Spike Neural Network Encoder for Quantized Time-Series Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in efficiently converting continuous time-series signals into input signals for spike neural networks, which are essential for applications like risk recognition, security vigilance, and autonomous driving, due to high power consumption and inefficiencies in signal processing.
Innovation Solution
An encoder that samples and quantizes continuous time-series signals, selects discrete quantum signals based on a quantum level threshold, and activates input neuron circuits to generate spike signals, ensuring efficient signal transmission to the spike neural network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If continuous time-series signals are converted into input signals for spike neural networks using existing technologies, then signal processing can be performed, but power consumption is high and processing efficiency is low
Solution Approach 1:
The encoder divides the continuous time-series signal into multiple discrete quantum signals through sampling and quantization. Each discrete quantum signal is then independently processed and transmitted to corresponding input neuron circuits, enabling parallel processing and reducing overall processing time while lowering power consumption
Solution Approach 2:
The encoder uses periodic sampling to convert continuous signals into discrete time-series quantum signals at regular intervals. This periodic action transforms the continuous signal processing task into discrete, manageable units that can be efficiently processed by the spike neural network with lower power consumption
2Loss of information
If all discrete quantum signals are transmitted to input neuron circuits, then complete signal information is preserved, but processing complexity and power consumption increase
Solution Approach 1:
The encoder extracts only the essential features from discrete quantum signals by comparing each signal with a reference level. Only signals that exceed the reference level (positive spikes) are selected for transmission to input neuron circuits, removing redundant information and simplifying processing while preserving critical signal characteristics
Solution Approach 2:
The encoder applies different processing rules to different portions of the signal based on their relationship to the reference level. Positive spikes are transmitted as active signals, negative spikes are suppressed, and zero-level signals are ignored, creating a localized quality transformation that reduces complexity while maintaining information integrity
Data Source
AI summary
Disclosed is operation method of an encoder that receives a continuous time-series signal and respectively transmits first to N-th input signals to first to N-th input neuron circuits of spike neural network circuit. The method of operating the encoder includes receiving the continuous time-series signal, generating a plurality of discrete quantum signals by sampling and quantizing the continuous time-series signal, selecting first to N-th discrete quantum signals among the plurality of discrete quantum signals, matching the selected first to N-th discrete quantum signals with the first to N-th input neuron circuits, respectively, identifying discrete quantum signals, each of which has a quantum level different from a quantum level of a previous discrete quantum signal, from among the second to N-th discrete quantum signals, and activating the input signals to be transmitted to the input neuron circuits corresponding to the identified discrete quantum signals and the first discrete quantum signal.