A software-hardware joint optimization method and system for communication in a liquid state machine model
A liquid state machine and joint optimization technology, applied in computer components, design optimization/simulation, biological neural network models, etc., can solve problems such as hardware function errors and low simulation efficiency, reduce data packet transmission delay, and ease communication Pressure, the effect of ensuring accuracy
Active Publication Date: 2022-05-03
NAT UNIV OF DEFENSE TECH
View PDF7 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The synchronization period in TrueNorth is 1ms. There is a problem with this fixed time synchronization method, that is, if the synchronization time is set too long, the hardware simulation efficiency will be very low, and most of the hardware will be idle; if the synchronization time is set If the value is too short, then this will cause the function of the hardware to malfunction
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0039] Such as image 3 As shown, the combined software and hardware optimization method for communication in the liquid state machine model of this embodiment includes:
[0040] 1) Initialize the liquid state machine model LSM in the SNN simulator, and set the initial pulse packet loss rate;
[0041] 2) In the SNN simulator, the liquid state machine model LSM is trained and reasoned, and the pulses transmitted in the liquid state machine model LSM are lost according to the pulse packet loss rate during the training and reasoning process, and finally after the training and Calculate the classification accuracy after reasoning;
[0042] 3) Judging whether the classification accuracy rate meets the requirements, if the classification accuracy rate does not meet the requirements, then jump to step 6); otherwise, jump to step 4);
[0043] 4) Map the liquid state machine model LSM to the on-chip network, simulate the communication between the liquid layer neurons in the liquid st...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a software and hardware joint optimization method and system for communication in a liquid state machine model. The invention includes training and reasoning of the liquid state machine model LSM in a pulse neural network (SNN) simulator. During the training and reasoning process According to the pulse packet loss rate, the pulse transmitted in the liquid state machine model LSM is lost, and the classification accuracy rate is calculated after training and reasoning are completed; if the classification accuracy rate meets the requirements, the liquid state machine model LSM is mapped to the network on chip Simulate and calculate the maximum transmission delay on the simulator. If the maximum transmission delay meets the requirements, increase the pulse packet loss rate and continue to iterate until the best pulse packet loss rate is found. On the premise of ensuring the classification accuracy, the present invention continuously loses packets through the determined pulse packet loss rate, so as to reduce the data packet transmission delay between the cores of the brain-inspired processor, so that the final LSM meets the class requirements when running in the brain-inspired processor. Brain computing real-time requirements.
Description
technical field [0001] The invention relates to a method for optimizing the communication efficiency of a liquid state machine network when it is executed on a network-on-chip brain-inspired computing platform, in particular to a method and system for joint software and hardware optimization of communication in a liquid state machine model. Background technique [0002] Spiking neural network (SNN), also known as the third-generation neural network, is an emerging discipline formed by the intersection of computer science and biological neuroscience. Compared with the traditional artificial neural network (ANN), SNN has achieved a higher level of biological neurosimulation, and it is easier to simulate the low-power, high-performance processing of the human brain. The spiking neural network has the following characteristics: the neurons in it are simulated biological neurons; when the membrane voltage of the neurons accumulates to a certain level, a Spike (pulse) will be emit...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F30/27G06N3/04G06N3/063G06K9/62G06F111/08
CPCG06F30/27G06N3/049G06N3/063G06F2111/08G06F18/24G06F18/214
Inventor 王蕾王世英康子扬曲连华李石明刘威张剑锋石伟龚锐冯权友张英励楠
Owner NAT UNIV OF DEFENSE TECH