Short-time memories in a network with randomly distributed connections.

Area of Science:

• Complex systems
• Nonlinear dynamics
• Network science

Background:

• Coupled map lattices (CMLs) exhibit complex behaviors and have been investigated for their potential in information processing.
• External periodic forcing is known to influence the dynamics of CMLs, potentially enabling memory-like properties.

Purpose of the Study:

• To investigate short-term memory formation in coupled map lattices under periodic external forcing.
• To analyze how network connectivity (regular vs. random) affects memory storage.
• To determine the parameter regimes (coupling and nonlinearity) that support single or multiple memory patterns.

Main Methods:

• Simulations of coupled map lattices with nearest-neighbor and random connection topologies.
• Analysis of system dynamics under varying coupling strengths and nonlinear parameters.
• Characterization of stored patterns and memory stability.

Main Results:

• Short-term memory storage is demonstrated in CMLs with periodic forcing.
• Network structure significantly influences memory capacity and stability.
• Specific ranges of coupling and nonlinear parameters are identified for storing distinct patterns.

Conclusions:

• Coupled map lattices are capable of forming short-term memories.
• The interplay between network topology and system parameters dictates the ability to store and retrieve information.
• This work provides insights into the design principles for CMLs as potential memory devices.