Audio Transfer Channel Isolation for Deterministic Safety Audio
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio systems face issues with deterministic low latency transfer of safety audio, as hang situations can occur due to delays or stops in non-safety audio sources, impacting occupant safety in vehicles.
Innovation Solution
The system employs separate memory operations and channels for safety and non-safety audio sources, ensuring independent and deterministic low latency transfer of safety audio by using direct memory access (DMA) engines and stream descriptor registers to manage audio sample blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If blocks of audio samples from multiple audio sources are transferred sequentially in the same memory channel, then device complexity is reduced, but reliability of safety audio transfer deteriorates due to hang situations caused by non-safety audio sources
Solution Approach 1:
The patent divides the memory transfer system into separate channels: a first memory channel dedicated to safety audio sources and a second memory channel for non-safety audio sources. This segmentation ensures that hang situations in non-safety audio transfers do not affect safety audio transfer reliability, while maintaining manageable device complexity through structured separation.
Solution Approach 2:
The patent extracts safety audio transfer operations from the general audio transfer pool by creating a dedicated memory channel and DMA engine for safety audio. This extraction isolates safety audio from potential hang situations caused by non-safety audio sources, ensuring deterministic low-latency transfer without being affected by other audio sources' performance issues.
2Reliability
If separate memory operations and channels are used for safety and non-safety audio sources, then reliability of safety audio transfer is improved, but device complexity increases
Solution Approach 1:
The patent implements segmentation by creating distinct memory channels and DMA engines for safety and non-safety audio. This structured division improves reliability through isolation while controlling complexity through systematic organization of the separate transfer paths.
Solution Approach 2:
The patent employs universal DMA engines that can handle both safety and non-safety audio transfers, with the first DMA engine capable of operating in both dedicated and shared modes. This multi-functionality reduces device complexity by reusing hardware resources while maintaining the reliability benefits of separate transfer paths.
3Productivity
If audio samples from multiple sources are transferred in the same memory operation, then productivity is improved, but latency of safety audio transfer increases due to sequential dependencies
Solution Approach 1:
The patent segments audio transfers into parallel memory operations on separate channels, allowing safety audio and non-safety audio to be transferred simultaneously without sequential dependencies. This eliminates waiting time for safety audio while maintaining overall system productivity through efficient parallel processing.
Solution Approach 2:
The patent adds a temporal dimension to audio transfer by enabling parallel operations across multiple memory channels. Instead of transferring audio samples in a single sequential dimension, the system utilizes multiple simultaneous transfer paths, reducing safety audio latency while maintaining high overall productivity.
4Device complexity
If a single DMA engine handles all audio sources, then device complexity is reduced, but deterministic low latency transfer of safety audio cannot be guaranteed
Solution Approach 1:
The patent segments the DMA engine functionality into a first DMA engine dedicated to safety audio and a second DMA engine for non-safety audio. This segmentation guarantees deterministic low-latency transfer for safety audio by isolating it from the variable performance of non-safety audio transfers, while maintaining manageable device complexity through specialized hardware design.
Solution Approach 2:
The patent configures the first DMA engine with pre-allocated memory channels and dedicated resources for safety audio transfers before any audio playback occurs. This preliminary setup ensures that safety audio transfer paths are ready and deterministic, eliminating latency variability without requiring complex runtime resource management.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Transferring audio samples in an audio system comprises receiving respective streams of audio samples at a same time from a safety audio source and non-safety audio sources, where a sampling frequency of the audio samples of the safety audio source is the same as the sampling frequency of the audio samples of the non-safety audio sources. A determination is made whether the audio samples which are received is from a safety audio source or non-safety audio sources. A first memory operation to transfer the audio samples from the safety audio source to an audio processor in a first memory channel and a second memory operation to transfer the audio samples from the non-safety audio sources to the audio processor in a second memory channel are performed based on the determination.