Multi-Stream Spike Cross Correlation Without Buffering Delays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cross correlation methods suffer from high latency and require buffering to perform bidirectional searches, which is inefficient and introduces delays in processing streaming data.
Innovation Solution
The proposed solution involves a multi-stream cross correlator (MCC) that uses a Cross Correlation Unit (CCU) with two main units: Produce Delay and Learn Delay. These units operate independently and utilize the stochastic nature of spike streams to produce a correlated spike stream with reduced memory requirements, eliminating the need for buffering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If established cross correlation methods are used to perform bidirectional search, then measurement precision is improved, but loss of time increases due to high latency and buffering requirements
Solution Approach 1:
The patent applies preliminary action by pre-processing the input signal to extract features and characteristics that are essential for cross-correlation analysis. This pre-processing step prepares the data in advance, allowing the correlation algorithm to work more efficiently with reduced computational requirements, thereby reducing latency without sacrificing accuracy
Solution Approach 2:
The patent extracts only the necessary components and features from the input signal that are relevant for cross-correlation, rather than processing the entire signal. This extraction approach reduces the amount of data that needs to be buffered and processed, directly addressing the latency issue while maintaining measurement precision
2Measurement precision
If memory buffer is used for bidirectional search, then measurement precision is improved, but device complexity increases
Solution Approach 1:
By performing preliminary processing and feature extraction on the input signal, the patent reduces the amount of data that needs to be stored in memory buffers. This pre-preparation of data allows the system to achieve accurate cross-correlation with minimal buffering, thereby reducing device complexity while maintaining measurement precision
Solution Approach 2:
The patent uses simplified representations or copies of the signal data rather than storing the complete original signal in buffers. By working with extracted features and condensed data representations, the system achieves the necessary correlation accuracy with reduced memory requirements and lower device complexity
3Measurement precision
If many comparisons are performed for cross correlation, then measurement precision is improved, but productivity decreases due to computational expense
Solution Approach 1:
The patent extracts and focuses only on the most relevant signal features and time points for cross-correlation analysis, eliminating the need to perform comparisons on unnecessary data. This selective approach maintains measurement precision by concentrating computational effort on critical elements while significantly improving processing speed and productivity
Solution Approach 2:
The patent applies partial action by performing cross-correlation comparisons only at strategically selected time points and offsets rather than exhaustively checking all possible combinations. This selective sampling approach achieves sufficient measurement precision with far fewer computations, thereby improving productivity and processing efficiency
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A multi-stream cross correlator for spiking neural networks, where each stream contains significant stochastic content. At least one event occurs, with a fixed temporal relationship across at least two streams. Each stream is treated as a Frame Of Reference (FOR), and subject to an adjustable delay based on comparison to the Other streams. For each spike of the FOR, a timing analysis, relative to the last and current FOR spikes, is completed by comparing Post and Pre accumulators. Also, a new timing analysis is begun, with the current FOR spike, by restarting the production of Post and Pre weighting functions, the values of which are accumulated, upon the occurrence of each Other spike, until a next FOR spike. A one-spike delay unit can be used, if time-neutral conflict resolution is used. The average spike rate of the FOR can be determined and used for the Post and Pre weighting functions.