A planning method for 5G network tracking area

Abstract

The present invention relates to a planning method for a 5G network tracking area, the steps of which are: obtaining road data information and cell parameters; determining the number of TA divisions according to the road data information and the cell parameters; determining the number of clusters according to the number of TA divisions, and determining the position of the initial cluster center according to the distance between the base station and the road; and using K-means clustering to plan the TA according to the similarity between the initial cluster center and the base station cell. Obtain road data and base station cell parameters of the area to be planned, and determine the maximum number of cells that the TA can accommodate. Select several cluster centers according to the distance from the cell to the road, cluster according to the similarity, and obtain clusters equal to the number of cluster centers, each cluster corresponding to a TA, avoiding base station cells covering the same section of road from being divided into different TAs, clustering base station cells with similar geographical density distribution characteristics and business distribution characteristics into the same TA, combining geographical features, considering user business characteristics, directly outputting planning results, and reducing the unreasonable risk of manual division.

Description

Technical Field

[0001] The present invention relates to the field of mobile communications, and more specifically, to a planning method for a 5G network tracking area. Background Art

[0002] With the exponential growth of mobile communication data service traffic, the number of terminal devices has increased dramatically and moved frequently, which greatly increases the difficulty of mobility management on the network side, especially user location management. In 5G networks, the network coverage area is artificially divided into multiple user tracking areas (Tracking Area, TA). Among them, each TA contains a certain number of cells. When a terminal user in the RRC idle state moves within the same TA range, there is no need to frequently perform location update reporting operations. Only when the terminal crosses the TA boundary will the location update operation be performed and location update signaling be generated. Therefore, if the TA division is too small, the number of cells contained and their coverage area are too small, the probability of the terminal user crossing or wandering across the TA boundary increases, which may bring additional location update signaling overhead. However, if the TA division is too large, when a call arrives, the network needs to send a paging message to all cells in the TA described by the user's current location, which brings huge paging resource overhead.

[0003] The existing 5G network TA planning methods usually adopt the TA planning method directly inherited from the 4G network, or plan based on the method of fixed number of cells / fixed area. Both methods follow similar planning principles, namely, TA should be a continuous area geographically, TA should use mountains and rivers as boundaries, TA boundaries should be located in areas with low traffic volume as much as possible, and TA scale should be comprehensively evaluated based on the existing network paging load, etc.

[0004] In the existing technology, Chinese invention patent CN105323769B discloses "a tracking area planning method", which was published on February 10, 2016; obtain the traffic model parameters and service quality parameters of the planning area, determine the calculated value of the number of single-cell single-carrier users according to the above parameters, and obtain the calculated value of the number of single-cell single-carrier users through simulation. Determine the maximum number of users in the tracking area list according to the paging frame related parameters, and determine the number of single-cell users according to the coverage factor and restriction factor of the planning area. When the number of single-cell users is less than the calculated value and the simulated value of the number of single-cell single-carrier users, determine the coverage area of ​​the tracking area in the tracking area list according to the maximum number of users and the number of single-cell users. This invention takes into account the impact of user service characteristics on the TA scale, which can improve the rationality of TA planning to a certain extent and reduce the performance fluctuations caused by subsequent network parameter adjustments. However, it also fails to combine geographical features, resulting in the planner's only basis being the TA coverage area, and unable to determine the preliminary geographical boundaries of the TA. Starting planning from different cells in the area to be planned will lead to different final planning results, and thus lead to a high risk of producing unreasonable planning results. Summary of the invention

[0005] In order to solve the technical defect that the existing TA planning method fails to combine geographical features for demarcation, resulting in unreasonable planning results, the present invention provides a planning method for 5G network tracking areas.

[0006] In order to achieve the above invention purpose, the technical solution adopted is:

[0007] A planning method for a 5G network tracking area comprises the following steps:

[0008] S1: Obtain road data information and cell parameters;

[0009] S2: Determine the number of TA divisions according to road data information and cell parameters;

[0010] S3: Determine the number of clusters according to the number of TA divisions, and determine the location of the initial cluster center according to the distance between the base station and the road;

[0011] S4: Based on the initial cluster center and the similarity between base station cells, K-means clustering is used for TA planning.

