Segment index mapping-based coverage group sampling data compression storage method

Abstract

The application relates to the technical field of chip test verification, in particular to a coverage group sampling data compression storage method based on segmented index mapping, which differentiates one-dimensional sampling points and multi-dimensional sampling points, further differentiates user-defined bins and automatically created bins for the multi-dimensional sampling points, adopts three different segmented index mapping strategies of base index, sequence index and offset index, realizes efficient compression storage of sampling data, replaces original data storage with data index and statistical coverage times in the sampling data, avoids storage overhead of metadata such as repeated storage of bin names and boundary ranges, greatly compresses data volume, only records effective sampling data, further reduces waste of storage resources, supports mixed use of user-defined bins and automatically created bins, and improves flexibility and universality of sampling data representation.

Description

Technical Field

[0001] This invention relates to the field of chip testing and verification technology, and in particular to a method for compressing and storing cover group sampling data based on segmented index mapping. Background Technology

[0002] In the field of chip verification, functional coverage in SystemVerilog is a key metric for measuring verification completeness. The covergroup structure, through a cross-coverage mechanism, combines multiple coverpoints to construct a multi-dimensional data space encompassing all possible combinations of variable states.

[0003] In existing technologies, the data model of cross-coverage can typically be abstracted as a high-dimensional matrix, with each state combination corresponding to a sampled data point being a high-dimensional coordinate point in the matrix. However, when facing large-scale, complex industrial designs, the number of generated auto bins increases exponentially with the increase in the dimension of cross-coverage. This method of recording sampled data often requires explicitly storing the definition information or complete coordinate data of these auto bins, resulting in storage resources being occupied by a large amount of redundant information and the sampled data record file becoming excessively large.

[0004] Furthermore, in actual simulations, state spaces with large cross-coverage ratios are often only partially covered, exhibiting characteristics of sparse matrices. If existing technologies use a fixed-structure storage method to cover and record every coordinate point in a high-dimensional matrix, a large number of invalid zero values ​​will be recorded. On the other hand, if a coordinate recording method is used, bandwidth will be wasted because each sampling point needs to carry complete multi-dimensional coordinate information.

[0005] Meanwhile, in the post-processing stage of sampled data, quickly mapping massive amounts of multidimensional coordinate data back to specific bin definitions is often a complex addressing process. Existing encoding formats cause the decoder to consume a large amount of computational resources when reconstructing the coverage model, resulting in decreased simulation throughput and slow coverage merging speed.

[0006] Therefore, how to reduce the storage overhead of coverage data and improve the processing performance of coverage data has become an urgent problem to be solved. Summary of the Invention

[0007] To address the aforementioned technical problems, the technical solution adopted by this invention is as follows: A method for compressing and storing cover group sampled data based on segmented index mapping, the method comprising the following steps: S1. For any sampling point, if the sampling point is a one-dimensional sampling point, then the base index corresponding to the one-dimensional coverage bin into which the sampling point falls is used as the data index corresponding to the sampling point. S2, If the sampling point is a multi-dimensional sampling point, determine whether the cross-section into which the sampling point falls is a user-defined cross-section. If the cross-section into which the sampling point falls is a user-defined cross-section, use the sequence index corresponding to the cross-section into which the sampling point falls as the data index of the sampling point. If the cross-section into which the sampling point falls is an automatically created cross-section, use the offset index corresponding to the cross-section into which the sampling point falls as the data index corresponding to the sampling point. S3, traverse all sampling points, for any data index that has a corresponding sampling point, use the number of sampling points corresponding to the data index as the coverage count corresponding to the data index, and form the sampling data with the data index and the coverage count corresponding to the data index; S4. Traverse all data indices that have corresponding sampling points to obtain several sample data. Based on the subspace to which the data index belongs in each sample data, determine the several sample data contained in each subspace. S5. For any subspace, form the corresponding sampling data sequence based on the sampling data contained in the subspace.

