Innovative tools and technologies for orbital angular momentum based Quantum Information protocols

High-dimensional quantum states (qudits) represent a pivotal resource in Quantum Information protocols. Indeed, by exploiting such states, it is possible to increase the amount of information exchanged between parties and to enhance the security of cryptographic schemes. Therefore, several platforms and devices have been proposed and implemented to produce and manipulate qudits. Among these, photonic platforms are particularly interesting since photons represent an ideal candidate for storing and transmitting information. Several degrees of freedom are employed for the information encoding, among them, we focused on the two components of light angular momentum, namely the Spin Angular Momentum (SAM) and the Orbital Angular Momentum (OAM). The former is related to the polarization, therefore, being bidimensional is used to codify qubits, while the latter is related to the field spatial structure and is unbounded, so it is suitable to encode high dimension quantum states. In this thesis work, we investigated the generation, engineering, and detection of OAM states, both proposing and demonstrating innovative approaches. Then, after addressing the main problems arising when working with this degree of freedom, we employed it in two quantum information protocols, namely a quantum machine learning method for state reconstruction and a quantum simulation scheme. We started our study from the sources of OAM states, focusing first on the direct generation of high dimensional entangled states through a Spontaneous Parametric Down Conversion (SPDC) process. Here, we certified the emitted OAM states, introducing a holographic method that outperformed the standard projective measurement approaches, both in time and resources needed for the reconstruction. Then, we addressed the problems connected with the probabilistic nature of SPDC sources, by proposing a platform interfacing a nearly deterministic Quantum Dot (QD) with OAM manipulation devices, in which we demonstrate the production of entangled high dimensional states in this degree of freedom. Thereafter, we investigate the manipulation of the photons state in the engineering of arbitrary qudits and the retrieval of the information stored in them. On the one hand, we optimized an experimental OAM-based QW photonic engineering platform through a black-box approach. Indeed, the latter, not needing to be imbued with a description of the setup, automatically accounts for experimental imperfections, resulting in high values for the fidelities of the engineered states. On the other hand, in our setup, we both enhanced the performance of a well-known measurement approach, namely the holographic techniques, through refined modeling of the beam propagation and demonstrated innovative machine learning-based reconstruction techniques, capable of reducing the losses present in interferometric and diffractive schemes. Finally, as we said, we make use of the potential connected with a high dimensional encoding inside two quantum information processing protocols. We first identified the OAM-based QW dynamics as the reservoir of a quantum extreme learning paradigm, introducing an estimation method for qubits properties that is effective and resource-efficient. Then, we exploited the connection between the QW evolution and the one-dimensional Dirac equation, and, by codifying the particle position in the OAM of single photons, we simulated the Zitterbewegung effect, i.e the trembling motion occurs during the free evolution of relativistic particles. Both implementations are prelaminar results, but they showcase promising performance in the application of the approaches to higher dimensional states and in the photonic simulation of more complex dynamics.