Towards real time multidimensional Hierarchical Graph Neuron (mHGN)

The Hierarchical Graph Neuron (HGN) has already been known that, it implements a single-cycle memorization and recall operation. The scheme also utilizes small response time that is insensitive to the increases in the number of stored patterns. In this improved approach, the architecture of multidimensional HGN (mHGN) is developed so, that it is not only suitable for scrutinizing 1D- or 2D-patterns; it can also work on multidimensional patterns. As the result, the mHGN architecture is able to recognize multidimensional patterns. Some approaches usually need a lot of training cycles and require complex deployment, when patterns need to be memorized and recalled. The multidimensional capabilities of mHGN can recognize not only concrete but also abstract patterns in real time manner. The mHGN architecture still encompasses a lightweight in-network processing algorithm which does not require expensive floating point computations; hence, it is still suitable for real-time applications.