[0008] Compared with the prior art, the present invention has significant advantages. Through the above technical solution, the coverage group sampling data compression and storage method provided by the present invention, based on segmented index mapping, achieves considerable technical progress and practicality, and has broad industrial application value. It has at least the following advantages: This invention distinguishes between one-dimensional and multi-dimensional sampling points, and further differentiates between user-defined and automatically created warehouses for multi-dimensional sampling points. It employs three different segmented index mapping strategies—base index, sequential index, and offset index—to achieve efficient compressed storage of sampled data. By replacing the original data storage with data indexing and statistical coverage counts in the sampled data, it avoids the storage overhead of repeatedly storing metadata such as warehouse names and boundary ranges, greatly compressing the data volume and recording only valid sampled data. This further reduces the waste of storage resources and supports the mixed use of user-defined and automatically created warehouses, improving the flexibility and versatility of sampled data representation. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1This is a flowchart illustrating a method for compressing and storing cover group sampling data based on segmented index mapping, provided in an embodiment of the present invention. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] This embodiment provides a method for compressing and storing cover group sampling data based on segmented index mapping. See [link to relevant documentation]. Figure 1 This is a flowchart illustrating a method for compressing and storing cover group sampled data based on segmented index mapping, provided by an embodiment of the present invention. The method includes the following steps: S1. For any sampling point, if the sampling point is a one-dimensional sampling point, then the base index corresponding to the one-dimensional coverage bin into which the sampling point falls is used as the data index corresponding to the sampling point. S2, If the sampling point is a multi-dimensional sampling point, determine whether the cross-section into which the sampling point falls is a user-defined cross-section. If the cross-section into which the sampling point falls is a user-defined cross-section, use the sequence index corresponding to the cross-section into which the sampling point falls as the data index of the sampling point. If the cross-section into which the sampling point falls is an automatically created cross-section, use the offset index corresponding to the cross-section into which the sampling point falls as the data index corresponding to the sampling point. S3, traverse all sampling points, for any data index that has a corresponding sampling point, use the number of sampling points corresponding to the data index as the coverage count corresponding to the data index, and form the sampling data with the data index and the coverage count corresponding to the data index; S4. Traverse all data indices that have corresponding sampling points to obtain several sample data. Based on the subspace to which the data index belongs in each sample data, determine the several sample data contained in each subspace. S5. For any subspace, form the corresponding sampling data sequence based on the sampling data contained in the subspace.

[0013] This embodiment describes a single sampling round as an example. Implementers should know that in multiple sampling rounds, each sampling round records sampling data according to the process provided in this embodiment.

[0014] A sampling point refers to the state value hit in the current sampling round. A sampling point can be a one-dimensional sampling point, in which case the sampling point corresponds to a single one-dimensional coverage dimension, or a multi-dimensional sampling point, in which case the sampling point corresponds to multiple one-dimensional coverage dimensions.

[0015] A one-dimensional coverage bin refers to a coverage bin that is defined by the user or automatically created by a tool under a corresponding one-dimensional coverage dimension. A one-dimensional coverage bin corresponds to a value range under a one-dimensional coverage dimension. The one-dimensional coverage bin that a sampling point falls into means that the value range of the one-dimensional coverage bin contains the state value corresponding to the sampling point.

[0016] For example, for an 8-bit binary variable, the state range of the variable in decimal representation is [0, 127]. If a one-dimensional coverage bin B1 with a corresponding state range of [0, 10] is defined under the one-dimensional coverage dimension of the variable, and the state value of sampling point P1 is 5, then sampling point P1 falls into the one-dimensional coverage bin B1. If the state value of sampling point P2 is 11, then sampling point P2 does not fall into the one-dimensional coverage bin B1.

[0017] A base index is a linear sequence number assigned to a one-dimensional cover bin within a one-dimensional cover dimension. It represents the position of the corresponding one-dimensional cover bin in the one-dimensional subspace corresponding to the one-dimensional cover dimension.

[0018] User-defined warehouses are regular warehouses set up by users. Sequential indexes are linear serial numbers assigned to user-defined warehouses to indicate the position of the corresponding warehouse among all user-defined warehouses.

