In many applications, modelling is a crucial step. The process of developing a mathematical description for a particular natural event is referred to as modelling (also called system). Price prediction in banking [85], channel estimation in communication systems [88], and controller design in industrial processes [80] are some examples.Modeling may be divided into two categories. The first method considers modelling a natural event using physical rules (for example, modelling an electrical circuit using Kirchhoff's current and voltage laws). The second method is to describe the natural phenomena as a black box model (or a grey box model if some physical understanding is taken into consideration) and identify the model using input-output data from a system experiment (e.g., provide a voltage excitation to an electrical circuit, and measure the current to obtain a model for the equivalent impedance). The second method, often referred to as system identification, is the focus of this thesis.The input sequence we give to excite the system isone of the most important aspects of system identification. Because system identification requires input-output data, it's important to create an input excitation that, in certain ways, optimises the information in the experiment (e.g., optimising a cost function related to the intended model application). Input design refers to the process of creating an input sequence for system identification, which is the subject of this thesis.The words input design and experiment design are differentiated. The specification of input and output signals, measurement instants, modified signals, and how to modify them are all options in experiment design (which is the focus of input design). Signal conditioning is also included (e.g., the choice of presampling filters [59, Paper 13]).