[0012] In the above scheme, the road data and base station cell parameters of the area to be planned are obtained, and the maximum number of cells that the TA can accommodate is determined. Several cluster centers are selected according to the distance from the cell to the road, and clustering is further performed according to the cell similarity to obtain clusters equal to the number of cluster centers. Each cluster corresponds to a TA, which avoids the base station cells covering the same section of road being divided into different TAs, resulting in frequent terminal location update operations and unnecessary signaling overhead. The base station cells with similar geographical density distribution characteristics and service distribution characteristics are clustered into the same TA, which combines geographical features and user service characteristics, and can obtain a reasonable balance between location update overhead and paging overhead, and directly output the planning results of the TA, reducing the unreasonable risk of manual division.

[0013] Preferably, in step S1, map information is acquired within the planning area to obtain road data information; according to the 5G network parameter planning, cell parameters of all base stations within the planning area are collected.

[0014] Preferably, the cell parameters include base station type, longitude, latitude, center frequency, transmission power, and affiliated AMF; base station types include outdoor macro base stations, outdoor micro base stations, and indoor distribution systems.

[0015] Preferably, in step S2, before determining the number of TA divisions, the maximum number of cells accommodated by a single TA is determined according to the cell parameters, and the number of cells M accommodated by a single TA satisfies:

[0016]

[0017] Among them, I cell N is the number of paging times per second in a single cell, sub / cell is the number of service users in a single cell, N CAPS N is the number of times the device attempts to call per second. RRC / cell The number of RRC connected users per cell;

[0018] The number of TAs divided k is the total number of base station cells N within the region. p The ratio to the maximum number of cells M that a single TA can accommodate is rounded up:

[0019]

[0020] According to the number of base stations M in a single TA and the average coverage area of ​​the base stations β, the radius r of a single TA can be approximately calculated:

[0021]

[0022] Preferably, in step S3, the base station cell is regarded as a sample point, denoted as point p∈P={p1,p2,…,p Np}, let the road be s, s∈S={s1,s2,…,s Ns}, the geographical distance between the base station cell and the road is calculated as:

[0023] Draw a tangent line for each road, and draw a vertical line segment between the base station cell and the road tangent line. The length of the line segment is recorded as the distance d from the base station p to the road s. (p,s) , all outdoor macro base stations in set P are recorded as set P'. For each outdoor macro base station p', the shortest distance to all roads is selected as the proximity index of the macro base station cell:

[0024]

[0025] The proximity indices of all outdoor macro stations are sorted in ascending order, and k outdoor macro stations are selected as the initial cluster centers.

[0026] Preferably, the initial cluster center is recorded as HEAD = {head1, head2, ..., head k}, in order to avoid the cluster centers being too close, the distance between the cluster centers should be no less than the TA diameter, i.e. 2r; the specific selection process is:

[0027] Ⅰ. When i=1, head i The outdoor macro base station ranked first for the proximity index;

[0028] Ⅱ. When 1 <i≤k,head i Sort by proximity index after head i-1 The first outdoor macro station whose distance to any element in HEAD is greater than 2r.

[0029] Preferably, in step S4, the following steps are included:

[0030] S41: For each sample point that is not a cluster center, calculate its similarity to each cluster center;

[0031] S42: assign each non-cluster center sample point to the target cluster;

[0032] S43: for each cluster, recalculate a new cluster center;

[0033] S44: If the set HEAD of cluster centers is updated, or the cluster to which a sample point belongs is changed, iterative steps S41-S43 are repeated, otherwise the iteration is terminated.