[0019] Automatically created warehouses refer to warehouses created automatically by the tool. The offset index is the sequence number assigned to an automatically created warehouse, which indicates the position of the corresponding warehouse among all automatically created warehouses.

[0020] The data index is used to uniquely identify the bin where the sampling point falls. The number of sampling points corresponding to the data index refers to the number of sampling points falling into the corresponding bin. The sampling data refers to the binary data composed of the data index and the coverage count, which realizes the compressed storage of the sampling point list.

[0021] A subspace refers to the state space defined under the cover group model. A subspace can be a one-dimensional subspace or a two-dimensional subspace, and a cover group model can include multiple subspaces.

[0022] A sampled data sequence refers to a sequence formed by arranging multiple sampled data in the same subspace. It should be noted that regardless of the arrangement order used to form the sampled data sequence, the decoding of the sampled data sequence is still processed at the sampled data granularity. Therefore, this embodiment does not restrict the arrangement order of the sampled data sequence, but it should be ensured that a single sampled data is not split. That is, in the sampled data sequence, after the data index of any sampled data, there must be the number of times that sampled data is covered.

[0023] Specifically, the value ranges corresponding to user-defined warehouses are allowed to overlap, and the union of all user-defined warehouses in a one-dimensional subspace is allowed to be smaller than the value range of the subspace.

[0024] In one specific implementation, for any one-dimensional covering bin, the method for obtaining the base index corresponding to the one-dimensional covering bin includes: Determine the subspace to which the one-dimensional cover warehouse belongs. If the subspace to which the one-dimensional cover warehouse belongs contains one-dimensional cover warehouses that are user-defined warehouses, then determine the base index corresponding to the one-dimensional cover warehouse according to the definition order of the one-dimensional cover warehouses in the subspace to which the one-dimensional cover warehouse belongs.

[0025] In this context, the subspace to which a one-dimensional cover repository belongs is a one-dimensional subspace. When the subspace to which the one-dimensional cover repository belongs contains a one-dimensional cover repository that is a user-defined repository, it is assumed that only the coverage of the user-defined repository needs to be considered, and the range of values ​​not defined by the user does not need to be considered. Therefore, the base index corresponding to the one-dimensional cover repository is determined directly according to the definition order of the one-dimensional cover repository in the subspace to which the one-dimensional cover repository belongs. For a one-dimensional subspace that contains a one-dimensional cover repository that is a user-defined repository, no automatic creation of the repository is performed.

[0026] The definition order can refer to the order in which users declare one-dimensional cover bins in the subspace of the cover group model. In this embodiment, the minimum value of the definition order is set to 0.

[0027] It should be noted that, in order to ensure the continuity of the value range of the one-dimensional subspace, base indexes are allocated for all types of covering bins, including regular covering bins, ignored covering bins, and illegal covering bins.

[0028] For example, in the coverage group model, two regular coverage bins B3 and B4 are defined sequentially under the one-dimensional subspace SP1, and one ignored coverage bin B5. The definition order of regular coverage bin B3 is 0, the definition order of regular coverage bin B4 is 1, and the definition order of ignored coverage bin B5 is 2.

[0029] In one specific implementation, after determining the subspace to which the one-dimensional cover compartment belongs, the method further includes: If the subspace to which the one-dimensional cover warehouse belongs does not contain a one-dimensional cover warehouse that is a user-defined warehouse, then based on the value range corresponding to the subspace to which the one-dimensional cover warehouse belongs, several enumeration values ​​corresponding to the subspace to which the one-dimensional cover warehouse belongs are generated. The base index corresponding to the one-dimensional cover warehouse is determined by the enumeration order of the enumeration values ​​corresponding to the one-dimensional cover warehouse among the several enumeration values ​​corresponding to the subspace to which the one-dimensional cover warehouse belongs.

