Intelligent channel allocation method based on imitation learning

By adopting the intelligent channel allocation method combined with FFR and IFR in the communication network, the cellular network capacity is optimized, and the problem of spectrum resource management in high-speed mobile communication scenarios is solved, efficient resource allocation and noise immunity are achieved, and system delay is reduced and improve user satisfaction.

CN119071930APending Publication Date: 2024-12-03蒋世嘉
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411239867.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-05
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively manage spectrum resources in high-speed mobile communication scenarios, resulting in network link congestion and increased delay, especially under the conditions of high frequency utilization and high capacity requirements, traditional orthogonal frequency division multiplexing technology resists The noise capability is weak and there is model error, making it difficult to meet the needs of complex communication scenarios.

Method used

Using an intelligent channel allocation method based on a combination of fractional frequency multiplexing (FFR) and integer frequency multiplexing (IFR), the cellular network capacity is optimized through MATLAB simulation and specific algorithms (such as SWF, TX-ITL, FLG), and the central area and Edge zone, reduce interference and reasonably allocate frequency channel resources.

Benefits of technology

It realizes that while preventing user data leakage, channel and power resources are adaptively allocated, system delay, user satisfaction is improved, and synchronous interference is eliminated when link multiplexing downlink frequency resources, and transmission rate is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119071930A_ABST
    Figure CN119071930A_ABST
Patent Text Reader

Abstract

The invention discloses an intelligent channel allocation method based on imitation learning, which discusses the most suitable channel allocation method in different communication environments from four dimensions, and provides the optimal allocation method of several channel allocation methods in different communication conditions through comparison of variable control modes. In addition, the simulation model is also suitable for actual communication conditions. Under the condition of high frequency resource demand, searching for a more reasonable method to save resources is of great significance, and the method hopes that a more optimized method can be designed to solve the difficulties of resource shortage and network congestion in the future.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to an intelligent channel allocation method based on imitation learning. Background Art

[0002] The air interface and related waveforms of future communication systems must support various high-speed application scenarios. However, the Orthogonal Frequency Division Multiplexing (OFDM) technology widely used in the fourth and fifth generation mobile communication systems faces serious subcarrier interference in high-speed mobile scenarios. In order to provide stable and reliable communication services in high-mobility scenarios, the 6G system can consider using the Orthogonal Time Frequency Space (OTFS) in the delay Doppler domain as its modulation scheme. However, with the rapid development of mobile Internet, the spectrum resources of traditional cellular wireless communication systems can no longer meet the current communication needs. The computing and communication tasks of a large number of computationally intensive and delay-sensitive mobile applications have caused an explosive growth in mobile data traffic in the network, which has led to a heavy burden on the backhaul link between the base station (BS) and the core network. Studies have confirmed that users' requests for content on the Internet are often concentrated, among which requests for 20% of popular content account for 80% of network traffic. The resulting repeated and intensive computing and communication put a great burden on the backhaul link between the base station and the core network. Since available frequency resources are limited, a more reasonable channel allocation method is needed to make full use of available frequency resources.

[0003] In order to meet the user's requirements for high frequency utilization and high capacity, Orthogonal Frequency Division Multiplexing (OFDM) technology is usually used to divide all available frequencies into multiple frequency channels, which can greatly overcome multipath fading. In a fast time-varying multipath channel environment, channel estimation of Orthogonal Frequency-Division Multiplexing (OFDM) systems has always been a challenging problem. Although the traditional oversampling complex exponential basis expansion model (OCE-BEM) reduces the inter-carrier interference (ICI) caused by Doppler frequency shift to a certain extent, its anti-noise ability is weak and there are inherent model errors, which greatly limits its application in complex communication scenarios. Although the above algorithm is adaptive, making its results more accurate and optimized. However, in the actual model, they only regard the entire communication cell as a whole, while in the actual transmission model, the cell is usually divided into two parts, which can reduce interference and improve frequency channel utilization.

