JP2022006693A - Information processing device, information processing method and program - Google Patents

Abstract

To provide an information processing device, an information processing method and a program which can efficiently analyze data, such as a prediction of time series data of received power, for example.SOLUTION: An information processing device comprises a control unit which: performs clustering for a plurality of partial regions included in a region which is to be a target of analysis; analyzes data of the partial region for each cluster obtained by the clustering; and acquires a result of the analysis for each partial region based on the result of the analysis for each cluster.SELECTED DRAWING: Figure 1

Description

The present invention relates to an information processing apparatus, an information processing method and a program.

Multiple wireless systems may share the same frequency. In such a case, it is necessary to solve the problem of interference in these multiple wireless systems.

When a frequency is shared by multiple different wireless systems, the secondary user uses the wireless station in order to reduce (preferably not affect) the existing wireless system provided by the primary user. There can be major restrictions on where and when it is possible.
Therefore, it is desirable to predict the future of frequency usage by primary users from the viewpoint of investment decisions of secondary users or formulation of usage plans.

For example, in the channel state prediction device described in Patent Document 1, future data of time-series data of busy state and idle state of a radio channel is predicted by a stochastic neural network as whether or not a primary user uses a frequency. (See Patent Document 1).

Japanese Unexamined Patent Publication No. 2019-101792

However, in the conventional technique, when predicting the frequency usage status by the primary user, it is necessary to construct an individual prediction model suitable for each place to be predicted.
For this reason, in the conventional technology, when expanding the range of the place to be predicted, it is necessary to construct a large number of prediction models in proportion to the size of the range, and the amount of calculation may be enormous. ..

The present invention has been made in consideration of such circumstances, and is an information processing device, an information processing method, and an information processing device capable of efficiently analyzing data such as prediction of time-series data of received power. The challenge is to provide the program.

As a configuration example, clustering is performed on a plurality of partial regions included in the region to be analyzed, data of the partial region is analyzed for each cluster obtained by the clustering, and the result of the analysis for each cluster is obtained. It is an information processing apparatus including a control unit which acquires the analysis result for each subregion based on the control unit.

As a configuration example, the information processing apparatus clusters a plurality of partial regions included in the region to be analyzed, analyzes the data of the partial region for each cluster obtained by the clustering, and analyzes the data of the partial region for each cluster. It is an information processing method that acquires the result of analysis for each subregion based on the result of the analysis.

As a configuration example, a step of clustering a plurality of subregions included in an area to be analyzed on a computer, a step of analyzing data of the subregion for each cluster obtained by the clustering, and the cluster. It is a program for executing a step of acquiring the analysis result for each subregion based on the analysis result for each subregion.

According to the information processing apparatus, information processing method and program according to the present invention, data analysis such as prediction of time-series data of received power can be efficiently performed.

It is a figure which shows the schematic structure of the information processing system which concerns on embodiment of this invention. It is a figure which shows the schematic structure of the mesh of the prediction target area which concerns on embodiment of this invention. It is a figure which shows an example of the topographical information of the prediction target area which concerns on embodiment of this invention. It is a figure which shows an example of the received power for each mesh of the prediction target area which concerns on embodiment of this invention. It is a figure which shows an example of the procedure of the clustering phase processing in the prediction processing which concerns on embodiment (1st Embodiment) of this invention. It is a figure which shows an example of the time series data of the received power for each mesh of the prediction target area which concerns on embodiment of this invention. It is a figure which shows an example of the time series data of the normalized received power for each mesh of the prediction target area which concerns on embodiment of this invention. (A) shows an example of time-series data of representative values in one cluster according to the embodiment of the present invention, and (B) shows an example of time-series data of representative values in another cluster according to the embodiment of the present invention. It is a figure which shows. (A) shows an example of time-series data of normalized received power classified into one cluster according to the embodiment of the present invention, and (B) is classified into another cluster according to the embodiment of the present invention. It is a figure which shows an example of the time series data of the normalized received power. It is a figure which shows an example of the procedure of the process of the post-clustering prediction phase in the prediction process which concerns on embodiment of this invention. (A) shows an example of time-series data of the received power of the representative value predicted for one cluster according to the embodiment of the present invention, and (B) is predicted for another cluster according to the embodiment of the present invention. It is a figure which shows an example of the time series data of the received power of a representative value. It is a figure which shows an example of the time series data of the received power predicted for each mesh which concerns on embodiment of this invention. It is a figure which shows an example of the procedure of the clustering phase processing in the prediction processing which concerns on embodiment (second embodiment) of this invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First Embodiment)
<Multiple wireless systems>
In this embodiment, a plurality of wireless systems share the same frequency, and the primary user is prioritized over the secondary user. Then, in the wireless system of the secondary user, the influence on the wireless system of the primary user may be reduced (preferably not affected) according to the usage status of the frequency in the wireless system of the primary user. Suppose it is necessary. Therefore, the secondary user predicts the frequency usage status in the primary user's wireless system.

