# US 2024/0202196 A1 · Methods and systems that generate component quantities for system configurations

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2024/0202196 A1
- Document ID: US20240202196A1
- Status: Published application
- Kind: A1
- Application number: 18/082,308
- Filed: 2022-12-15
- Published: 2024-06-20
- Earliest claimed: 2022-12-15
- Assignees: VMware, Inc.
- Applicants: VMware LLC
- Inventors: Ajitesh Gupta
- Web page: https://patentgrep.com/patent/US20240202196A1

## CPC

- G06F 16/24554: {Unary operations; Data partitioning operations}
- G06F 16/2454: {Optimisation of common expressions}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The current document is directed to methods and systems that generate lists of component types and quantities needed for system installations based on parameter values that characterize the system, environment, and application domain, referred to as “input values.” An implementation of a private-5G-network component-type-and-quantity-determination system is disclosed. An initial model used to generate component types and quantities is generated from information acquired from various information sources, including system vendors, designers, and/or administrators. The initial model is used to generate lists of component types and quantities on behalf of requesting entities during an initial period of system operation. Over time, as the final types and quantities of components used for particular installations are determined and reported back to the system, the system uses these reported results to refine the initial model, generate more specific models for particular types or classes of system installations and to refine the more specific models.

## Claims

### Claim 1

A component-types-and-quantities-estimator system (“CTQ estimator”) that estimates the types and quantities of system components needed for a system installation, the CTQ estimator comprising:
a computing platform that provides an execution environment for CTQ-estimator functionality, the computer platform implemented within one or more computer systems that each includes one or more processors, one or memories, and one or more mass-storage appliances or devices; and
computer instructions, stored in one of the one or more memories, that, when executed in the execution environment, control the one or more computer systems to:
receive a request for a list of component types and quantities for installation of a system, the received request including a set of parameter values, and return the requested list of component types and quantities; and
receive a final-configuration report that includes a list of component types and quantities installed in a system installation and a set of parameter values and stores the final-configuration report for use in refining component models used for determining the component types and quantities for installation of systems and for partitioning a model, comprising component models, that corresponds to a parameter-value space or subspace into sub-models corresponding to different parameter-value subspaces.

### Claim 2 (depends on 1)

The CTQ estimator of claim 1 wherein each component type is associated with a single-variate or multivariate polynomial expression used to estimate the quantity of components of the component type needed for the system installation.

### Claim 3 (depends on 2)

The CTQ estimator of claim 2
wherein the values of variables in the polynomial expression are computed from the values of parameters and/or from estimated component quantities; and
wherein the coefficients in the polynomial expression are initially generated from vendor-supplied information, developer-supplied information, and other information sources for component quantities in system installations and stored in the component model for the component type.

### Claim 4 (depends on 3)

The CTQ estimator of claim 3 wherein component models are refined using component quantities returned in final-configuration reports.

### Claim 5 (depends on 4)

The CTQ estimator of claim 4 wherein a component model for a component type is refined using a least-squares method that minimizes the squared differences of component quantities estimated from the polynomial expression associated for the component type and an average of the reported component quantities for the component type.

### Claim 6 (depends on 3)

The CTQ estimator of claim 3 wherein a model comprises an ordered set of component models, one component model for each type of component.

### Claim 7 (depends on 3)

The CTQ estimator of claim 3 wherein the set of component models is ordered so that, for a particular component model that includes a variable computed from an estimated quantity of a different component type, the component model for the different component type precedes the particular component model in the ordered set of component models.

### Claim 8 (depends on 1)

The CTQ estimator of claim 1 wherein, upon receiving the request for a list of component types and quantities for installation of a system from a requestor entity, the CTQ estimator
extracts the set of parameter values from the request;
uses the set of parameter values to identify a model that represents a class of system installation characterized by the set of parameter values;
for each component type for which a component model is included in the model,
uses the component model for the component type to generate an estimate of the quantity of components of the component type needed for the system installation;
includes the estimated component quantities generated from the component models in a list of component types and quantities; and
returns the list of component types and quantities to the requestor entity.

### Claim 9 (depends on 9)

The CTQ estimator of claim 9 wherein the CTQ estimator uses the component model for the component type to generate an estimate of the quantity of components of the component type needed for the system installation by using the polynomial expression for the component type to compute the component quantity by replacing each variable in the polynomial expression with a parameter value or with a component quantity estimated for a different component type and replacing each coefficient in the polynomial expression by a current value for the coefficient stored in the component model.