[0030] In this scenario, if the subspace to which the one-dimensional cover warehouse belongs does not contain a one-dimensional cover warehouse defined by the user, it is considered that the warehouse needs to be automatically created. During automatic creation, the warehouse is created at the smallest granularity of the subspace, resulting in several automatically created warehouses. Similarly, based on the value range corresponding to the subspace to which the one-dimensional cover warehouse belongs, several enumeration values ​​corresponding to the subspace to which the one-dimensional cover warehouse belongs are generated. Also based on the value range and the smallest granularity corresponding to the subspace to which the one-dimensional cover warehouse belongs, several enumeration values ​​corresponding to the subspace to which the one-dimensional cover warehouse belongs are generated. The creation order of the automatically created warehouses corresponds to the enumeration values ​​of the automatically created warehouses. In this embodiment, the minimum value of the enumeration order is set to 0.

[0031] Specifically, to ensure the regularity of the multidimensional subspace structure when a one-dimensional subspace participates in the construction of a multidimensional subspace, for a one-dimensional subspace, even if the user defines an ignore cover warehouse, illegal cover warehouse, etc. for the one-dimensional subspace in the cover group model, as long as the user does not define a regular cover warehouse for the one-dimensional subspace in the cover group model, it is considered that the subspace does not contain a one-dimensional cover warehouse defined by the user. That is, a user-defined warehouse does not refer to any type of cover warehouse set by the user, but only to a regular type of cover warehouse set by the user. The reason is that a one-dimensional subspace that does not contain a user-defined warehouse will still be projected in the multidimensional subspace at the smallest granularity.

[0032] For example, in the cover group model, the one-dimensional subspace SP2 has a value range of {1,2,3} and a minimum granularity of 1. SP2 has an ignored cover bin B6. Then the enumeration order of enumeration value 1 is 0, the enumeration order of enumeration value 2 is 1, and the enumeration order of enumeration value 3 is 2.

[0033] In one specific implementation, the method for obtaining the sequential index corresponding to the cross-warehouse of a user-defined warehouse includes: Determine the subspace to which the cross-link belongs, and determine the sequence index corresponding to the cross-link based on the definition order of the cross-link in the subspace to which it belongs.

[0034] The definition order of cross-repositories can refer to the order in which users declare cross-repositories in the multidimensional subspace of the coverage group model. In this embodiment, the minimum value of the definition order of cross-repositories is set to 0.

[0035] In one specific implementation, the method for obtaining the offset index corresponding to the cross-warehouse for automatically created warehouses includes: Determine the number of one-dimensional coverage dimensions corresponding to the subspace to which the cross-bin belongs; For any one-dimensional coverage dimension, determine the subspace corresponding to that one-dimensional coverage dimension, and use the number of one-dimensional coverage bins contained in the subspace corresponding to that one-dimensional coverage dimension as the size of the base space of that one-dimensional coverage dimension. The base index corresponding to the cross-section is projected onto the one-dimensional coverage dimension, and this is taken as the projection value of the cross-section in the one-dimensional coverage dimension. Traverse all one-dimensional coverage dimensions corresponding to the subspace to which the cross-cell belongs, and obtain the projection value of the cross-cell under all one-dimensional coverage dimensions corresponding to its subspace and the size of the base space of all one-dimensional coverage dimensions corresponding to the subspace to which the cross-cell belongs; The offset index corresponding to the cross-cell is calculated based on the projection values ​​of the cross-cell under all one-dimensional coverage dimensions of its subspace and the base space size of all one-dimensional coverage dimensions of the subspace to which the cross-cell belongs.

[0036] The base space size represents the number of one-dimensional cover bins contained in the corresponding one-dimensional cover dimension, which is also the number of base indexes in the corresponding one-dimensional cover dimension.

[0037] The projection value refers to the base index obtained by mapping the cross-bin to a one-dimensional coverage dimension.

[0038] In one specific implementation, the step of calculating the offset index corresponding to the cross-cell based on the projection values ​​of the cross-cell in all one-dimensional coverage dimensions corresponding to its subspace and the base space size of all one-dimensional coverage dimensions corresponding to the subspace to which the cross-cell belongs, includes: The offset value is determined based on the number of user-defined cross warehouses contained in the subspace to which the cross warehouse belongs; Based on the offset value, the projection values ​​of the cross-cell in all one-dimensional coverage dimensions corresponding to its subspace, and the base space size of all one-dimensional coverage dimensions corresponding to the subspace to which the cross-cell belongs, the offset index corresponding to the cross-cell is calculated using a multidimensional coordinate linear normalization algorithm.