In the present embodiment, the secondary user has an information processing system including an information processing device that predicts the frequency usage status in the primary user's wireless system. The information processing apparatus and the information processing system may be provided integrally with the wireless system of the secondary user, or may be provided separately from the wireless system of the secondary user.
The wireless system of the primary user may be, for example, one wireless system, or may be a system in which two or more wireless systems owned by the same primary user are collectively captured. , Or two or more wireless systems owned by two or more different primary users may be captured together.

<Information processing system>
FIG. 1 is a diagram showing a schematic configuration of an information processing system 1 according to an embodiment of the present invention.
The information processing system 1 includes an information processing device 11, a received power database 21, an environment information database 22, a cluster time series database 23, a predicted received power database 24, and a line 31.

The information processing device 11, the received power database 21, the environment information database 22, the cluster time series database 23, and the predicted received power database 24 are connected via a line 31.
Here, the line 31 may be, for example, a wired line, a wireless line, or a line including wired and wireless.

As another configuration example, one or more of the received power database 21, the environment information database 22, the cluster time series database 23, and the predicted received power database 24 are provided integrally with the information processing device 11. May be used.

<Information processing equipment>
The information processing device 11 includes an input unit 111, an output unit 112, a storage unit 113, a communication unit 114, and a control unit 115.
The control unit 115 includes a prediction unit 131.

The input unit 111 inputs information from the outside.
As an example, the input unit 111 has an operation unit operated by a user (person), and information corresponding to the operation accepted by the operation unit may be input. As another example, the input unit 111 may be connected to an external device such as a portable storage device and input information output from the external device.

The output unit 112 outputs information to the outside.
As an example, the output unit 112 has a display unit provided with a screen for displaying and outputting information, and the display unit may display and output information on the screen. As another example, the output unit 112 has a speaker that outputs information by sound, and the information may be output by the speaker. As another example, the output unit 112 may be connected to an external device such as a portable storage device and output information to the external device.

The storage unit 113 stores information.
The storage unit 113 includes, for example, a ROM (Read Only Memory) and a RAM (Random Access Memory).
The storage unit 113 stores, for example, a predetermined program and various parameters.

The communication unit 114 communicates with another device via the line 31.
In the present embodiment, the communication unit 114 communicates with the received power database 21, the environment information database 22, the cluster time series database 23, and the predicted received power database 24, respectively, via the line 31.
Here, in the present embodiment, for convenience of explanation, the communication unit 114 is shown as a functional unit separate from the input unit 111 and the output unit 112, but the functions of the communication unit 114 are included in the input unit 111 and the output unit 112. It may be caught.

The control unit 115 performs various processes in the information processing device 11.
In the present embodiment, the control unit 115 has a CPU (Central Processing Unit), and performs various processes by executing a program stored in the storage unit 113. The control unit 115 may use various parameters stored in the storage unit 113 when performing various processes.

The prediction unit 131 predicts the frequency usage status in the wireless system of the primary user.
The prediction unit 131 is an example of an analysis unit. The analysis unit may be configured to perform analysis other than the prediction according to the present embodiment, for example.

<Various databases>
The received power database 21 stores information about the received power to be predicted. The information is stored in advance, for example, before the prediction process.
The environment information database 22 stores information about the environment. The information is stored in advance, for example, before the prediction process.
The cluster time series database 23 stores time series information for each cluster. The information is stored, for example, in the middle of the prediction process.
The predicted received power database 24 stores information regarding the received power of the predicted result.

Here, in the present embodiment, for convenience of explanation, a configuration in which each information is stored in a separate database is shown, but as another configuration example, two or more databases may be integrated.
Moreover, the name of each database may be arbitrary.

<Prediction processing>
The prediction processing performed by the information processing apparatus 11 in the information processing system 1 will be described with reference to FIGS. 2 to 12.

FIG. 2 is a diagram showing a schematic configuration of a mesh 231 of a prediction target region 211 according to an embodiment of the present invention.
FIG. 2 shows an example of the prediction target area 211.
The prediction target area 211 is a geographical area. In the example of FIG. 2, the square-shaped prediction target region 211 is shown, but the shape of the prediction target region 211 may be arbitrary. Further, in the example of FIG. 2, the two-dimensional prediction target region 211 is shown, but it may be applied to the three dimensions.

