Nested Hash Tables for Fast Simulation Data Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current simulation systems for training moving assets, such as aircraft, face challenges in efficiently storing and replaying complex simulation data, particularly for long-duration simulations, as the volume of data exceeds what can be reasonably stored in random access memory (RAM), requiring slower persistent storage and involving sequential data reading to restore an accurate state.
Innovation Solution
The use of nested hash tables in RAM to quickly locate and retrieve raw binary data stored in persistent storage, allowing for fast or slow playback, non-contiguous time jumping, and reconstruction of simulation states by mapping points in time to data indices, enabling efficient data retrieval and presentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If simulation data is stored in persistent storage for long-duration simulations, then data storage capacity is sufficient, but data retrieval speed deteriorates due to sequential reading requirements
Solution Approach 1:
The patent segments simulation data into discrete event records with unique identifiers, storing them in persistent storage as individual units. This segmentation allows the system to retrieve only specific events needed for playback rather than reading entire files sequentially, thus maintaining fast retrieval speed while accommodating large volumes of simulation data.
Solution Approach 2:
The patent introduces an intermediary indexing system that maps simulation time points to event record locations in persistent storage. This intermediary layer enables direct access to specific events without sequential scanning, resolving the contradiction between storing large quantities of data and retrieving them quickly.
2Speed
If simulation parameters are recorded sequentially in RAM, then data access speed is fast, but storage capacity is insufficient for complex long-duration simulations
Solution Approach 1:
The patent extracts only the essential event records and their metadata from the full simulation data, storing these compact representations in RAM for fast access. The complete simulation data remains in persistent storage, thus achieving fast access speed for critical information while maintaining sufficient storage capacity for the entire simulation dataset.
Solution Approach 2:
The patent transitions from storing all simulation data in a single memory space to a multi-dimensional storage architecture where frequently accessed event indexes reside in RAM while the full dataset resides in persistent storage. This dimensional separation allows the system to overcome RAM capacity limitations while maintaining fast access to critical simulation events.
3Quantity of substance
If simulation data is stored in persistent storage, then storage capacity is sufficient, but system complexity increases due to coordination between RAM and persistent storage
Solution Approach 1:
The patent performs preliminary organization of simulation data into standardized event records with unique identifiers and time stamps before storage. This pre-organization simplifies the coordination between RAM and persistent storage by establishing a consistent data structure that requires minimal complex processing during playback operations, thus reducing system complexity while maintaining adequate storage capacity.
Data Source
AI summary
An apparatus is provided for simulation of an asset to train a user to use the asset. The apparatus includes persistent storage to store raw binary data transformable into messages associated with simulation of the asset, and a random access memory (RAM) to store a protocol name, sender identifier and message type of the messages in respective hash tables. These hash tables further include a fourth hash table that maps times of the messages to the raw binary data in the persistent storage. In response to a request for playback of messages from a selected time, the apparatus locates the raw binary data for a message chronologically last in time before the selected time using the four nested hash tables in the RAM, retrieves the raw binary data for the message from its location in the persistent storage medium, and transforms the raw binary data into the message for presentation.