[Cluster associative memory formation in a three-layer network of phase oscillators]

A three-layer network model of oscillatory associative mermory is proposed. The network is capable to store binary images that can be retrieved if an appropriate stimulus has been applied. Binary images are encoded in the form of the spatial distribution of oscillatory phase clusters in-phase (+1) and anti-phase relative to the base periodic signal. The information is loaded into the network using a set of interlayer connection weights. A condition for error-free pattern retrieval has been obtained, which imposes a certain limitation on the maximal number of patterns to be stored in the memory (storage capacity). It has been shown that the capacity can be significantly increased by the generation of optimal pattern alphabet (basic pattern set). The number of stored patterns can reach values of the network size (the number of oscillators in the layer), which is significantly higher than the capacity of traditional oscillatory memory models. The dynamical and information characteristics of the retrieval process based on the optimal alphabet including the estimations of attraction basins and the admissible input pattern discrepancy for error-free retrieval have been investigated.