[0039] The offset value is equal to the total number of cross-bins for user-defined bins in the subspace, which can be obtained by adding 1 to the maximum value of the sequential index.

[0040] Specifically, the projection values ​​of the cross-cell in all one-dimensional coverage dimensions corresponding to its subspace can be represented as multidimensional coordinates (V1, V2, ..., V...). N The size of the base space for each one-dimensional covering dimension can be represented as R1, R2, ..., R N Then the offset index is represented as: Where offset represents the offset value, and ∏ represents the cumulative multiplication sign. The value is 1.

[0041] In one specific implementation, the method further includes the following steps: For any subspace, if the subspace corresponds to a single one-dimensional coverage dimension, then the model information corresponding to the subspace is formed by the subspace name, subspace type, base space size, and bin information of the subspace. The subspace type of the subspace is a one-dimensional subspace type, and the bin information of the subspace includes several one-dimensional coverage bin information. The bin information of a single one-dimensional coverage bin includes the bin name, base index, and value range of the corresponding one-dimensional coverage bin. The base space size of the subspace is the number of one-dimensional coverage bins contained in the subspace. If the subspace corresponds to multiple one-dimensional coverage dimensions, then the model information corresponding to the subspace is formed by the subspace name, subspace type, dimension information, total space size, and bin information. The subspace type is a multi-dimensional subspace type. The dimension information of the subspace includes the subspace names corresponding to all one-dimensional coverage dimensions of the subspace. The total space size of the subspace is the product of the base space sizes corresponding to all one-dimensional coverage dimensions of the subspace. The bin information of the subspace includes several cross bin information. Each cross bin information includes the bin name and the multi-dimensional coordinate value of the corresponding cross bin. The multi-dimensional coordinate value of the corresponding cross bin is composed of the projection values ​​of the corresponding cross bin under all one-dimensional coverage dimensions of the subspace. The target model information is formed from the model information corresponding to each subspace.

[0042] The model information describes the partitioning method of the corresponding subspace. The subspace name is a unique identifier of the subspace defined by the user. For example, the subspace name of the one-dimensional subspace SP1 is SP1.

[0043] Subspace types include one-dimensional subspace types and multi-dimensional subspace types.

[0044] The size of the base space refers to the total number of one-dimensional covering bins in a one-dimensional subspace, while the size of the global space refers to the total number of possible cross bins in a multi-dimensional subspace, which is equal to the product of the sizes of the base spaces in each dimension.

[0045] Warehouse information describes the detailed attributes of each warehouse. One-dimensional cover warehouse information includes warehouse name, base index, and value range, while cross warehouse information includes warehouse name and multi-dimensional coordinate values.

[0046] Target model information refers to complete metadata aggregated from model information of all subspaces.

[0047] For example, the one-dimensional subspace SP3 includes one-dimensional cover bins B7 and B8. The one-dimensional cover bin information of one-dimensional cover bin B7 includes bin name B7, base index 0, and value range [0,15]. The one-dimensional cover bin information of one-dimensional cover bin B8 includes bin name B4, base index 1, and value range [16,31].

[0048] The one-dimensional subspace SP4 includes a one-dimensional cover bin B9. The one-dimensional cover bin information of the one-dimensional cover bin B9 includes the bin name B9, the base index 0, and the value range [0,15].

[0049] The dimensional information of the two-dimensional subspace SP5 is SP3 and SP4. The two-dimensional subspace includes cross-bin B10. The cross-bin information includes the name of the cross-bin B10 and the multi-dimensional coordinate value (1, 0) of the cross-bin, indicating that hitting cross-bin B10 requires hitting B7 and B9 at the same time.

[0050] In one specific implementation, the target model information and the sampling data sequences corresponding to each subspace are used for storage or sent to the sampling receiver.