[0004] In the process of implementing the above system optimization operations, a series of decision-making problems are faced. Reinforcement learning has some advantages in solving decision-making problems in communication networks, but it still has limitations. Due to the complexity of communication scenarios and the uncertainty of user service requests, it is currently necessary to propose a solution that can optimize the allocation of network cache resources while satisfying the user's willingness to provide resources. Technically, it can solve the problems of high-dimensional state space and complex discrete action space that are difficult to solve with existing technologies, and achieve the goal of minimizing system latency. Summary of the invention

[0005] The present invention mainly aims at the link congestion problem caused by redundant content transmission in the network, and provides an intelligent channel allocation method based on fractional frequency reuse (FFR) and integer frequency reuse (IFR).

[0006] The technical solution of the present invention is to provide an intelligent channel allocation method based on fractional frequency reuse (FFR) and integer frequency reuse (IFR) with the following steps:

[0007] Step 1: Initialize the communication scenario;

[0008] Step 2: Calculate the cellular network capacity by using MATLAB simulation;

[0009] Step 3: Calculate the distance between each cell;

[0010] Step 4: Analyze the fractional frequency multiplexing data and divide the central area and the edge area;

[0011] Step 5: Establish a channel allocation method using integer frequency reuse method, SWF algorithm, transmit interference temperature limit (TX-ITL) algorithm, and forward-looking game (FLG) algorithm;

[0012] Step 6: Calculate the total capacity of the network.

[0013] Preferably, the step 2 is specifically:

[0014] The cellular network capacity is calculated using MATLAB simulation (see Figure 1 ), in order to construct a cellular network, it is necessary to determine the row and column numbers of the network, which can be set according to the specific situation. For example, this is the following 24-unit network with 4 rows and 6 columns.

[0015] Based on the row number, column number, and cell number, a program can be created to generate the coordinates of each cell according to the following formula (1):

[0016] (m, n): Nc=(M-1)×C+N (1)

[0017] Where Nc represents the number of the focused cell, C represents the column number just set for the network, and (m, n) represents the coordinates of the Nc cells. According to this formula, the coordinates of each cell in the cellular network can be defined.

[0018] Preferably, the step 3 is specifically:

[0019] If the radius of the hexagon is set to R, the length from the center of the cell to one side can be easily calculated as The distance between each cell can be calculated using the vertical distance and horizontal distance according to the Pythagorean theorem. Therefore, we only need to focus on calculating the vertical distance and horizontal distance. As shown in the figure, the calculation rules for the vertical distance between different rows are the same, while the calculation rules for the horizontal distance are different. If we want to find the distance between cells (M1, N1) and (M2, N2), when M1 and M2 are both odd or both even, the horizontal distance between them is exactly the difference between the two column numbers multiplied by the horizontal distance between two adjacent cells:

[0020] (N1-N2)×√3R(2)

[0021] However, when M1 is odd and M2 is even or M1 is even and M2 is odd, the adjustment of the horizontal distance is different. There is a gap of (√3r) / 2 in the horizontal direction between the two rows. Therefore, when calculating the distance, the network needs to be divided into odd rows and even rows.

[0022] When M1 and M2 are both odd or even, the distance between them can be calculated by the following formula:

[0023]

[0024] where D is specific to the distance between cells (M1, N1) and (M2, N2).

[0025] Because in the construction of the network, the even rows are moved to the right by (√3R) / 2 from the odd rows. Therefore, when calculating the distance between the odd row cells and the even row cells, it is necessary to divide it into two cases. For example, set M1 to an odd number and M2 to an even number. Therefore, the cell (M1, N1) is on the odd row and the cell (M2, N2) is on the even row. When N1 is smaller than N2, the distance between them can be calculated by the following formula:

[0026]

[0027] When N1 is smaller than N2, the distance between them can be calculated by the following formula:

[0028]

[0029] Equation (4) can also be applied to the case where M1 is an even number and M2 is an odd number and N1 is greater than N2. In addition, equation (5) can also be applied to the case where M1 is an odd number and M2 is an even number and N1 is less than N2. This is because there is always a (√3r) / 2 horizontal gap between the same columns of odd and even rows.

