Modeling of artificial semiconductor quantum neural network from first principles

Position-based qubits in the chain of coupled semiconductor quantum dots controlled electrostatically are remaining one of the most prominent ways of implementation of compact quantum computer [1-2]. We formulate N input and M ouput quantum neural network with k hidden layers. We model propagation of electron wavepacket and utlize its functionality. We derive generalized tight-bidning model from Schroedinger formalism of N perturbatively interacting particles and find fundamental analytical solutions governing the dynamics of wavepackets in quantum neural network. Furthermore we train neural network in direction of implementation certain logical operations using numerical quasi one-dimensional approximation by usage of properties of electrostatically coupled single-electron lines [3-5].

Literature

[1]. T. Fujisawa et al. , Time-dependent single-electron transport through quantum dots,

Reports on Progress in Physics, Volume 69, Number 3, 2006

[2]. A.R.Mills et al., Shuttling a single charge across a one-dimensional array of silicon quantum dots,

Nature Communications, Vol. 10, 1063, 2019.

[3]. K.Pomorski, Electrostatically interacting Wannier qubits in curved space, Arxiv: 2112.13470 ,2021

[4]. K.Pomorski, Equivalence between finite state stochastic machine, non-dissipative and dissipative tight-binding and Schroedinger model, Arxiv: 2208.09758, 2022.

[5]. K.Pomorski et al., Analytic view on coupled single-electron lines, Semicond. Sci. Technol., Vol. 34, 125015, 2019.