A multi-objective performance evaluation method for edge-cloud collaborative video crowd sensing

Through storage transmission and complexity measurement, a performance loss model is built, and the cost function is constructed using RCcost and L(D) to construct a cost function, the multi-objective performance optimization of the intelligent computing system is achieved, the problems of homogeneous parameters and normalized measurement of the end-edge cloud system are solved, and the multi-objective performance optimization of the intelligent computing system is achieved.

CN118337639BActive Publication Date: 2025-03-18HANGZHOU DIANZI UNIV
2 Cites 0 Cited by

Patent Information

Application Number
CN202410592683.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-14
Publication Date
2025-03-18
Estimated Expiration
2044-05-14

AI Technical Summary

Technical Problem

The existing technology is difficult to realize homogeneous measurement of heterogeneous parameters and normalized measurements under the end-edge cloud system, which makes it difficult to achieve multi-objective performance optimization of intelligent computing systems.

Method used

Through storage transmission and complexity measurement, a performance loss model is built, and the cost function is constructed using RCcost and L(D) to conduct multi-objective performance evaluation to realize homogeneous parameters and normalized measurement under the end-edge cloud system.

Benefits of technology

It realizes multi-objective performance optimization of intelligent computing systems, supports unified measurement of heterogeneous parameters such as equipment cost, power consumption cost, storage cost, and transmission cost, and provides unified measurement and evaluation support for end-edge cloud collaborative processing.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present invention discloses a multi-objective performance evaluation method for edge-cloud collaborative video group intelligent perception, which measures the cost of bitstream data transmission based on bitrate, measures the cost of bitstream data storage based on bitrate and corresponding storage architecture, and measures the complexity cost based on the consumption of hardware resources and energy resources; homogenizes the measurement of storage, transmission and complexity costs based on the consumed costs to obtain parameters of unified dimension measurement; converts the static hardware cost of calculation and storage into the hardware cost of calculation and storage required for each camera video; converts the dynamic energy consumption cost of calculation and storage into the power consumption cost required for encoding and analyzing each camera video; converts the transmission and storage cost into the cost required for each video transmission based on the broadband network fee, and all costs are measured based on currency; constructs the performance loss of a machine vision algorithm based on the multi-objective performance under distortion and distortion-free conditions, constructs a cost function through the performance loss and the parameters, and performs multi-objective performance evaluation based on the merits of performance under different parameters.
Need to check novelty before this filing date? Find Prior Art