[0030] Preferably, step 4 is specifically as follows: Fractional frequency reuse (FFR) is a method for saving frequency channels. Although it can reduce interference by using the IFR3 method, it also increases the waste of frequency channels. Users are accustomed to using all frequency channels in a cell, but the available frequency channels are reduced by one third in the IFR3 method. Therefore, the FFR method can make an appropriate balance between interference and available frequency channels. Therefore, setting the same frequency channel in the center of the cell can save frequency resources and reduce interference.

[0031] Preferably, the step 5 is specifically:

[0032] 1) IFR3 method

[0033] Integer Frequency Reuse (IFR) is a frequency reuse method. The IFR3 method combines three cells into a cluster. These three cells share all frequency channels on average. This method increases the distance between cells using the same frequency, thereby reducing interference from the same frequency. Through regularity analysis, the coordinates of the same reuse factor are found.

[0034] 2) SWF algorithm

[0035] Water injection (SWF) is a power allocation algorithm that allocates power to different frequency channels based on the channel performance and water level. The water level is a threshold for judging channel performance. When the noise and interference are much lower than the water level, more power will be allocated to the channel. The power Pn allocated to the channel depends on the formula:

[0036] p n =(wc n ) + (6)

[0037] W represents the water level, and Cn represents the noise and interference normalized by the channel gain in channel n. Cn can be obtained by the following formula:

[0038]

[0039] Where σ2 represents the noise power, In represents the interference power in channel n, |H n | 2 represents the channel n gain. The symbol is defined as:

[0040] (x) + =max{0,x} (8)

[0041] Since the total power transmitted is known, the water level can be defined by the following formula:

[0042]

[0043] P represents the total power transmitted, Pi represents the power on channel i, and N represents the total number of channels in the network.

[0044] 3) TX-ITL algorithm

[0045] The transmit interference temperature limit (TX-ITL) algorithm sets a limit Sk for the performance of the channel based on the SWF algorithm. In the TX-ITL algorithm, the power allocation of channel n is based on the formula:

[0046]

[0047] This means that when Cn is above the limit, channel n will not be used due to having poor performance.

[0048] 4) FLG algorithm

[0049] The forward-looking game (FLG) algorithm is similar to the SWF algorithm, especially in the FLG algorithm, the current power allocation is based on the power allocation in the previous channel, and the current power allocation on channel n is updated by the following formula:

[0050]

[0051] Where Pn* represents the previous power allocation on channel n, and φn can be calculated as follows:

[0052]

[0053] Cn* represents the previous noise and interference normalized by the channel gain in channel n. According to the FLG algorithm, the performance of the channel can be predicted by the previous performance, so that power can be allocated to the channel with better performance.

[0054] Preferably, the step 6 is specifically:

[0055] Because there is fading when the signal is transmitted, the matrix ρm,n is set to represent the power loss when the signal is transmitted from cell m to cell n. The power loss depends on the transmission distance and the communication environment. When making comparisons, the environment is always considered to be the same, so that the ρm,n matrix depends only on the transmission distance.

[0056] Assume that ρm,n is calculated according to the following formula:

[0057]

[0058] D(m,n) represents the distance between cell m and cell n. When making comparisons, all other identical parameters can be considered to be 1. The capacity of cell n can be calculated according to Shannon's formula:

[0059] C n =B n ×lb(1+SINR n ) (14)

[0060] Where B represents the frequency bandwidth and SINR represents the signal to interference plus noise radio in each cell. The SINR in cell n can be calculated by the following formula:

[0061]

[0062] Where Pn represents the transmission power from the base station in cell n, Pm represents the transmission power from the base station in cell m, σ2 represents the noise power, and H(m,n) represents the same frequency channel in cell m and cell n, which means that there will be same frequency interference. In this formula, the numerator is the useful transmission power, and the denominator is the combination of the same frequency interference and noise. The capacity of each cell can be obtained by formulas (14) and (15), and the total capacity of the network can be calculated by combining the capacities of all cells in the network.