### Claim 10 (depends on 1)

The CTQ estimator of claim 1 wherein the CTQ estimator includes a comprehensive data structure that comprises a set of model-partition data structures, each model-partition data structure including:
a model corresponding to a particular parameter-value space or subspace, the model comprising a set of component models for each component type;
a parameter-value-space data structure that represents the dimensions and quantization levels for the parameter-value space or subspace, each dimension corresponding to a parameter, each dimension associated with a number of quantization levels computed from values of the parameter, each quantization level associated with an expression used to determine whether a value for the parameter falls into the quantization level, and a count of the processed final-configuration reports that include a value that falls into the quantization level; and
stored final-configuration reports for system installations associated with parameter values that fall within the particular parameter-value space or subspace.

### Claim 11 (depends on 10)

The CTQ estimator of claim 10 wherein the CTQ estimator traverses the comprehensive data structure to identify the model-partition data structure corresponding to a set of parameter values.

### Claim 12 (depends on 11)

The CTQ estimator of claim 11 wherein, when the parameter-value space or subspace associated with a model partition can be partitioned into two parameter-value subspaces that each corresponds to at least a threshold number of final-configuration reports, the CTQ estimator replaces the model partition with two new model partitions that represent the two parameter-value subspaces so that, over time, the CTQ estimator contains increasingly specific models for smaller and more specific parameter-value subspaces.

### Claim 13 (depends on 12)

The CTQ estimator of claim 12 wherein the comprehensive data structure is a tree-like data structure with model partitions as leaf nodes and internal nodes containing expressions that evaluate to a Boolean value when evaluated using a parameter value selected from a set of parameter values.

### Claim 14

A method that automatically estimates the types and quantities of system components needed for a system installation, the method comprising:
receiving a request for a list of component types and quantities for installation of a system, the received request including a set of parameter values;
using the set of parameter values to traverse a comprehensive data structure to identify a model that represents a class of system installation characterized by the set of parameter values;
for each component type for which a component model is included in the model,
using the component model for the component type and a polynomial expression associated with the component type to generate an estimate of the quantity of components of the component type needed for the system installation;
including the estimated component quantities generated from the component models in a list of component types and quantities; and
returning the list of component types and quantities to the requestor entity.

### Claim 15 (depends on 14)

The method of claim 14 wherein the model represents a parameter-value space or subspace corresponding to a particular class of system installations and comprises an ordered set of component models, each corresponding to a component type, that each contains a set of one or more coefficient values for the polynomial expression associated with the associated component type.

### Claim 16 (depends on 15)

The method of claim 15 wherein using the component model for the component type and a polynomial expression associated with the component type to generate an estimate of the quantity of components of the component type needed for the system installation further comprises:
replacing each variable in the polynomial expression with a parameter value or with a component quantity estimated for a different component type;
replacing each coefficient in the polynomial expression by a current value for the coefficient stored in the component model for the component type; and
evaluating the polynomial expression with variable values and coefficient values to generate an estimated component quantity.

### Claim 17 (depends on 15)

The method of claim 15 wherein comprehensive data structure comprises a set of model-partition data structures, each model-partition data structure including:
a model corresponding to a particular parameter-value space or subspace;
a parameter-value-space data structure that represents the dimensions and quantization levels for the particular parameter-value space or subspace, each dimension corresponding to a parameter, each dimension associated with a number of quantization levels computed from values of the parameter, each quantization level associated with an expression used to determine whether a value for the parameter falls into the quantization level, and a count of the processed final-configuration reports that include a value that falls into the quantization level; and
stored final-configuration reports for system installations associated with parameter values that fall within the particular parameter-value space or subspace.

### Claim 18 (depends on 17)

The method of claim 17 further comprising, when the parameter-value space or subspace associated with a model partition can be partitioned into two parameter-value subspaces that each corresponds to at least a threshold number of final-configuration reports, replacing the model partition with two new model partitions that represent the two parameter-value subspaces so that, over time, the comprehensive data structure contains increasingly specific models for smaller and more specific parameter-value subspaces.

### Claim 19 (depends on 12)

The CTQ estimator of claim 12 wherein the comprehensive data structure is a tree-like data structure with model partitions as leaf nodes and internal nodes containing expressions that evaluate to a Boolean value when evaluated using a parameter value selected from a set of parameter values.

### Claim 20

A physical data-storage device that contains a set of computer instructions that, when executed by one or more processors of one or more computer systems, each having one or more memories and one or more mass-storage device, control the one or more computer systems to:
receive a request for a list of component types and quantities for installation of a system, the received request including a set of parameter values;
use the set of parameter values to traverse a comprehensive data structure to identify a model that represents a class of system installation characterized by the set of parameter values;
for each component type for which a component model is included in the model,
using the component model for the component type and a polynomial expression associated with the component type to generate an estimate of the quantity of components of the component type needed for the system installation;
include the estimated component quantities generated from the component models in a list of component types and quantities; and
return the list of component types and quantities to the requestor entity.