The prediction target area 211 is divided into mesh 231 which is a plurality of partial areas. In this embodiment, for convenience of explanation, each partial region is referred to as a mesh, but it may be referred to by any other name. Further, in the example of FIG. 2, in order to make the figure easier to see, only one mesh 231 is designated with a reference numeral.
In the example of FIG. 2, each mesh 231 has a square shape, but the shape of the mesh 231 may be arbitrary.
Further, in the example of FIG. 2, the shapes of the plurality of meshes 231 are the same, but as another configuration example, the plurality of meshes 231 may include meshes having different shapes.

FIG. 3 is a diagram showing an example of topographical information of the prediction target area 211 according to the embodiment of the present invention. For convenience of explanation, the topographical information will be referred to as area topographical information 311.
In the example of FIG. 3, an example of the area topographical information 311 is shown for the same area as the prediction target area 211 shown in FIG.
The area topographic information 311 is divided into mesh 331 which is a plurality of partial areas. In the example of FIG. 3, only one mesh 331 is designated for easy viewing.

Here, each mesh 231 shown in FIG. 2 and each mesh 331 shown in FIG. 3 correspond to the same partial region. That is, in the example of FIG. 3, the topographical information of each mesh 331 represents the topographical information of each mesh 231 shown in FIG.

In the example of FIG. 3, four types of terrain marks 351 to 354 are used as marks representing each of the four types of terrain.
That is, the terrain of each mesh 331 is indicated by the terrain mark (any of the terrain marks 351 to 354) represented by the mesh 331.
The number of types of terrain may be two or three, or may be five or more.
As the type of terrain, any terrain may be used, and for example, flat land, mountains, rivers, forests and the like may be used.

The information of the prediction target area 211 and the area topography information 311 may be stored in the environment information database 22, and may be acquired from the environment information database 22 by the information processing apparatus 11.
Further, the information of the prediction target area 211 and the area topography information 311 may be displayed on the screen by, for example, the information processing apparatus 11.
In the present embodiment, for convenience of explanation, the information of the prediction target area 211 and the area topography information 311 are described as separate information, but these information may be integrated.

FIG. 4 is a diagram showing an example of the received power for each mesh 231 of the prediction target region 211 according to the embodiment of the present invention.
In the graph shown in FIG. 4, the horizontal axis represents time and the vertical axis represents received power. As the received power, for example, RSSI (Received Signal Strength Information) may be used, and the unit is [dBm].
FIG. 4 shows the received power characteristic 411 of one mesh 231.

In the present embodiment, the time-series data of the received power in the past and the data of the environmental information are stored in the received power database 21 and the environmental information database 22 in a unified unit for each mesh 231.

In the present embodiment, for convenience of explanation, the prediction process will be described separately as a clustering phase and a post-clustering prediction phase.
FIG. 5 is a diagram showing an example of a procedure for processing the clustering phase in the prediction processing according to the embodiment (first embodiment) of the present invention.

FIG. 5 shows a normalization unit T1, a clustering unit T2, a representative value acquisition unit T3, an RMSE calculation unit T4, a reference value determination unit T5, and a cluster, which are functional units of the prediction unit 131 of the information processing apparatus 11. The number determination unit T6 is shown.
Here, in the present embodiment, these functional parts are integrated.
If these functional units are provided separately, the functional unit that performs the initial or intermediate processing sends the processed information to the next and subsequent functional units, and the functional unit after the next stage has the information. To receive.

The received power database 21 stores the mesh ID and the time series data of the received power. The mesh ID and the time series data of the received power are associated with each other.
The mesh ID is information that identifies each mesh 231 of the prediction target area 211.
The time-series data of the received power is the time-series data of the received power in each mesh 231 of the prediction target area 211.

FIG. 6 is a diagram showing an example of time-series data of received power for each mesh 231 of the prediction target region 211 according to the embodiment of the present invention.
In the graph shown in FIG. 6, the horizontal axis represents time and the vertical axis represents received power.
FIG. 6 shows the received power characteristics 611 to 614 of each of the four meshes 231.
In the example of FIG. 6, the received power characteristics 611 to 614 are shown for the four meshes 231. However, the received power characteristics are stored in the received power database 21 for all the meshes 231.

The environment information database 22 stores the mesh ID and the environment information. The mesh ID and the environment information are associated with each other.
The environmental information is the environmental information in each mesh 231 of the prediction target area 211.
As an example, the environmental information may be the topographical information shown in FIG.
For example, the environmental information may be information on various environments such as geographical information such as satellite images, traffic volume information, and weather information.
Geographical information may include, for example, information such as building density or divisions such as residential or commercial areas.
Further, the environmental information may include, for example, two or more types of environmental information.

(Processing as the normalization unit T1 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 reads and acquires the time-series data of the received power for each mesh 231 stored in the received power database 21 for all the meshes 231 to be predicted, and the acquired mesh 231. Normalize the time series data of the received power for each.
Further, the prediction unit 131 stores information (normalization information) regarding the conversion of the normalization performed for each mesh 231. In the present embodiment, the normalization information is used when performing an inverse transformation with respect to the transformation of the normalization.