[0063] Beneficial Effects

[0064] By using deep reinforcement learning algorithms, the base station-side server can adaptively allocate channel and power resources, deploy cache resources, and guide the establishment of user communication links. This centralized approach can continuously and efficiently achieve global control of the network and system performance optimization while preventing user data leakage. It can not only ensure the normal communication of cellular users, but also minimize the average system delay while meeting the user's personal wishes, thereby maximizing user satisfaction. Using this transmission scheme, when the link reuses downlink frequency resources, the co-frequency interference caused by the reuse can be eliminated, and the reused transmission rate within the system can be improved compared to systems that do not adopt this scheme.

[0065] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 The calculation of the cellular network capacity of the present invention is simulated by using MATLAB;

[0067] Figure 2 It is a schematic diagram of a flow chart of calculating the distance between each cell of the present invention;

[0068] Figure 3 It is a schematic diagram of the IFR3 method in the cellular network of the present invention.

[0069] Figure 4 It is a schematic diagram of cellular network capacity (3×10 network capacity).

[0070] Figure 5 It is a schematic diagram of cellular network capacity (5×6 network capacity).

[0071] Figure 6-8 Schematic diagram of the busyness of the indicator channel when the number of users is 3, 7 and 15 respectively

[0072] Figure 9-10 Schematic diagram of network capacity when the cell radius is 1 and the radius is 10 respectively

[0073] Fig.11 This is a diagram of the network capacity when the power budget at the cell center (Pc) is set to 1 and the power budget at the cell edge (Pe) is set to 2.

[0074] Fig.12 This is a diagram of the network capacity when Pc is set to 2 and Pe is set to 1. DETAILED DESCRIPTION

[0075] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0076] The following is a further description of the intelligent channel allocation method based on fractional frequency reuse (FFR) and integer frequency reuse (IFR) of the present invention in conjunction with the accompanying drawings and specific embodiments:

[0077] Provided is an intelligent channel allocation method based on fractional frequency reuse (FFR) and integer frequency reuse (IFR) having the following steps: comprising the following steps,

[0078] Step 1: Initialize the communication scenario;

[0079] Step 2: Calculate the cellular network capacity by using MATLAB simulation;

[0080] The cellular network capacity is calculated using MATLAB simulation (see Figure 1 ), in order to construct a cellular network, it is necessary to determine the row and column numbers of the network, which can be set according to the specific situation, such as Figure 1 As shown, this network is a 24-unit network with 4 rows and 6 columns.

[0081] In a cellular network, the structure of the network will also affect the capacity of the network. Taking a 30-cell network as an example, it can be composed of three rows and ten columns or five rows and six columns. In this embodiment, MATLAB is used to simulate these two structures and calculate their capacities.

[0082] Based on the row number, column number, and cell number, a program can be created to generate the coordinates of each cell according to the following formula (1):

[0083] (m, n): Nc=(M-1)×C+N (1)

[0084] Where Nc represents the number of the focused cell, C represents the column number just set for the network, and (m, n) represents the coordinates of the Nc cells. According to this formula, the coordinates of each cell in the cellular network can be defined.

[0085] according to Figure 4 , 5As shown, in all methods, there is a small difference between the capacities of the two networks, with the 3×10 network capacity being slightly higher than the 5×6 network. This is because in the 5×6 network, the cells are more concentrated, which will have more interference from other cells near them, and the opposite is true for the 3×10 network.

[0086] This suggests that in a network with a certain number of cells, the network should be constructed more centrally to achieve higher capacity.