[0034] Preferably, in step S41, the similarity between any two points Calculated as the inverse of the absolute value of the difference between the two point features T:

[0035]

[0036] For any sample point p, its eigenvalue T is defined as: the distance X from the point to the cluster center (p , head) 、The geographical distribution density of the point Y (p) And the business distribution characteristics Z of this point (p) The weighted sum of is:

[0037] T (p) =w1*X (p,head) +w2*Y (p) +w3*Z (p) (6)

[0038] The distance X between point p and the cluster center (p) Calculation: Based on the latitude and longitude information of the sample points, the plane coordinates (x (p) ,y (p) ); Calculate the Chebyshev distance from each point p to the center of each cluster based on the plane coordinate information and round it up:

[0039]

[0040] Geographical distribution density Y of point p (p) It is defined as the ratio of the sum of the distances d between any two points in the region to the sum of the distances from point p to other points:

[0041]

[0042] The traffic distribution characteristic Z of point p (p) UL for a certain period of time at this point (p) , Downlink (p) Flow ratio:

[0043]

[0044] Preferably, in step S42, each non-cluster center sample point is assigned to the target cluster, and the specific steps are:

[0045] Ⅰ. For a non-cluster center sample point, select the cluster center with the greatest similarity as the target cluster center head target ;

[0046] Ⅱ. Determine sample points and head target Whether they belong to the same AMF, if so, they belong to the head target If it belongs to the cluster, end the step; if it does not belong to the cluster, target Eliminate the selection range of this sample point.

[0047] Preferably, in step S43, the distance from each point in the cluster to the cluster center, the geographical distribution density and the business distribution characteristics are calculated to obtain the characteristic value of each point, and then the average characteristic value in the cluster is calculated, and the point with the smallest difference from the average characteristic value is the new cluster center of the cluster.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The present invention provides a planning method for a 5G network tracking area, which obtains road data and base station cell parameters of the area to be planned, and determines the maximum number of cells that a TA can accommodate. According to the distance from the cell to the road, several cluster centers are selected, and clustering is further performed according to the cell similarity to obtain clusters equal to the number of cluster centers. Each cluster corresponds to a TA, so that base station cells covering the same section of road are not divided into different TAs, resulting in frequent terminal location update operations and unnecessary signaling overhead. Base station cells with similar geographical density distribution characteristics and service distribution characteristics are clustered into the same TA, which combines geographical features and takes into account user service characteristics, and can obtain a reasonable balance between location update overhead and paging overhead, and directly output the planning result of the TA, reducing the unreasonable risk of manual division. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is a flow chart of the method of the present invention;

[0051] Figure 2 is a flow chart of a method according to an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of the TA planning results of the present invention. DETAILED DESCRIPTION

[0053] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0054] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0055] Example 1

[0056] like Figure 1 As shown, a planning method for a 5G network tracking area includes the following steps:

[0057] S1: Obtain road data information and cell parameters;

[0058] S2: Determine the number of TA divisions according to road data information and cell parameters;

[0059] S3: Determine the number of clusters according to the number of TA divisions, and determine the location of the initial cluster center according to the distance between the base station and the road;

[0060] S4: Based on the initial cluster center and the similarity between base station cells, K-means clustering is used for TA planning.

[0061] In the above scheme, the road data and base station cell parameters of the area to be planned are obtained, and the maximum number of cells that the TA can accommodate is determined. Several cluster centers are selected according to the distance from the cell to the road, and clustering is further performed according to the cell similarity to obtain clusters equal to the number of cluster centers. Each cluster corresponds to a TA, which avoids the base station cells covering the same section of road being divided into different TAs, resulting in frequent terminal location update operations and unnecessary signaling overhead. The base station cells with similar geographical density distribution characteristics and service distribution characteristics are clustered into the same TA, which combines geographical features and user service characteristics, and can obtain a reasonable balance between location update overhead and paging overhead, and directly output the planning results of the TA, reducing the unreasonable risk of manual division.

[0062] Preferably, in step S1, map information is acquired within the planning area to obtain road data information; according to the 5G network parameter planning, cell parameters of all base stations within the planning area are collected.

[0063] Preferably, the cell parameters include base station type, longitude, latitude, center frequency, transmission power, and affiliated AMF; base station types include outdoor macro base stations, outdoor micro base stations, and indoor distribution systems.

