR, 2024.

[0049] 1 Parameter setting system 10 Parameter setting device 20 Cellular network simulation device 30 Cellular network 101 Input device 102 Display device 103 External I / F 103a Recording medium 104 Communication I / F 105 RAM 106 ROM 107 Auxiliary storage device 108 Processor 109 Bus 110 Simulation environment search unit 111 Search information management unit 112 Real environment search unit 113 Environment search information storage unit

Claims

1. A calculation device for calculating setting values ​​for parameters related to handover in a cellular network, comprising: a first acquisition unit that uses an environment simulating the cellular network to acquire simulated environment search information represented by a set of setting values ​​that optimize an objective function composed of an index representing a predetermined communication quality and an objective function value corresponding to the setting values; a second acquisition unit that uses the simulated environment search information and real environment search information represented by a set of the parameter setting values ​​and an objective function value when the setting values ​​are set in the cellular network to acquire a set of setting values ​​that ensure the communication quality of the cellular network does not fall below a predetermined allowable value and an objective function value when the setting values ​​are set in the cellular network as new real environment search information; and a calculation unit that calculates, among the setting values ​​included in the real environment search information, the setting value that is evaluated as having the best objective function value when the setting values ​​are set in the cellular network, as a setting value to be set in the cellular network.

2. The calculation device according to claim 1, wherein the second acquisition unit repeats the acquisition of new real-world exploration information until predetermined conditions are met.

3. The calculation device according to claim 1 or 2, wherein the second acquisition unit acquires the new real-environment exploration information by using SaMSBO (Safe Multi-Source Bayesian Optimization), a type of multitask Bayesian optimization, where the real-environment exploration information is the data of the main task, the simulated-environment exploration information is the data of the auxiliary task, and the tolerance value is a threshold for safely performing the main task's exploration.

4. A calculation method comprising: a computer for calculating setting values ​​for parameters related to handover in a cellular network, which performs the following steps: a first acquisition step of acquiring simulated environment search information represented by a pair of setting values ​​that optimize an objective function composed of an index representing a predetermined communication quality, and an objective function value corresponding to the setting values, using an environment that simulates the cellular network; a second acquisition step of acquiring new real environment search information, which is a pair of setting values ​​that ensure the communication quality of the cellular network does not fall below a predetermined tolerance value, and an objective function value that occurs when the setting values ​​are set in the cellular network, using the simulated environment search information and real environment search information represented by a pair of the parameter setting values ​​and the objective function value that occurs when the setting values ​​are set in the cellular network; and a calculation step of calculating the setting value to be set in the cellular network, which is evaluated as the setting value that best results in the objective function value when the setting values ​​are set in the cellular network, from among the setting values ​​included in the real environment search information.

Images