[0051] The sampling receiver refers to the downstream system or module that needs to obtain the sampling data, such as data analysis servers, visualization front-ends, and edge computing nodes.

[0052] Specifically, the target model information and the sampling data sequences corresponding to each subspace can be sent to the sampling receiver together, or the static model information can be sent first, and then the sampling data sequences corresponding to each subspace can be sent according to the sampling round increment. The sampling receiver can independently decode the geometric distribution of the original coverage group model based on the model information, and then decode the sampling information of the current sampling round by combining the sampling data sequences corresponding to each subspace.

[0053] In this embodiment, by distinguishing between one-dimensional and multi-dimensional sampling points, and further distinguishing between user-defined warehouses and automatically created warehouses for multi-dimensional sampling points, three different segmented index mapping strategies—base index, sequential index, and offset index—are adopted respectively, achieving efficient compressed storage of sampling data. In the sampling data, the original data storage is replaced by data index and statistical coverage count, avoiding the storage overhead of duplicate storage of metadata such as warehouse names and boundary ranges, greatly compressing the data volume, recording only valid sampling data, and further reducing the waste of storage resources. It supports the mixed use of user-defined warehouses and automatically created warehouses, improving the flexibility and versatility of sampling data representation.

[0054] While specific embodiments of the invention have been described in detail by way of example, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of the invention. Those skilled in the art should also understand that various modifications can be made to the embodiments without departing from the scope and spirit of the invention. The scope of this invention is defined by the appended claims.

Claims

1. A method for compressing and storing cover group sampled data based on segmented index mapping, characterized in that, The method includes the following steps: S1. For any sampling point, if the sampling point is a one-dimensional sampling point, then the base index corresponding to the one-dimensional coverage bin into which the sampling point falls is used as the data index corresponding to the sampling point. S2, If the sampling point is a multi-dimensional sampling point, determine whether the cross-section into which the sampling point falls is a user-defined cross-section. If the cross-section into which the sampling point falls is a user-defined cross-section, use the sequence index corresponding to the cross-section into which the sampling point falls as the data index of the sampling point. If the cross-section into which the sampling point falls is an automatically created cross-section, use the offset index corresponding to the cross-section into which the sampling point falls as the data index corresponding to the sampling point. S3, traverse all sampling points, for any data index that has a corresponding sampling point, use the number of sampling points corresponding to the data index as the coverage count corresponding to the data index, and form the sampling data with the data index and the coverage count corresponding to the data index; S4. Traverse all data indices that have corresponding sampling points to obtain several sample data. Based on the subspace to which the data index belongs in each sample data, determine the several sample data contained in each subspace. S5. For any subspace, form the corresponding sampling data sequence based on the sampling data contained in the subspace.

2. The method for compressing and storing cover group sampling data based on segmented index mapping according to claim 1, characterized in that, For any one-dimensional covering repository, the methods for obtaining the corresponding base index include: Determine the subspace to which the one-dimensional cover warehouse belongs. If the subspace to which the one-dimensional cover warehouse belongs contains one-dimensional cover warehouses that are user-defined warehouses, then determine the base index corresponding to the one-dimensional cover warehouse according to the definition order of the one-dimensional cover warehouses in the subspace to which the one-dimensional cover warehouse belongs.

3. The method for compressing and storing cover group sampling data based on segmented index mapping according to claim 2, characterized in that, After determining the subspace to which the one-dimensional cover compartment belongs, the following is also included: If the subspace to which the one-dimensional cover warehouse belongs does not contain a one-dimensional cover warehouse that is a user-defined warehouse, then based on the value range corresponding to the subspace to which the one-dimensional cover warehouse belongs, several enumeration values ​​corresponding to the subspace to which the one-dimensional cover warehouse belongs are generated. The base index corresponding to the one-dimensional cover warehouse is determined by the enumeration order of the enumeration values ​​corresponding to the one-dimensional cover warehouse among the several enumeration values ​​corresponding to the subspace to which the one-dimensional cover warehouse belongs.