[0064] Preferably, in step S2, before determining the number of TA divisions, the maximum number of cells accommodated by a single TA is determined according to the cell parameters, and the number of cells M accommodated by a single TA satisfies:

[0065]

[0066] Among them, I cell N is the number of paging times per second in a single cell, sub / cell is the number of service users in a single cell, N CAPS N is the number of times the device attempts to call per second. RRC / cell The number of RRC connected users per cell;

[0067] The number of TAs divided k is the total number of base station cells N within the region. p The ratio to the maximum number of cells M that a single TA can accommodate is rounded up:

[0068]

[0069] According to the number of base stations M in a single TA and the average coverage area of ​​the base stations β, the radius r of a single TA can be approximately calculated:

[0070]

[0071] Preferably, in step S3, the base station cell is regarded as a sample point, denoted as point p∈P={p1,p2,…,p Np}, let the road be s, s∈S={s1,s2,…,s Ns}, the geographical distance between the base station cell and the road is calculated as:

[0072] Draw a tangent line for each road, and draw a vertical line segment between the base station cell and the road tangent line. The length of the line segment is recorded as the distance d from the base station p to the road s. (p,s) , all outdoor macro base stations in set P are recorded as set P'. For each outdoor macro base station p', the shortest distance to all roads is selected as the proximity index of the macro base station cell:

[0073]

[0074] The proximity indices of all outdoor macro stations are sorted in ascending order, and k outdoor macro stations are selected as the initial cluster centers.

[0075] Preferably, the initial cluster center is recorded as HEAD = {head1, head2, ..., head k}, in order to avoid the cluster centers being too close, the distance between the cluster centers should be no less than the TA diameter, i.e. 2r; the specific selection process is:

[0076] Ⅰ. When i=1, head i The outdoor macro base station ranked first for the proximity index;

[0077] Ⅱ. When 1 <i≤k,head i Sort by proximity index after head i-1 The first outdoor macro station whose distance to any element in HEAD is greater than 2r.

[0078] Preferably, in step S4, the following steps are included:

[0079] S41: For each sample point that is not a cluster center, calculate its similarity to each cluster center;

[0080] S42: assign each non-cluster center sample point to the target cluster;

[0081] S43: for each cluster, recalculate a new cluster center;

[0082] S44: If the set HEAD of cluster centers is updated, or the cluster to which a sample point belongs is changed, iterative steps S41-S43 are repeated, otherwise the iteration is terminated.

[0083] Preferably, in step S41, the similarity between any two points Calculated as the inverse of the absolute value of the difference between the two point features T:

[0084]

[0085] For any sample point p, its eigenvalue T is defined as: the distance X from the point to the cluster center (p , head) 、The geographical distribution density of the point Y (p) And the business distribution characteristics Z of this point (p) The weighted sum of is:

[0086] T (p) =w1*X (p,head) +w2*Y (p) +w3*Z (p) (6)

[0087] The distance X between point p and the cluster center (p) Calculation: Based on the latitude and longitude information of the sample points, the plane coordinates (x (p) ,y (p) ); Calculate the Chebyshev distance from each point p to the center of each cluster based on the plane coordinate information and round it up:

[0088]

[0089] Geographical distribution density Y of point p (p) It is defined as the ratio of the sum of the distances d between any two points in the region to the sum of the distances from point p to other points:

[0090]

[0091] The traffic distribution characteristic Z of point p (p) UL for a certain period of time at this point (p) , Downlink (p) Flow ratio:

[0092]

[0093] Preferably, in step S42, each non-cluster center sample point is assigned to the target cluster, and the specific steps are:

[0094] Ⅰ. For a non-cluster center sample point, select the cluster center with the greatest similarity as the target cluster center head target ;

[0095] Ⅱ. Determine sample points and head target Whether they belong to the same AMF, if so, they belong to the head target If it belongs to the cluster, end the step; if it does not belong to the cluster, target Eliminate the selection range of this sample point.

[0096] Preferably, in step S43, the distance from each point in the cluster to the cluster center, the geographical distribution density and the business distribution characteristics are calculated to obtain the characteristic value of each point, and then the average characteristic value in the cluster is calculated, and the point with the smallest difference from the average characteristic value is the new cluster center of the cluster.

[0097] Example 2

[0098] like Figure 2 and Figure 3 As shown, the TA planning method of the embodiment of the present invention includes:

[0099] Step 1: Obtain road data information and cell parameters

[0100] The road data is obtained by obtaining map information within the planning area. The road data includes: road name (name), level (level), longitude (lon), latitude (lat), contour type (shape) and open street map code (osm_id), as shown in Table 1.

