Dynamics of coupled mode solitons in bursting neural networks

Abstract

Using an electrically coupled chain of Hindmarsh-Rose neural models, we analytically derived the nonlinearly coupled complex Ginzburg-Landau equations. This is realized by superimposing the lower and upper cutoff modes of wave propagation and by employing the multiple scale expansions in the semidiscrete approximation. We explore the modified Hirota method to analytically obtain the bright-bright pulse soliton solutions of our nonlinearly coupled equations. With these bright solitons as initial conditions of our numerical scheme, and knowing that electrical signals are the basis of information transfer in the nervous system, it is found that prior to collisions at the boundaries of the network, neural information is purely conveyed by bisolitons at lower cutoff mode. After collision, the bisolitons are completely annihilated and neural information is now relayed by the upper cutoff mode via the propagation of plane waves. It is also shown that the linear gain of the system is inextricably linked to the complex physiological mechanisms of ion mobility, since the speeds and spatial profiles of the coupled nerve impulses vary with the gain. A linear stability analysis performed on the coupled system mainly confirms the instability of plane waves in the neural network, with a glaring example of the transition of weak plane waves into a dark soliton and then static kinks. Numerical simulations have confirmed the annihilation phenomenon subsequent to collision in neural systems. They equally showed that the symmetry breaking of the pulse solution of the system leaves in the network static internal modes, sometime referred to as Goldstone modes.

On the possibility of rogue wave generation based on the dynamics of modified Burridge-Knopoff model of earthquake fault

“…We now focus on high degree of oceanic inter plate coupling at subduction zones in which 𝛾 ≠ 0 [Nkomom et al, 2021;Akishin et al, 2000;Pelap et al, 2016]. Consequently we consider small amplitude oscillation wave solution to equation ( 10), by employing the multiple-scale method [Remoissenet, 1986;Nfor et al, 2021;Nfor, 2021;Nfor et al, 2018;Nfor and Mokoli, 2016]. We initially introduce the change of variable…”

Modulational Instability and Discrete Localized Modes in Two Coupled Atomic Chains with Next-Nearest-Neighbor Interactions

“…Enormous effort has been made to obtain analytical solutions to a number of discrete nonlinear partial differential equations. Some prominent analytical methods includes: quadrature method, ansatz method, method of separation of variables, Painlevé method, Hirota method, IST method, B äcklund transform method, Darboux transformation method among others [11,18,51,[53][54][55][56]. Furthermore, a more powerful analytical method like the discrete generalized (m, N − m)-fold Darboux transformation technique; have already been developed to obtain two-component localized wave solutions for a reduced semi-discrete two-component coupled system on a two-chain lattice [57].…”

“…Since the discovery of solitary waves by John Scott Russell, solitons have been experimentally observed in many spatially discrete physical systems, such as granular materials, nonlinear Hamiltonian lattices [10], electrical transmission lines, and neural networks [11][12][13][14]. The recurrence phenomenon was originally observed in the Fermi-Pasta-Ulam lattice [15,16].…”

Dynamics of interval Cohen‐Grossberg neural networks with time‐varying delays based on LMI computation

“…The dynamics of different classes of neural networks have been widely studied in recent years. [1][2][3][4][5][6][7][8][9][10][11][12][13] For the Cohen-Grossberg neural networks:…”

“…A good neural network is expected to be robust against such uncertainties. The dynamics of different classes of neural networks have been widely studied in recent years 1–13 …”