Here, in the present embodiment, the normalization conversion is performed so that the fluctuation pattern (variation element) of the time series data of the received power can be compared between different meshes 231.
As the normalization, for example, normalization that converts the time-series data so that the value of the time-series data is in the range of 0 or more and 1 or less may be used.
In normalization according to the normal distribution, the mean value is 0 and the standard deviation is 1.
In this embodiment, the same normalization method is used for all meshes 231. However, as another configuration example, different normalization methods may be used for some meshes 231.

FIG. 7 is a diagram showing an example of time-series data of normalized received power for each mesh 231 of the prediction target region 211 according to the embodiment of the present invention.
In the graph shown in FIG. 7, the horizontal axis represents time and the vertical axis represents normalized received power.
FIG. 7 shows the normalized received power characteristics 711 to 714 in each of the four meshes 231.
In the example of FIG. 7, the received power characteristics 711 to 714 are shown for the four meshes 231 but the normalized received power characteristics are required for all the meshes 231.

(Processing as the clustering unit T2 of the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 acquires time-series data of normalized received power for all meshes 231 to be predicted.
Further, the prediction unit 131 reads and acquires the environmental information of each mesh 231 (each mesh 331) stored in the environment information database 22 for all the meshes 231 to be predicted.
Then, the prediction unit 131 clusters all the meshes 231 to be predicted based on the time-series data of the normalized received power and the environment information for each mesh 231.

Here, in the present embodiment, the prediction unit 131 classifies the plurality of meshes 231 into two or more clusters by clustering.
As the clustering method, various methods may be used. In this embodiment, for example, as a clustering method in which the number of clusters (k) is specified, a method such as k-hape or k-means is used. May be used. Any value may be set as the initial value of the number of clusters in clustering.

Each cluster aggregates a mesh 231 similar with respect to normalized received power time series data and environmental information. As a result, the cluster ID of each cluster is associated with the mesh IDs of all the meshes 231 classified in the cluster.
The cluster ID is information that identifies each cluster.

Here, in the present embodiment, as a preferable example, clustering of the mesh 231 is performed based on the time-series data of the normalized received power in each mesh 231 and the environmental information, but as another configuration example, each mesh is performed. Clustering of the mesh 231 may be performed based on the time series data of the normalized received power in 231. In this case, each cluster aggregates a mesh 231 similar with respect to the normalized received power time series data.

FIG. 8A is a diagram showing an example of time-series data of normalized received power classified into one cluster according to the embodiment of the present invention.
FIG. 8B is a diagram showing an example of time-series data of normalized received power classified into other clusters according to the embodiment of the present invention.
In the graphs shown in FIGS. 8A and 8B, the horizontal axis represents time and the vertical axis represents normalized received power.
FIG. 8A shows normalized received power characteristics 812, 814 classified into one cluster.
FIG. 8 (b) shows the normalized received power characteristics 811 and 813 classified into another cluster.
In addition, in the example of FIG. 8A and FIG. 8B, the case where two meshes 231 are classified into each cluster is shown, but the number of meshes 231 classified into each cluster is determined. It may be one, or it may be three or more.

(Processing as the representative value acquisition unit T3 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 acquires time-series data of the normalized received power for each mesh 231 included in each cluster for all the clusters.
Then, the prediction unit 131 determines the representative value of the time-series data of the received power normalized for each cluster, and acquires the time-series data consisting of the representative value for each cluster.

Here, the representative value may be determined at regular time intervals, for example, from the time-series data of the normalized received power.
Further, as the representative value, for example, a median value, an average value, or the like may be used.

FIG. 9A is a diagram showing an example of time-series data of representative values in one cluster according to the embodiment of the present invention.
FIG. 9B is a diagram showing an example of time-series data of representative values in other clusters according to the embodiment of the present invention.
In the graphs shown in FIGS. 9A and 9B, the horizontal axis represents time and the vertical axis represents normalized received power.

FIG. 9A shows the received power characteristic 911 of the representative value in one cluster.
The example of FIG. 9A corresponds to the example of FIG. 8A, and for reference, the received power characteristics 812 and 814 from which the representative value was obtained are shown.
FIG. 9B shows the received power characteristic 912 as a representative value in the other cluster.
The example of FIG. 9B corresponds to the example of FIG. 8B, and for reference, the received power characteristics 811 and 813 from which the representative value was obtained are shown.

(Processing as RMSE calculation unit T4 possessed by prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 acquires the time-series data of the received power of the representative value in each cluster for all the clusters, and is normalized for each mesh 231 for all the meshes 231. Acquire time series data of received power.
Then, the prediction unit 131 calculates a predetermined root mean square error (RMSE: Root Mean Square Error) for each cluster for all clusters.

