Convex Optimization for Compressed Sensor Target Tracking

Solution Overview

Problem

Current sensor and data processing systems for analyzing object locations and motions are bulky, costly, and power-intensive due to the need for large data collection and processing, especially when using uncompressed data, which can be mitigated by compressive sensing but still require significant reconstruction efforts.

Innovation Solution

The method involves representing physical attributes of objects in a path space associated with compressively sensed data, using convex optimization to select points likely to correspond to observed data without full reconstruction, thereby reducing computational complexity and enabling distributed processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If compressive sensing is used to collect only limited samples, then the size, cost, and power consumption of sensors are reduced, but the data processing burden increases substantially during reconstruction

Engineering Contradiction:

Improvepower consumption of sensorVSAvoiddata processing burden

Core Design Contradiction:

Use of energy by moving objectVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by performing convex optimization directly on the compressed sensing measurements before reconstruction. The method formulates the target extraction problem as a convex optimization task that operates on the compressed domain data, avoiding the need for full image reconstruction. This preliminary processing of compressed data reduces the subsequent computational burden while maintaining extraction accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent extracts only the essential actionable information (target presence, location, motion) directly from the compressed sensing measurements without performing full image reconstruction. By taking out only the necessary information elements needed for the application, the system avoids the computational overhead of reconstructing the entire image while still obtaining the required targets.

Inventive Principle:

Principle #2Taking out (Extraction)

2Loss of information

If full reconstruction of compressed data is performed to extract actionable knowledge, then complete image information is obtained, but the computational complexity and processing time increase significantly

Engineering Contradiction:

Improveimage information completenessVSAvoidprocessing time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The patent extracts only the essential actionable information (target presence, location, motion) directly from the compressed sensing measurements without performing full image reconstruction. By taking out only the necessary information elements needed for the application, the system avoids the computational overhead of reconstructing the entire image while still obtaining the required targets.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies partial action by performing convex optimization on a subset of the compressed measurements that are most relevant for target extraction. Rather than processing all compressed data for full reconstruction, the method selectively processes only the portions necessary for detecting and characterizing targets, reducing computational time while maintaining extraction accuracy.

Inventive Principle:

Principle #16Partial or excessive action

3Measurement precision

If a large number of computations are performed on collected sensor data, then actionable knowledge is extracted accurately, but the system becomes bulky and costly

Engineering Contradiction:

Improvetarget detection accuracyVSAvoidsystem size and cost

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by performing convex optimization directly on the compressed sensing measurements before reconstruction. The method formulates the target extraction problem as a convex optimization task that operates on the compressed domain data, avoiding the need for full image reconstruction. This preliminary processing of compressed data reduces the subsequent computational burden while maintaining extraction accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces the traditional mechanical approach of collecting full-resolution data and then processing it with a substitute method: performing convex optimization directly on compressed measurements. This substitution changes the processing paradigm from reconstruction-based to optimization-based extraction, reducing computational requirements and system complexity while maintaining detection accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS11726197B2Systems and methods for efficient targeting

Publication Date: 2023.08.15 QUALCOMM INC

AI summary

A system for determining the physical path of an object can map several candidate paths to a suitable path space that can be explored using a convex optimization technique. The optimization technique may take advantage of the typical sparsity of the path space and can identify a likely physical path using a function of sensor observation as constraints. A track of an object can also be determined using a track model and a convex optimization technique.