[0101] Table 1 Road data

[0102]

[0103] Collect the cell parameters of all base stations in the planning area, including base station type, longitude, latitude, station height, center frequency, transmission power, AMF and other information, as shown in Table 2.

[0104] Table 2 Cell parameters

[0105]

[0106] Step 2: Determine the number of TA divisions

[0107] The number of cells M accommodated by a single TA is calculated according to formula (1). The number of paging times per second in a single cell is I cell According to the paging mechanism specified by 3GPP and the existing network parameter configuration, the number of users served by a single cell is set to 3200. sub / cell =600, the number of times the device attempts to call per second N CAPS According to the operator's requirements, it is set to 120. The number of RRC connected users per cell is N. RRC / cell According to the equipment capacity, it is set to 2400. Substituting into formula (1), we can get M≤106.

[0108] The number of TA divisions k is calculated according to formula (2). Where, the total number of base station cells in the planned area is N p There are 196 of them. Substituting them into formula (2), we get k=2.

[0109] The radius r of a single TA is approximately calculated according to formula (3). The base station cell coverage area β is related to the cell transmission power, station height, frequency band, and coverage scenario. Under the conditions of covering dense urban areas, total transmission power of 200W, station height of 30m, and carrier frequency of 3.5GHz, the cell coverage area β is approximately 66510m 2 Substituting into formula (3) we get r≈1500m.

[0110] Step 3: Determine k=2 initial cluster centers

[0111] The geographical distance between each base station cell and the road is calculated to obtain the proximity of all base station cells in the area to be planned.

[0112] The proximity indexes of all outdoor macro stations are sorted in ascending order, and two outdoor macro stations are selected as the initial cluster centers. The selection process is as follows: the outdoor macro station with the highest proximity index is recorded as head1, and the first outdoor macro station with a proximity index lower than head1 and a distance greater than 3000m (i.e., 2r) from any element in the set HEAD is recorded as head2.

[0113] Step 4: K-means clustering to achieve TA planning

[0114] For each non-cluster center sample point, calculate its similarity to the two cluster centers mentioned above. According to formulas (7), (8), and (9), calculate the Chebyshev distance X between the sample point and the cluster center respectively. (p) , the geographical distribution density of sample points Y (p) and business distribution characteristics Z (p) , substituting into formula (6) we can get the characteristic value T of the sample point (p) Among them, w1, w2 and w3 are weighting coefficients, which can be adjusted manually. Furthermore, according to the characteristic value T of the sample point (p) Substitute into formula (5) to calculate the similarity between any two points

[0115] For any non-cluster center sample point, select the cluster center with the greatest similarity as the target cluster center and assign the sample point to the corresponding cluster. Then, calculate the average feature value within each cluster and reselect k cluster centers. Iterate until the cluster center set HEAD is no longer updated and the affiliation of any point is no longer changed.

[0116] The obtained base station cell clustering result is the TA division result. In the embodiment of the present invention, the TA planning result is shown as follows: Figure 3 shown.

[0117] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A planning method for a 5G network tracking area, characterized in that: The following steps are involved: S1: Obtain road data information and cell parameters; S2: Determine the number of TA divisions based on road data information and cell parameters; the number of TA divisions k is the total number of base station cells N within the geographical area. p The ratio to the maximum number of cells M that a single TA can accommodate is rounded up; S3: Determine the number of clusters according to the number of TA divisions, and determine the location of the initial cluster center according to the distance between the base station and the road; wherein: The base station cell is regarded as a sample point, denoted as point p∈P={p1,p2,…,p Np }, let the road be s, s∈S={s1,s2,…,s Ns }, the geographical distance between the base station cell and the road is calculated as: Draw a tangent line for each road, and draw a vertical line segment between the base station cell and the road tangent line. The length of the line segment is recorded as the distance d from the base station p to the road s. (p,s) , all outdoor macro base stations in set P are recorded as set P'. For each outdoor macro base station p', the shortest distance to all roads is selected as the proximity index of the macro base station cell: Sort the proximity indices of all outdoor macro stations in ascending order, and select k outdoor macro stations as the initial cluster centers; S4: Based on the initial cluster center and the similarity between base station cells, K-means clustering is used for TA planning.

