Computational Storage Inference Decoys for Abnormal Video Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The inefficiency in managing storage spaces for multimedia data and the vulnerability of data security due to large data sizes and potential data leakage through abnormal access in computational storage systems.
Innovation Solution
A computational storage system utilizing neural network models to generate fake inference video data in response to abnormal access, efficiently managing storage space and preventing data leakage by generating secure alternative data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multimedia data is stored in storage devices, then data storage capacity is increased, but storage space management efficiency deteriorates
Solution Approach 1:
The patent segments multimedia data into multiple versions with different quality levels (original quality, reduced quality, thumbnail). This segmentation allows the system to store and manage different representations of the same data, enabling efficient retrieval and usage based on specific needs without managing the entire large-volume original data for every operation.
Solution Approach 2:
The patent applies parameter changes by creating multiple versions of multimedia data with different quality parameters (resolution, file size, compression level). This allows the system to optimize storage space by storing lower-quality versions for common operations while preserving original quality data only when needed, thus improving storage management efficiency while maintaining data storage capacity.
2Quantity of substance
If multimedia data is stored in storage devices, then data storage capacity is increased, but data security deteriorates
Solution Approach 1:
The patent applies local quality by providing different quality levels of data to different users or access scenarios. When abnormal access is detected, the system provides lower-quality or degraded versions of multimedia data to the intruder, while maintaining access to high-quality original data for authorized users. This local differentiation of data quality based on access context enhances data security without sacrificing storage capacity.
Solution Approach 2:
The patent converts the potential harm of data theft into a benefit by detecting abnormal access patterns and responding by providing degraded-quality data to intruders. The system transforms the security threat into an opportunity to identify and respond to unauthorized access, using the data quality differentiation as a security mechanism that protects the integrity and confidentiality of the original multimedia data.
3Reliability
If fake inference video data is generated using neural network models, then data security is improved, but device complexity is increased
Solution Approach 1:
The patent uses copying by creating synthetic versions of multimedia data through neural network models. When abnormal access is detected, the system generates fake inference video data that mimics the appearance of real multimedia content but contains no actual sensitive information. This copying approach enhances data security by providing plausible-looking data to intruders while protecting the original data, without requiring fundamentally new system architectures.
Solution Approach 2:
The patent replaces traditional mechanical security measures (such as physical access controls or simple encryption) with intelligent systems based on neural networks. The system uses machine learning models to detect abnormal access patterns and generate synthetic data dynamically, substituting complex computational processes for simpler but less effective security mechanisms. This substitution improves data security through intelligent detection and response while managing system complexity through software-based solutions.
Data Source
AI summary
A computational storage system includes a storage device storing a plurality of neural network models as well as base data and an event table both corresponding to each of the plurality of neural network models, and a computing device configured to generate inference video data corresponding to original video data, based on the plurality of neural network models, the base data, and the event table, wherein the computing device is further configured to, when an abnormal access to the computational storage system is identified, generate fake inference video data based on a security level of the abnormal access and output the generated fake inference video data, and wherein the security level indicates a data leakage path of the abnormal access.