The RMSE is a value acquired for each cluster with respect to the square of the difference from the representative value for the time-series data of the normalized received power of all the meshes 231 included in the cluster.
Here, it is considered that the smaller the RMSE of the cluster, the higher the similarity of the mesh 231 included in the cluster.
Any other index may be used as an index for determining the similarity of the mesh 231 for each cluster.

(Processing as the reference value determination unit T5 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 acquires the calculated RMSE.
Then, the prediction unit 131 determines whether or not the RMSE is equal to or less than a predetermined reference value.
As a result of this determination, when the prediction unit 131 determines that the RMSE exceeds a predetermined reference value (that is, is not equal to or less than a predetermined reference value), the prediction unit 131 processes the cluster number designation unit T6.
On the other hand, as a result of this determination, when the prediction unit 131 determines that the RMSE is equal to or less than a predetermined reference value, for all the clusters, the time series data of the received power of the representative value in each cluster, the cluster ID, and each of them. The mesh ID of the mesh 231 included in the cluster and the normalization information for each mesh 231 are stored in the cluster time series database 23.

(Processing as the cluster number designation unit T6 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 increases the number of clusters (number of clusters) by one in the processing of the clustering unit T2 when the calculated RMSE value exceeds a predetermined reference value. To specify.

Here, in the present embodiment, the clustering unit T2 performs clustering with a designated number of clusters. By increasing the number of clusters in clustering, the accuracy of classification of mesh 231 is improved.
The cluster number designation unit T6 may specify the number of clusters so as to increase the number of clusters by two or more.
Further, in the present embodiment, when it is determined that the result of clustering is insufficient, the number of clusters is increased and the clustering is performed again. However, as another configuration example, parameters other than the number of clusters are changed. A configuration may be used in which clustering is performed again.

FIG. 10 is a diagram showing an example of a procedure for processing a post-clustering prediction phase in the prediction process according to the embodiment of the present invention.
FIG. 10 shows a time-series prediction unit T11 and a denormalization unit T12, which are functional units of the prediction unit 131 of the information processing apparatus 11.
When these functional units are provided separately, the functional unit that performs the initial or intermediate processing sends the processed information to the next-stage functional unit, and the next-stage functional unit receives the information. do.

(Processing as the time series prediction unit T11 possessed by the prediction unit 131)
The time-series prediction unit T11 acquires the time-series data of the received power of the representative value in each cluster for all the clusters from the cluster time-series database 23.
As another example, the time-series data of the received power of the representative value in each cluster may be output from the reference value determination unit T5 to the time-series prediction unit T11 for all the clusters.
In the information processing apparatus 11, the prediction unit 131 reads and acquires the time-series data of the received power of the representative value in each cluster for all the clusters from the cluster time-series database 23.
Then, the prediction unit 131 predicts the time-series data of the reception power of the representative value in each cluster in the future for all the clusters based on the time-series data of the reception power of the representative value in each cluster.

Here, any method may be used as the method for predicting the time series data.
As a model for predicting time series data, for example, an autoregressive (AR) model, an autoregressive integrated moving average (ARIMA) model, a Kalman filter model, a particle filter model, or , LSTM (Long Short-Term Memory) and other models of neural networks may be used.

FIG. 11A is a diagram showing an example of time-series data of received power of a representative value predicted for one cluster according to the embodiment of the present invention.
FIG. 11B is a diagram showing an example of time-series data of received power of representative values predicted for other clusters according to the embodiment of the present invention.
In the graphs shown in FIGS. 11A and 11B, the horizontal axis represents time and the vertical axis represents the predicted received power.

FIG. 11A shows the received power characteristic 1111 of the predicted representative value in one cluster.
The example of FIG. 11 (A) corresponds to the example of FIG. 9 (A).
FIG. 11B shows the received power characteristic 1112 of the predicted representative value in another cluster.
The example of FIG. 11B corresponds to the example of FIG. 9B.

(Processing as the denormalization unit T12 of the prediction unit 131)
The time-series prediction unit T11 outputs the time-series data of the received power of the representative value predicted for each cluster to the denormalization unit T12 for all the clusters.
Further, the denormalization unit T12 acquires the normalization information for each mesh 231 for all the meshes 231 from the cluster time series database 23.
As another example, the normalization information for each mesh 231 may be output from the reference value determination unit T5 to the denormalization unit T12 for all the meshes 231.
In the information processing apparatus 11, the prediction unit 131 acquires time-series data of the received power of the representative value predicted for each cluster for all the clusters.
Further, the prediction unit 131 reads and acquires the normalization information for each mesh 231 for all the meshes 231 from the cluster time series database 23.