[0087] Step 3: Calculate the distance between each cell;

[0088] If the radius of the hexagon is set to R, the length from the center of the cell to one side can be easily calculated as The distance between each cell can be calculated using the vertical distance and horizontal distance according to the Pythagorean theorem. Therefore, we only need to focus on calculating the vertical distance and horizontal distance. As shown in the figure, the calculation rules for the vertical distance between different rows are the same, while the calculation rules for the horizontal distance are different. If we want to find the distance between cells (M1, N1) and (M2, N2), when M1 and M2 are both odd or both even, the horizontal distance between them is exactly the difference between the two column numbers multiplied by the horizontal distance between two adjacent cells:

[0089] (N1-N2)×√3R(2)

[0090] However, when M1 is odd and M2 is even or M1 is even and M2 is odd, the adjustment of the horizontal distance is different. There is a gap of (√3r) / 2 in the horizontal direction between the two rows. Therefore, when calculating the distance, the network needs to be divided into odd rows and even rows.

[0091] When M1 and M2 are both odd or even, the distance between them can be calculated by the following formula:

[0092]

[0093] where D is specific to the distance between cells (M1, N1) and (M2, N2).

[0094] Because in the construction of the network, the even rows are moved to the right by (√3R) / 2 from the odd rows. Therefore, when calculating the distance between the odd row cells and the even row cells, it is necessary to divide it into two cases. For example, set M1 to an odd number and M2 to an even number. Therefore, the cell (M1, N1) is on the odd row and the cell (M2, N2) is on the even row. When N1 is smaller than N2, the distance between them can be calculated by the following formula:

[0095]

[0096] When N1 is smaller than N2, the distance between them can be calculated by the following formula:

[0097]

[0098] Equation (4) can also be applied to the case where M1 is an even number and M2 is an odd number and N1 is greater than N2. In addition, equation (5) can also be applied to the case where M1 is an odd number and M2 is an even number and N1 is less than N2. This is because there is always a (√3r) / 2 horizontal gap between the same columns of odd and even rows.

[0099] The specific distance calculation process is as follows: Figure 2 shown.

[0100] Step 4: Fractional frequency reuse data analysis, dividing the central area and the edge area; Fractional frequency reuse (FFR) is a method of saving frequency channels. Although it can reduce interference by using the IFR3 method, it also increases the waste of frequency channels. Users are accustomed to using all frequency channels in the cell, but the available frequency channels are reduced by one third in the IFR3 method. Therefore, the FFR method can make an appropriate balance between interference and available frequency channels. Therefore, setting the same frequency channel in the center of the cell can save frequency resources and reduce interference.

[0101] Step 5: Establish a channel allocation method using integer frequency reuse method, SWF algorithm, transmit interference temperature limit (TX-ITL) algorithm, and forward-looking game (FLG) algorithm;

[0102] 1) IFR3 method

[0103] Integer Frequency Reuse (IFR) is a frequency reuse method. The IFR3 method combines three cells into a cluster. These three cells share all frequency channels on average. This method increases the distance between cells using the same frequency, thereby reducing interference from the same frequency. Through regularity analysis, the coordinates of the same reuse factor are found. Figure 3 As shown, there is a cycle of "gray, yellow, blue" cells in odd rows and a cycle of "blue, gray, yellow" cells in even rows. According to the rule, the coordinates of the same reuse factor can be found.

[0104] 2) SWF algorithm

[0105] Water injection (SWF) is a power allocation algorithm that allocates power to different frequency channels based on the channel performance and water level. The water level is a threshold for judging channel performance. When the noise and interference are much lower than the water level, more power will be allocated to the channel. The power Pn allocated to the channel depends on the formula:

[0106] p n =(wcn ) + (6)

[0107] W represents the water level, and Cn represents the noise and interference normalized by the channel gain in channel n. Cn can be obtained by the following formula:

[0108]

[0109] Where σ2 represents the noise power, In represents the interference power in channel n, |H n | 2 represents the channel n gain. The symbol is defined as:

[0110] (x) + =max{0,x} (8)

[0111] Since the total power transmitted is known, the water level can be defined by the following formula:

[0112]

[0113] P represents the total power transmitted, Pi represents the power on channel i, and N represents the total number of channels in the network.

