Optical implementation of the Hopfield neural network using multiple fiber nets

Area of Science:

• Optoelectronics
• Artificial Neural Networks
• Optical Computing

Background:

• Hopfield networks are a type of recurrent neural network used for associative memory.
• Traditional implementations can be complex and lack efficient global connectivity.
• Optical systems offer potential for high-speed parallel processing.

Purpose of the Study:

• To propose and experimentally demonstrate an associative memory system based on the Hopfield model.
• To utilize multiple optical fiber nets for interconnections, representing synaptic weights.
• To leverage the advantages of optical fiber networks for achieving global connections and efficient matrix-vector multiplication.

Main Methods:

• Developing an associative memory model using the Hopfield network architecture.
• Implementing interconnections with multiple optical fiber nets, where coupling ratios act as synaptic weights.
• Designing an experimental setup for a 5x5 binary pattern associative memory.
• Storing three distinct binary patterns within the optical fiber neural network.

Main Results:

• Successful experimental demonstration of the proposed optical fiber neural network.
• Achieved global connections between 2-D units, enabling simultaneous weight matrix and input vector multiplication.
• Stored three 5x5 binary patterns and successfully retrieved them from the associative memory.

Conclusions:

• Optical fiber neural networks provide an effective platform for implementing associative memory based on the Hopfield model.
• The proposed method allows for easy achievement of global connections and efficient parallel processing.
• This demonstrates a viable optical approach for advanced neural network applications.