Then, the prediction unit 131 responds to the normalization information for all meshes 231 based on the time-series data of the received power of the representative value predicted for each cluster and the normalization information for each mesh 231. By performing denormalization, time-series data of the predicted received power is acquired for each mesh 231.
The prediction unit 131 stores the time-series data, the mesh ID, and the cluster ID of the received power predicted for each mesh 231 in the predicted received power database 24 for all the meshes 231.

Here, in denormalization, the inverse conversion of the normalization conversion is performed.
Specifically, the prediction unit 131 reverse-normalizes the time-series data of the received power of the representative value predicted for the cluster including the mesh 231 for each mesh 231 to receive the received power in the mesh 231. Reflect the absolute value of. That is, the prediction unit 131 predicts the representative value for each cluster, and reflects the absolute value of the received power in each mesh 231 included in the cluster in the prediction result for each mesh 231. , Get the time series data of the predicted received power.

FIG. 12 is a diagram showing an example of time-series data of received power predicted for each mesh 231 according to the embodiment of the present invention.
In the graph shown in FIG. 12, the horizontal axis represents time and the vertical axis represents normalized received power.

FIG. 12 shows the received power characteristics 1211-1214 predicted for each mesh 231 for the four meshes 231.
The example of FIG. 12 corresponds to the example of FIG. 6, and the received power characteristics 1211 to 1214 predicted for each mesh 231 shown in FIG. 12 are the received power characteristics 611 to each of the mesh 231 shown in FIG. Corresponds to 614.

<About the first embodiment>
As described above, in the information processing system 1 according to the present embodiment, in the information processing apparatus 11, mesh 231 having high similarity is aggregated by clustering based on the data of each mesh 231 and analyzed for each cluster. , Data can be analyzed efficiently.
For example, as an analysis of time-series data, it is possible to efficiently predict the frequency usage status in other wireless systems that use the same frequency.

Further, in the information processing system 1 according to the present embodiment, in the information processing apparatus 11, mesh 231 having high similarity may be aggregated by clustering based on the data and environment information of each mesh 231. In this case, for example, for example. The effect is that geographical information is taken into account. Specifically, for example, depending on the geographical situation, the influence of having a periodic pattern in the flow of people can be considered.

In the present embodiment, for example, when analyzing time-series data for a plurality of meshes 231 by classifying the plurality of meshes 231 into clusters and using the same number of prediction models as the number of clusters, sufficient accuracy is maintained. As it is, the number of models can be reduced to reduce the calculation load and calculation time. Therefore, for example, even when the area to be analyzed is expanded, the calculation load and the calculation time can be suppressed, and efficient analysis can be realized.
For example, if the area of (10 km x 10 km) is divided by a mesh of (50 m x 50 m), the number of meshes will be 40,000 (= 200 x 200), but the future is predicted for each cluster as a smaller number of clusters. As a result, efficient analysis is realized.

As the analysis, various analyzes may be used, for example, future prediction, past analysis, analysis of the magnitude of the received power fluctuation for each mesh, and the like may be used.

(Second Embodiment)
In the present embodiment, the parts different from the first embodiment will be described in detail, and the parts similar to the first embodiment will not be described in detail.
In the present embodiment, generally, the clustering phase of the prediction processing performed in the information processing system 1 is different from that of the first embodiment in that the clustering phase is different from the example of FIG. 5 according to the first embodiment. It is similar in that. In the present embodiment, the post-clustering prediction phase of the prediction process performed in the information processing system 1 may be, for example, the same as the example of FIG. 10 according to the first embodiment.
In this embodiment, for convenience of explanation, the configurations other than the functional unit related to the clustering phase will be described with the same reference numerals as those in the first embodiment.

<Clustering phase in prediction processing>
FIG. 13 is a diagram showing an example of the procedure of the clustering phase processing in the prediction processing according to the embodiment (second embodiment) of the present invention.

FIG. 13 shows a normalization unit T31, a dimension reduction unit T32, a clustering unit T33, a representative value acquisition unit T34, and an RMSE calculation unit T35, which are functional units of the prediction unit 131 of the information processing apparatus 11. A reference value determination unit T36 and a cluster number designation unit T37 are shown.
When these functional units are provided separately, the functional unit that performs the initial or intermediate processing sends the processed information to the next-stage functional unit, and the next-stage functional unit receives the information. do.

(Processing as the normalization unit T31 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 performs the same processing as the normalization unit T1 shown in FIG. 5 according to the first embodiment as the processing as the normalization unit T31.

(Processing as the dimension reduction unit T32 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 acquires time-series data of normalized received power for all meshes 231 to be predicted.
Further, the prediction unit 131 reads and acquires the environmental information of each mesh 231 (each mesh 331) stored in the environment information database 22 for all the meshes 231 to be predicted.
Then, the prediction unit 131 performs a predetermined dimension reduction process for all the meshes 231 to be predicted based on the time-series data of the normalized received power and the environmental information for each mesh 231.