[0114] 3) TX-ITL algorithm

[0115] The transmit interference temperature limit (TX-ITL) algorithm sets a limit Sk for the performance of the channel based on the SWF algorithm. In the TX-ITL algorithm, the power allocation of channel n is based on the formula:

[0116]

[0117] This means that when Cn is above the limit, channel n will not be used due to having poor performance.

[0118] 4) FLG algorithm

[0119] The forward-looking game (FLG) algorithm is similar to the SWF algorithm, especially in the FLG algorithm, the current power allocation is based on the power allocation in the previous channel, and the current power allocation on channel n is updated by the following formula:

[0120]

[0121] Where Pn* represents the previous power allocation on channel n, and φn can be calculated as follows:

[0122]

[0123] Cn* represents the previous noise and interference normalized by the channel gain in channel n. According to the FLG algorithm, the performance of the channel can be predicted by the previous performance, so that power can be allocated to the channel with better performance.

[0124] Step 6: Calculate the total capacity of the network.

[0125] Because there is fading when the signal is transmitted, the matrix ρm,n is set to represent the power loss when the signal is transmitted from cell m to cell n. The power loss depends on the transmission distance and the communication environment. When making comparisons, the environment is always considered to be the same, so that the ρm,n matrix depends only on the transmission distance.

[0126] Assume that ρm,n is calculated according to the following formula:

[0127]

[0128] D(m,n) represents the distance between cell m and cell n. When making comparisons, all other identical parameters can be considered to be 1. The capacity of cell n can be calculated according to Shannon's formula:

[0129] C n =B n ×lb(1+SINR n ) (14)

[0130] Where B represents the frequency bandwidth and SINR represents the signal to interference plus noise radio in each cell. The SINR in cell n can be calculated by the following formula:

[0131]

[0132] Where Pn represents the transmission power from the base station in cell n, Pm represents the transmission power from the base station in cell m, σ2 represents the noise power, and H(m,n) represents the same frequency channel in cell m and cell n, which means that there will be same frequency interference. In this formula, the numerator is the useful transmission power, and the denominator is the combination of the same frequency interference and noise. The capacity of each cell can be obtained by formulas (14) and (15), and the total capacity of the network can be calculated by combining the capacities of all cells in the network.

[0133] The simulation results show that the number of users indicates the busyness of the channel. Because all power allocation methods are combined with the FFR frequency allocation method in the simulation, the average number of cell center frequency channels and cell edge frequency channels needs to be set separately in the simulation. For example, the number of cell center frequency channels is set to 5, the number of cell edge frequency channels is set to 10, and users are randomly allocated in the cell center and cell edge. The comparison when the average number of users is 3, 7, and 15 is shown in Figure 2. Figure 6-8 As shown in the figure. When the number of users is small, the capacity of IFR3 is much higher than other methods, while the capacity of SWF and TX-ITL algorithms is quite similar. However, as the number of users increases, the capacity of SWF, TX-ITL and FLG increases a lot, while the capacity of IFR3 and IFR1 increases very little. When the average number of usage is equal to 15 of the total number of available channels, the capacity of SWF and TX-ITL is much higher than that of IFR3, but the capacity of FLG does not increase significantly. The results show that higher capacity can be achieved in busy networks using SWF and TX-ITL algorithms.

[0134] The cell radius, which indicates the cell size, can also affect the network capacity. For example, the comparison between the radius of 1 and the radius of 10 shows the result. Fig. 9 , 10 As shown below. The capacity of all methods increases with the increase of radius. In addition, the capacity of the IFR1 method network increases the most because the increase of cell radius means the distance between each cell increases, therefore, the interference from other cells decreases, and IFR1 is the method most affected by radius because it is affected by all cells around them, therefore, when the cell radius increases, the interference decreases much more than other methods.