4. The method for compressing and storing cover group sampling data based on segmented index mapping according to claim 3, characterized in that, The methods for obtaining the sequential index corresponding to the cross-warehouse of a user-defined warehouse include: Determine the subspace to which the cross-link belongs, and determine the sequence index corresponding to the cross-link based on the definition order of the cross-link in the subspace to which it belongs.

5. The method for compressing and storing cover group sampling data based on segmented index mapping according to claim 4, characterized in that, The methods for obtaining the offset index corresponding to the cross-warehouse for automatically created warehouses include: Determine the number of one-dimensional coverage dimensions corresponding to the subspace to which the cross-bin belongs; For any one-dimensional coverage dimension, determine the subspace corresponding to that one-dimensional coverage dimension, and use the number of one-dimensional coverage bins contained in the subspace corresponding to that one-dimensional coverage dimension as the size of the base space of that one-dimensional coverage dimension. The base index corresponding to the cross-section is projected onto the one-dimensional coverage dimension, and this is taken as the projection value of the cross-section in the one-dimensional coverage dimension. Traverse all one-dimensional coverage dimensions corresponding to the subspace to which the cross-cell belongs, and obtain the projection value of the cross-cell under all one-dimensional coverage dimensions corresponding to its subspace and the size of the base space of all one-dimensional coverage dimensions corresponding to the subspace to which the cross-cell belongs; The offset index corresponding to the cross-cell is calculated based on the projection values ​​of the cross-cell under all one-dimensional coverage dimensions of its subspace and the base space size of all one-dimensional coverage dimensions of the subspace to which the cross-cell belongs.

6. The method for compressing and storing cover group sampling data based on segmented index mapping according to claim 5, characterized in that, The step of calculating the offset index corresponding to the cross-cell based on the projection values ​​of the cross-cell in all one-dimensional coverage dimensions of its subspace and the base space size of all one-dimensional coverage dimensions of the subspace to which the cross-cell belongs includes: The offset value is determined based on the number of user-defined warehouses contained in the subspace to which the cross warehouse belongs; Based on the offset value, the projection values ​​of the cross-cell in all one-dimensional coverage dimensions corresponding to its subspace, and the base space size of all one-dimensional coverage dimensions corresponding to the subspace to which the cross-cell belongs, the offset index corresponding to the cross-cell is calculated using a multidimensional coordinate linear normalization algorithm.

7. The method for compressing and storing cover group sampling data based on segmented index mapping according to claim 1, characterized in that, The method further includes the following steps: For any subspace, if the subspace corresponds to a single one-dimensional coverage dimension, then the model information corresponding to the subspace is formed by the subspace name, subspace type, base space size, and bin information of the subspace. The subspace type of the subspace is a one-dimensional subspace type, and the bin information of the subspace includes several one-dimensional coverage bin information. The bin information of a single one-dimensional coverage bin includes the bin name, base index, and value range of the corresponding one-dimensional coverage bin. The base space size of the subspace is the number of one-dimensional coverage bins contained in the subspace. If the subspace corresponds to multiple one-dimensional coverage dimensions, then the model information corresponding to the subspace is formed by the subspace name, subspace type, dimension information, total space size, and bin information. The subspace type is a multi-dimensional subspace type. The dimension information of the subspace includes the subspace names corresponding to all one-dimensional coverage dimensions of the subspace. The total space size of the subspace is the product of the base space sizes corresponding to all one-dimensional coverage dimensions of the subspace. The bin information of the subspace includes several cross bin information. Each cross bin information includes the bin name and the multi-dimensional coordinate value of the corresponding cross bin. The multi-dimensional coordinate value of the corresponding cross bin is composed of the projection values ​​of the corresponding cross bin under all one-dimensional coverage dimensions of the subspace. The target model information is formed from the model information corresponding to each subspace.

8. The method for compressing and storing cover group sampling data based on segmented index mapping according to claim 7, characterized in that, The target model information and the sampling data sequences corresponding to each subspace are used for storage or sent to the sampling receiver.

Images