Here, in the present embodiment, the dimension reduction process extracts only the information suitable for clustering with respect to the time-series data of the normalized received power and the environmental information for each mesh 231 to generate a new feature amount. It is a process. In this embodiment, the feature amount is used as a feature amount for clustering.
Any method may be used as the method for performing the dimension reduction processing, and for example, a method of principal component analysis (PCA: Principal Component Analysis) may be used.

(Processing as the clustering unit T33 of the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 acquires the feature amount for clustering for all the meshes 231 to be predicted.
Then, the prediction unit 131 clusters all the meshes 231 to be predicted based on the feature amount for clustering.
As the clustering method, for example, the same method as the clustering unit T2 shown in FIG. 5 according to the first embodiment may be used.

Here, in the present embodiment, as a preferable example, the feature amount for clustering the mesh 231 is obtained based on the time-series data and the environmental information of the normalized received power in each mesh 231. As a configuration example, a feature amount for clustering the mesh 231 may be obtained based on the time-series data of the normalized received power in each mesh 231.

(Processing as the representative value acquisition unit T34 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 acquires the mesh ID of the mesh 231 included in each cluster for all the clusters.
Further, in the information processing apparatus 11, the prediction unit 131 acquires the time-series data of the received power normalized to each mesh 231 for all the meshes 231.
Then, the prediction unit 131 determines the representative value of the time-series data of the received power normalized for each cluster, and acquires the time-series data consisting of the representative value for each cluster.
As a method for acquiring time-series data consisting of representative values, for example, a method similar to the representative value acquisition unit T3 shown in FIG. 5 according to the first embodiment may be used.

(Processing as the RMSE calculation unit T35 of the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 performs the same processing as the RMSE calculation unit T4 shown in FIG. 5 according to the first embodiment as the processing as the RMSE calculation unit T35.

(Processing as the reference value determination unit T36 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 performs the same processing as the reference value determination unit T5 shown in FIG. 5 according to the first embodiment as the processing as the reference value determination unit T36.

(Processing as the cluster number determination unit T37 possessed by the prediction unit 131)
In the information processing apparatus 11, the prediction unit 131 performs the same processing as the cluster number determination unit T6 shown in FIG. 5 according to the first embodiment as the processing as the cluster number determination unit T37.

<About the second embodiment>
As described above, the information processing system 1 according to the present embodiment can obtain the same effect as that of the first embodiment, and further, by using the feature amount for clustering, as compared with the first embodiment, for example. Therefore, it is possible to improve the accuracy of analysis.

<Configuration example>
As a configuration example, in the information processing apparatus 11, a plurality of partial regions (in the example of FIG. 2, the prediction target region 211) included in the region (prediction target region 211 in the example of FIG. 2) to be an analysis target (prediction target in the embodiment) are included in the information processing apparatus 11. Clustering is performed on the mesh 231), partial region data (received power data in the embodiment) is analyzed for each cluster obtained by the clustering, and partial region analysis is performed based on the analysis result for each cluster. The control unit 115 for acquiring the result of the above is provided.

As a configuration example, in the information processing apparatus 11, the control unit 115 performs clustering based on the data for each partial region.
As a configuration example, in the information processing apparatus 11, the control unit 115 performs clustering based on the data and the environmental information for each partial region.
As an example of the configuration, in the information processing apparatus 11, the control unit 115 performs clustering based on the feature amount for clustering obtained by reducing the dimension (example of FIG. 13).
As an example of the configuration, in the information processing apparatus 11, the data is the data of the received power of the wireless system. In addition, the analysis is a prediction of future received power data for the received power.
As a configuration example, in the information processing apparatus 11, the control unit 115 performs clustering based on the result of normalizing the data for each sub-region for each sub-region, and for each sub-region based on the result of the analysis for each cluster. Perform denormalization for normalization and obtain the result of analysis for each subregion.

As an example of the configuration, in the information processing method (in the embodiment, the method of processing performed by the information processing device 11), the information processing device 11 performs clustering on a plurality of partial regions included in the region to be analyzed, and clustering is performed. The data of the subregion is analyzed for each cluster obtained by, and the result of the analysis for each subregion is acquired based on the result of the analysis for each cluster.

As an example of the configuration, the program includes a step of clustering a plurality of partial regions included in the region to be analyzed on a computer (in the embodiment, a computer constituting the information processing apparatus 11), and a cluster obtained by the clustering. It is a program for executing the step of analyzing the data of each sub-region and the step of acquiring the result of the analysis of each sub-region based on the result of the analysis for each cluster.

