Module implementations

Module implementations describe the (real-world) hardware components (i.e., harvester, load, converter, or capacitor) and can - for example - contain simple mathematical representations of (ideal) components, complex analytical models, or data sets from experimental campaigns. To define and use a certain module implementation, the module-specific interfaces have to be implemented and the desired module has to be configured accordingly.

Available module implementations

Simba provides a number of pre-defined module implementations which can be configured individually and thus already cover a large number of different battery-free systems.

Module	ModuleTypes
Harvesters	Artificial, IVCurve, SolarPanel, TEG
Converter	Diode (converter-less), BuckBoost, LDO, BQ25570
Capacitor	IdealCapacitor, TantulumCapacitor
Load	ConstantLoad, TaskLoad, JITLoad, JITLoadAdvanced

Configuration

To select the desired module implementation, each module (i.e., Harvester, Converter, Capacitor etc.) can be configured using a dedicated configuration (file) in .json format as follows:

{
  "type": "<ModuleType>",
  "settings" : "<ModuleSettings>"
}

<ModuleType> is the name of the subtype/implementation of the module and <ModuleSettings>  contain module-specific settings. For module-specific settings, check out the description of each module implementation.

Module interfaces

In order to create a new Simba-compatible module implementation, a number of methods have to implemented accordingly. This include common methods that are called for all modules (e.g., reset, update_state etc.) and module-specific methods that are only required for a certain type of module (e.g., get_input_efficiency is certainly only required for the converter module).

In the following, we briefly describe the mandatory common and module-specific methods for each module type.

Common methods

• reset(): Initialize the module (e.g., load required data, set up variables) and set initial state.
• update_state(): Update the module's internal state and log data if required - for more module-specific information, see below.
• process_log(): Called at the end of the simulation to process logging data from module implementation (if required).
• get_log(): Retrieve (detailed) logging data from module implementation as Pandas DataFrame datatype.
• get_stats(): Derive module-specific statistics from logging data and provide this data as a dict data type (for more details, refer to Trade-off exploration).
• get_next_change(): Inform the simulation core of the next expected update within the module (i.e., returns the time until next update in seconds) - for more module-specific information, see below.

Harvester-specific methods

The Harvester modules represent energy sources that supply the sensor node with incoming power.

Common methods

• update_state(time, dt, v_in): Update state if necessary (i.e., if environmental conditions have changed and a different harvesting current applies) and log (actual and maximal) harvesting current and applied voltage at given time.
• get_next_change(time): Inform the simulation core when the harvesting conditions change.

Harvester-specific methods (mandatory)

• get_current(time, v_in): Return the harvesting current (in A) at the specified voltage at the current time.
• get_ocv(time): Return open-circuit voltage of harvester at the current time.

Harvester-specific methods (optional)

• plot_iv_curve(), plot_irradiance(), ... : Harvester-specific plot functions to support users while using Simba.

Converter-specific methods

The Converter modules represent the voltage converter(s) between harvester and capacitors ("input") as well as capacitor and load ("output").

Common methods

• update_state(time, dt, cap_voltage): Update state if necessary (i.e., turn-on/turn-off thresholds have been reached) and log converter-specific values (e.g., efficiencies, quiescent current, state) at given time.
• get_next_change(time): Inform the simulation core when the converter changes its state (e.g., when the next MPP sampling takes place).

Converter-specific methods (mandatory)

• get_input_operating_voltage(v_cap): Return (operating) voltage of converter between harvester and capacitor.
• get_input_efficiency(v_cap, i_in): Return efficiency of converter between harvester and capacitor.
• get_output_operating_voltage(v_cap): Return output voltage of converter between capacitor and load.
• get_output_efficiency(v_cap, i_out): Return efficiency of converter between capacitor and load.
• get_quiescent(v_cap): Return quiescent current (in A) depending on voltage level.
• get_next_threshold(v_cap, i_total): If any, return the next voltage threshold at which the converter will change its state.

Capacitor-specific methods

The Capacitor modules describe the energy buffer between harvester and load.

Common methods

• update_state(time, dt, i_total): Update the capacitor's state (i.e., state of charge) according to the current flow and the capacitor's leakage and log capacitor voltage (and leakage) at the given time.
• get_next_change(i_total, voltage_threshold): Inform the simulation core when the given voltage threshold (e.g., supplied by the load or converter) would be reached, given the actual current.

Capacitor-specific methods (mandatory)

• get_voltage(): Provide the current voltage of the capacitor.

Load-specific methods

The Load modules represent the load (e.g., sensor node) to be powered by the harvester and capacitor. Note that the load module must include/model the power consumption of both the MCU and any peripherals attached to it.

Common methods

• update_state(time, dt, v_out, v_cap): Update the load's internal state (e.g., currently running task etc.) accordingly and log any load-specific values (e.g., state, voltages, current consumption) at the given time.
• get_next_change(time): Inform the simulation core when the load changes its state (e.g., when the next task is scheduled).
• get_next_threshold(v_cap, i_total): If any, return the next voltage threshold at which the load will change its state.

Load-specific methods (mandatory)

• get_current(v_out): Return the load's current consumption in the current state.
• get_state(): Return the load's current internal state (for logging purposes only).