[0135] The allocation of power budget also affects the capacity of the network because the FFR method divides the frequency channels into two parts: the cell center channel and the cell edge channel. For example, assuming that the number of channels in the cell center and the cell edge is 5, the power budget of the cell center (Pc) is set to 1 and the power budget of the cell edge (Pe) is set to 2. Fig.11 The result when Pc is set to 2 and Pe is set to 1 is as follows Fig.12 The results show that when more power is allocated in the cell center, the capacity of the network will be higher, which means that the channel in the cell center has better performance than the channel at the cell edge because the channel in the cell center is farther away from the interference channel, so more power should be allocated in the cell center.

[0136] In order to obtain higher capacity under limited frequency resources in a specific communication environment, it is necessary to carefully select the channel allocation method. The appropriate channel allocation method can achieve better performance by using fewer resources. This paper discusses the most suitable channel allocation method under different communication environments from four dimensions. By comparing the way of controlling variables, the optimal allocation method of several channel allocation methods under different communication conditions is given. In addition, the simulation model is also suitable for actual communication situations. In the case of high demand for frequency resources, it is of great significance to find more reasonable methods to save resources. I hope that more optimized methods can be designed in the future to solve the difficulties of resource shortage and network congestion.

[0137] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An intelligent channel allocation method based on fractional frequency reuse (FFR) and integer frequency reuse (IFR), characterized in that The method comprises the following steps, Step 1: Initialize the communication scenario; Step 2: Calculate the cellular network capacity by using MATLAB simulation; Step 3: Calculate the distance between each cell; Step 4: Analyze the fractional frequency multiplexing data and divide the central area and the edge area; Step 5: Establish a channel allocation method using integer frequency reuse method, SWF algorithm, transmit interference temperature limit (TX-ITL) algorithm, and forward-looking game (FLG) algorithm; Step 6: Calculate the total capacity of the network.

2. The method according to claim 1, wherein step 2 specifically comprises: The calculation of cellular network capacity is done by MATLAB simulation. In order to construct the cellular network, the row and column numbers of the network need to be determined. This can be set according to the specific situation. For example, this is a 24-cell network with 4 rows and 6 columns as follows; Based on the row number, column number, and cell number, a program can be made to generate the coordinates (m, n) of each cell according to the following formula (1): Nc = (M-1) × C + N (1) Where Nc represents the number of the focused cell, C represents the column number just set for the network, and (m, n) represents the coordinates of Nc cells; according to this formula, the coordinates of each cell in the cellular network can be defined.

3. The method according to claim 1, wherein step 3 specifically comprises: If the radius of the hexagon is set to R, the length from the center of the cell to one side can be easily calculated as The distance between each cell can be calculated using the vertical distance and the horizontal distance according to the Pythagorean theorem; therefore, we only need to focus on calculating the vertical distance and the horizontal distance; as shown in the figure, the calculation rules for the vertical distance between different rows are the same, while the calculation rules for the horizontal distance are different; if we want to find the distance between cells (M1, N1) and (M2, N2), when M1 and M2 are both odd or both even, the horizontal distance between them is exactly the difference between the two column numbers multiplied by the horizontal distance between two adjacent cells: (N1-N2)×√3R(2) However, when M1 is odd and M2 is even or M1 is even and M2 is odd, the adjustment of the horizontal distance is different; there is a gap of (√3r) / 2 in the horizontal direction between the two rows; therefore, when calculating the distance, the network needs to be divided into odd rows and even rows; When M1 and M2 are both odd or even, the distance between them can be calculated by the following formula: where D is specific to the distance between cells (M1, N1) and (M2, N2); Because in the construction of the network, the even rows are moved to the right from the odd rows by (√3R) / 2; therefore, when calculating the distance between the odd row cells and the even row cells, it is necessary to divide it into two cases; for example, set M1 to an odd number and M2 to an even number; therefore, the cell (M1, N1) is on the odd row and the cell (M2, N2) is on the even row; when N1 is less than N2, the distance between them can be calculated by the following formula: When N1 is smaller than N2, the distance between them can be calculated by the following formula: Equation (4) can also be applied to the case where M1 is an even number, M2 is an odd number, and N1 is greater than N2; in addition, equation (5) can also be applied to the case where M1 is an odd number, M2 is an even number, and N1 is less than N2; because there is always a (√3r) / 2 horizontal gap between the same columns of odd and even rows.