A program for realizing the function of an arbitrary component in an arbitrary device such as the information processing apparatus 11 described above is recorded on a computer-readable recording medium, and the program is read into a computer system and executed. You may try to do it. The term "computer system" as used herein includes hardware such as an operating system (OS: Operating System) or peripheral devices. The "computer-readable recording medium" refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, a CD (Compact Disc) -ROM, or a storage device such as a hard disk built in a computer system. .. Further, the "computer-readable recording medium" is a volatile memory (RAM) inside a computer system that serves as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, it shall include those that hold the program for a certain period of time.

Further, the above program may be transmitted from a computer system in which this program is stored in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the "transmission medium" for transmitting a program means a medium having a function of transmitting information, such as a network such as the Internet or a communication line such as a telephone line.
Further, the above program may be for realizing a part of the above-mentioned functions. Further, the above program may be a so-called difference file that can realize the above-mentioned functions in combination with a program already recorded in the computer system. The difference file may be called a difference program.

The functions of any component in any device described above may be realized by a processor. For example, each process in the embodiment may be realized by a processor that operates based on the information of the program or the like and a computer-readable recording medium that stores the information of the program or the like. Here, in the processor, for example, the functions of each part may be realized by individual hardware, or the functions of each part may be realized by integrated hardware. For example, the processor may include hardware, which may include at least one of a circuit that processes a digital signal and a circuit that processes an analog signal. For example, the processor may be configured with one or more circuit devices mounted on a circuit board, or one or both of one or more circuit elements. An IC (Integrated Circuit) or the like may be used as the circuit device, and a resistor or a capacitor may be used as the circuit element.

Here, the processor may be, for example, a CPU. However, the processor is not limited to the CPU, and various processors such as GPU (Graphics Processing Unit) or DSP (Digital Signal Processor) may be used. Further, the processor may be, for example, a hardware circuit based on an ASIC (Application Specific Integrated Circuit). Further, the processor may be composed of, for example, a plurality of CPUs, or may be composed of a hardware circuit by a plurality of ASICs. Further, the processor may be composed of, for example, a combination of a plurality of CPUs and a hardware circuit by a plurality of ASICs. Further, the processor may include, for example, one or more of an amplifier circuit or a filter circuit for processing an analog signal.

Although the embodiments of the present invention have been described in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs and the like within a range that does not deviate from the gist of the present invention.

1 ... Information processing system, 11 ... Information processing device, 21 ... Received power database, 22 ... Environmental information database, 23 ... Cluster time series database, 24 ... Predicted received power database, 31 ... Line, 111 ... Input unit, 112 ... Output Unit, 113 ... Storage unit, 114 ... Communication unit, 115 ... Control unit, 131 ... Prediction unit, 211 ... Prediction target area, 231 ... 811 to 814, 911 to 912, 1111 to 1112, 1211-1214 ... Received power characteristics, T1, T31 ... Normalization part, T2, T33 ... Clustering part, T3, T34 ... Representative value acquisition part, T4, T35 ... RMSE part , T5, T36 ... Reference value determination unit, T6, T37 ... Cluster number specification unit, T11 ... Time series prediction unit, T12 ... Inverse normalization unit, T32 ... Dimension reduction unit

Claims (8)

Clustering is performed on a plurality of subregions included in the region to be analyzed, data of the subregion is analyzed for each cluster obtained by the clustering, and the subregion is analyzed based on the result of the analysis for each cluster. It is equipped with a control unit that acquires the results of each analysis.
Information processing equipment. The control unit performs the clustering based on the data for each partial region.
The information processing apparatus according to claim 1. The control unit performs the clustering based on the data and the environmental information for each partial region.
The information processing apparatus according to claim 2. The control unit performs the clustering based on the feature amount for clustering obtained by reducing the dimension.
The information processing apparatus according to any one of claims 1 to 3. The data is data on the received power of the wireless system.
The analysis is a prediction of future received power data for the received power.
The information processing apparatus according to any one of claims 1 to 4. The control unit performs the clustering based on the result of normalizing the data for each subregion for each subregion, and for the normalization for each subregion based on the result of the analysis for each cluster. Perform denormalization to obtain the results of analysis for each subregion.
The information processing apparatus according to any one of claims 1 to 5. Information processing equipment
Clustering is performed for multiple subregions included in the region to be analyzed.
The data of the partial region is analyzed for each cluster obtained by the clustering, and the data is analyzed.
The result of the analysis for each subregion is acquired based on the result of the analysis for each cluster.
Information processing method. On the computer
A step of clustering multiple subregions contained in the region to be analyzed, and
A step of analyzing the data of the partial region for each cluster obtained by the clustering, and
A step of acquiring the analysis result for each subregion based on the analysis result for each cluster, and
A program to execute.

Images