mna.h File Reference

Defines functions to perform voltage stimulation on a Nanowire Network using the Modified Nodal Analysis (MNA) algorithm. More...

#include "device/network.h"
#include "device/component.h"
#include "interface/interface.h"
#include "interface/mea.h"

Go to the source code of this file.

Functions
int	voltage_stimulation (network_state ns, const connected_component cc, const interface it, double io[])
	Perform the voltage stimulation of the Nanowire Network by using the Modified Nodal Analysis algorithm. It does not update the conductance value of the network, basically ignoring its plasticity and treating it as a static device. The voltage distribution is stored in the vector V of the ns parameter. More...
int	voltage_stimulation_mea (network_state ns, const connected_component cc, const MEA mea, double io[])
	Perform the voltage stimulation of a Nanowire Network connected with a MEA. See voltage_stimulation for more details. More...

Detailed Description

Defines functions to perform voltage stimulation on a Nanowire Network using the Modified Nodal Analysis (MNA) algorithm.

This file defines the functions required to perform voltage stimulation of a Nanowire Network through the Modified Nodal Analysis (MNA) algorithm. The MNA algorithm is used to calculate the voltage distribution across the network without considering its plasticity, effectively treating it as a static device. The voltage distribution is stored in the vector V of the ns parameter.

The file defines two different functions for the stimulation of the Nanowire Network: one using an interface, and one using a MEA. The MEA implementation simply converts the MEA into an interface before passing it to the main function.

Note

The MNA implementation leverages UMFPACK functions for efficient computation.

Function Documentation

voltage_stimulation()

int voltage_stimulation	(	network_state	ns,
		const connected_component	cc,
		const interface	it,
		double	io[]
	)

Perform the voltage stimulation of the Nanowire Network by using the Modified Nodal Analysis algorithm. It does not update the conductance value of the network, basically ignoring its plasticity and treating it as a static device. The voltage distribution is stored in the vector V of the ns parameter.

Note

This function uses the UMFPACK functions to perform efficient calculations. Previously, some algorithms for systems solution were studied with the aim to find the most efficient one. The considered factorization methods are the following:

• backward substitution -> not applicable as the elements on the bottom right of the diagonal are != 0
• cholesky -> not applicable as the matrix is not positive-defined
• LU decomposition -> costs n^3
• Gauss-Jordan elimination -> costs n^3

Parameters

[in,out]	ns	The Nanowire Network equivalent electrical circuit on which performing the MNA. Only the voltage value of the nodes belonging to the passed CC will be modified.
[in]	cc	The connected component of ns to stimulate.
[in]	it	The interface of the Nanowire Network with the external world, including sources, grounds and loads.
[in,out]	io	An array with an entry for each source. As input parameter it contains the voltage applied to a source, as output it contains the current drawn from that node.

Returns

0 if the computation successfully terminates, -1 if an error occurs (e.g. if the sources/grounds/loads nanowires are not connected).

voltage_stimulation_mea()

int voltage_stimulation_mea	(	network_state	ns,
		const connected_component	cc,
		const MEA	mea,
		double	io[]
	)

Perform the voltage stimulation of a Nanowire Network connected with a MEA. See voltage_stimulation for more details.

Parameters

[in,out]	ns	The Nanowire Network equivalent electrical circuit on which performing the MNA. Only the voltage value of the nodes belonging to the passed CC will be modified.
[in]	cc	The connected component of ns to stimulate.
[in]	mea	The mea-interface of the Nanowire Network with the external world, including sources, grounds and loads.
[in,out]	io	An array with an entry for each source. As input parameter it contains the voltage applied to a source, as output it contains the current drawn from that node.

Returns

0 if the computation successfully terminates, -1 if an error occurs (e.g. if the sources/grounds/loads nanowires are not connected).