2. A method for planning a 5G network tracking area according to claim 1, characterized in that: In step S1, map information is obtained within the planning area to obtain road data information; according to the 5G network parameter planning, cell parameters of all base stations within the planning area are collected.

3. A method for planning a 5G network tracking area according to claim 2, characterized in that: The cell parameters include base station type, longitude, latitude, center frequency, transmission power, and AMF; base station types include outdoor macro base stations, outdoor micro base stations, and indoor distribution systems.

4. A method for planning a 5G network tracking area according to claim 1, characterized in that: In step S2, before determining the number of TA divisions, the maximum number of cells accommodated by a single TA is determined according to the cell parameters, and the number of cells M accommodated by a single TA satisfies: Among them, I cell N is the number of paging times per second in a single cell, sub / cell is the number of service users in a single cell, N CAPS N is the number of times the device attempts to call per second. RRC / cell The number of RRC connected users per cell; The number of TAs divided k is the total number of base station cells N within the region. p The ratio to the maximum number of cells M that a single TA can accommodate is rounded up: According to the number of base stations M in a single TA and the average coverage area of ​​the base stations β, the radius r of a single TA can be approximately calculated:

5. A method for planning a 5G network tracking area according to claim 1, characterized in that: The initial cluster center is denoted as HEAD = {head1, head2, ..., head k }, in order to avoid the cluster centers being too close, the distance between the cluster centers should be no less than the TA diameter, i.e. 2r; the specific selection process is: Ⅰ. When i=1, head i The outdoor macro base station ranked first for the proximity index; Ⅱ. When 1 <i≤k,head i Sort by proximity index after head i-1 The first outdoor macro station whose distance to any element in HEAD is greater than 2r.

6. A method for planning a 5G network tracking area according to claim 1, characterized in that: In step S4, the following steps are included: S41: For each sample point that is not a cluster center, calculate its similarity to each cluster center; S42: assign each non-cluster center sample point to the target cluster; S43: for each cluster, recalculate a new cluster center; S44: If the set HEAD of cluster centers is updated, or the cluster to which a sample point belongs is changed, iterative steps S41-S43 are repeated, otherwise the iteration is terminated.

7. A method for planning a 5G network tracking area according to claim 6, characterized in that: In step S41, the similarity between any two points Calculated as the inverse of the absolute value of the difference between the two point features T: For any sample point p, its eigenvalue T is defined as: the distance X from the point to the cluster center (p,head) 、The geographical distribution density of the point Y (p) And the business distribution characteristics Z of this point (p) The weighted sum of is: T (p) =w1*X (p,head) +w2*Y (p) +w3*Z (p) (6) Among them, w1, w2, w3 are weighting coefficients; The distance X between point p and the cluster center (p) Calculation: Based on the latitude and longitude information of the sample points, the plane coordinates (x (p) ,y (p) ); Calculate the Chebyshev distance from each point p to the center of each cluster based on the plane coordinate information and round it up: Geographical distribution density Y of point p (p) It is defined as the ratio of the sum of the distances d between any two points in the region to the sum of the distances from point p to other points: Business distribution characteristics Z of point p (p) The uplink UL within a certain period of time at this point (p) , Downlink (p) Flow ratio:

8. A method for planning a 5G network tracking area according to claim 6, characterized in that: In step S42, each non-cluster center sample point is assigned to the target cluster. The specific steps are as follows: Ⅰ. For a non-cluster center sample point, select the cluster center with the greatest similarity as the target cluster center head target ; Ⅱ. Determine sample points and head target Whether they belong to the same AMF, if so, they belong to the head target If it belongs to the cluster, end the step; if it does not belong to the cluster, target Eliminate the selection range of this sample point.

9. A method for planning a 5G network tracking area according to claim 6, characterized in that: In step S43, the distance from each point in the cluster to the cluster center, the geographical distribution density and the business distribution characteristics are calculated to obtain the characteristic value of each point, and then the average characteristic value in the cluster is calculated. The point with the smallest difference from the average characteristic value is the new cluster center of the cluster.

Images