4. The method as claimed in claim 1, wherein step 4 is specifically: Fractional frequency reuse (FFR) is a method for saving frequency channels. Although it can reduce interference by using the IFR3 method, it also increases the waste of frequency channels. Users are accustomed to using all frequency channels in a cell, but the available frequency channels are reduced by one third in the IFR3 method. Therefore, the FFR method can make an appropriate balance between interference and available frequency channels. Therefore, setting the same frequency channel in the center of the cell can save frequency resources and reduce interference.

5. The method according to claim 1, wherein step 5 specifically comprises: 1) IFR3 method Integer Frequency Reuse (IFR) is a frequency reuse method. The IFR3 method combines three cells into a cluster. These three cells share all frequency channels equally. This method increases the distance between cells using the same frequency, thereby reducing interference from the same frequency. Through regularity analysis, the coordinates of the same reuse factor are found; 2) SWF algorithm Water injection (SWF) is a power allocation algorithm that allocates power to different frequency channels based on the channel performance and water level. The water level is a threshold for judging channel performance. When the noise and interference are much lower than the water level, more power will be allocated to the channel. The power Pn allocated to the channel depends on the formula: p n =(w-c n ) + (6) W represents the water level, Cn represents the noise and interference normalized by the channel gain in channel n; Cn can be obtained by the following formula: in, σ2 represents the noise power, In represents the interference power in channel n, |H n | 2 represents the channel n gain; the symbol is defined as: (x) + =max{0,x} (8) Since the total power transmitted is known, the water level can be defined by the following formula: P represents the total power transmitted, Pi represents the power on channel i, and N represents the total number of channels in the network; 3) TX-ITL algorithm The transmit interference temperature limit (TX-ITL) algorithm sets a limit Sk for the performance of the channel based on the SWF algorithm; in the TX-ITL algorithm, the power allocation of channel n is based on the formula: This means that when Cn is above the limit, channel n will not be used due to its poor performance; 4) FLG algorithm The forward-looking game (FLG) algorithm is similar to the SWF algorithm, especially in the FLG algorithm, the current power allocation is based on the power allocation in the previous channel, and the current power allocation on channel n is updated by the following formula: Where Pn* represents the previous power allocation on channel n, and φn can be calculated as follows: Cn* represents the previous noise and interference normalized by the channel gain in channel n; according to the FLG algorithm, the performance of the channel can be predicted by the previous performance, so that power can be allocated to the channel with better performance.

6. The method according to claim 1, wherein step 6 specifically comprises: Because there is fading when the signal is transmitted, the matrix ρm,n is set to represent the power loss when the signal is transmitted from cell m to cell n; the power loss depends on the transmission distance and the communication environment; when making a comparison, the environment is always considered to be the same, so that the ρm,n matrix depends only on the transmission distance; Assume that ρm,n is calculated according to the following formula: D(m, n) represents the distance between cell m and cell n; when comparing, all other identical parameters can be considered to be 1; the capacity of cell n can be calculated according to Shannon's formula: C n =B n ×lb(1+SINR n ) (14) Where B represents the frequency bandwidth and SINR represents the signal to interference plus noise radio in each cell; the SINR in cell n can be calculated by the following formula: Where Pn represents the transmission power from the base station in cell n, Pm represents the transmission power from the base station in cell m, σ2 represents the noise power, and H(m,n) represents the same frequency channel in cell m and cell n, which means that there will be same frequency interference; in this formula, the numerator is the useful transmission power, and the denominator is the combination of the same frequency interference and noise; the capacity of each cell can be obtained by formulas (14) and (15), and the total capacity of the network can be calculated by combining the capacities of all cells in the network.