Data Generator library is a Java library offering tools for device simulation suitable for IoT testing environments or as independent data sources. Device emulation supports consists of various types such as Sensors and Actuators, each represents different roles and responsibilities with different data generation methods.

This library provides two methods of data generation:


generated data are returned to the user by calling the requestData(params) method on Device. This allows the user to get data on demand with custom parameters. The number and types of parameters for this method depends mostly on the DataFeed implementation provided for the Device.

Continuous data delivery

Is achieved by wrapping instance of Device with ActiveDevice object. In this case the requestData() method is called periodically by ActiveDevice. This composition can be considered as a one, standalone device.

For data manipulation, the user is advised to perform the following additional steps:

  1. Configure a NetworkAdapter to enable data handling. (This is done the same way as in real-life applications.) This way data can be sent to valid endpoint using the various network protocols like REST, MQTT and many more.

  2. Enable CoAPControlServer from Controller package to allow ad-hoc configuration of remote devices. To access CoAP server created by particular devices use simple tool provided by the library: CoAPControlClient, or any other CoAP client available on the market.

  3. For testing purposes, it is recommended to create an Observer. Using the Observer pattern, the user is able to catch important information from the data creation process without using debugging tools or viewing log messages. For information on deploying.

Users can utilize three main data generators types:


interface adds possibility to define List of DataFeed. This provides possibility to simulate data generation for multiple independent data sources. For example commonly known DHT11 temperature-humidity sensor is working with two values simultaneously. This requires possibility to define multiple DataFeeds for one device.


extends Sensor and restrict its functionality to use only one DataFeed as it is convenient to sometimes utilize sensor that is monitoring only one value.


different type of Device than Sensor. Offers methods to simulate real life behaviour of actuators. Important part of this concept is StateMachine.class which holds information about current state and time duration, that is needed for actuator to finish simulated action. Handles any transitions between states if necessary.

The library provides few preconfigured devices that are ready for use. The idea behind them is to simplify the creation process for the user in case he wants to create commonly known devices but to offer him freedom in configuring his own.

Sensors Actuators










The packages currently offered are the following:


(Sensor, SingleSensor, Actuator) back and bone of library, brings parts of the library together. Gives meaning for generated data which are acquired from individual data feeds. Holds information like UUID, label, and description.


tool to provide data streams to different components (e.g. time simulation, simulated devices). This interface is ready for further user customization, as it is not possible to define all situations that could be needed in the future. So far the library provides examples, that can be divided into three parts: SSJ (Stochastic Simulation in Java) - Java library providing a tool for stochastic data simulation developed at Université de Montréal ( Expression approach, where a user defines an expression (function) which solutions represents expected data stream “Logic” driven approach. A user is fully capable to customize data generating class as long as it implements DataFeed interface


provides a collection of classes that allow the user to manipulate with configuration of existing devices, therefore influence ongoing processes. Thanks to usage of CoAP [1], user is able to interact with local or remote devices. For this purpose, library supports both server and client CoAP perspectives.


provides facilities that enrich functionality of Device with possibility to transfer generated data over the network to desired destination. Package implements support of protocols such as HTTP and MQTT.


Principle behind creation of simulated device is simple. First, the user has to determinate desired properties of simulated device and choose corresponding or best fitting representation offered by Data Generator library. Next step is to correctly choose or configure method for data simulation itself (in case of Sensor it would be the right choice of`DataFeed` and for Actuator it would be correct configuration of StateMachine). Finally, the last step (or steps) is to assign all required functionality that is offered by the library to the device.

As we can see, the functionality of the library is achieved by putting together right pieces. === Sensor example

On demand data pooling

A base for our device is properly configured data feed. For this example

DataFeed dataFeed = new NormalDistVariateDataFeed(30, 7);

Create Device with DataFeed. In this case, simple Sensor will be enough.

SimpleSensor device = new Thermometer("ThermometerExample", dataFeed);

Get generated data


Continuous data delivery - ActiveDevice

Now, we know how to set up a simple device. Define device as we did in previous examples In order to simulate device in time - to be able to send data periodically in set intervals, we need to wrap it with running environment called time simulation Like device, ActiveDevice also needs some data feed to be able to determinate mentioned time intervals

DataFeed timeFeed = new LinearDataFeed(2000);

We create an instance of ActiveDevice to simulate our device in time.

ActiveDevice active = new ActiveDeviceImpl(timeFeed, device);

At this point, the device needs to configure. Finally, we start the simulation.


Actuator example

The behaviour of an Actuator is specified by the StateMachine configuration.

Example of state machine of two states and two transitions. Note that, input event is not specified, as there is only one possible transition from each state (input event is generated and determinated automatically byt the state machine):

new StateMachine.Builder()


new StateMachine.Builder()
            .to("Extended", "extend", "extending", 10)
            .to("Retracted", "retract", "retracting", 10)

Finally, assign created state machine to Actuator instance:

Actuator actuator = new BasicActuator("actuator")

Add additional capabilities for Device

To register device to CoAP server simply call method:


Send generated data to desired destination using Network Adapter component:

NetworkAdapter na = new Rest(httpEndpoint, new JSONWrapper());