Microcontroller-based implementation of a data extraction algorithm from a radar signal

Precise and low-cost localization is fundamental in a variety of applications. The focus of the current research is on augmented reality for multitude of applications from gaming to training of personnel in various industries. In such applications, one key requirement is that of identifying a precise position on a given object, possibly in a complex scenario, like a cluttered environment, with a lot of background noise (one example could be that of identifying a precise location on a jet engine when training a maintenance operator with augmented reality). A typical approach when performing localization is that of using trilateration, starting from range information that has been collected through three or more radiofrequency receivers. The process can be facilitated by using an active target, that generates a unique return signal used to unambiguously identify the target. A further choice that must be made is the selection of the considered radiofrequency transceiver. To minimize the implementation cost, a low-complexity chirp RADAR (i.e., a Frequency Modulated Continuous Wave RADAR), can be used to derive the range information. Having to identify the signal reflected by a target in a cluttered environment, the system uses an active target equipped with a small and low-cost transponder that generates, in response to the RADAR signal, a bandpass signal that can be separated from the clutter which is generally located around baseband in the down-converted RADAR signal. It is therefore necessary to have a specific demodulator that recovers this bandpass signal generated by the active target and processes the signal to generate the range information needed for trilateration. Given this general scenario, the thesis work has been focused on the study, simulation, and implementation of the demodulator able to recover the bandpass signal generated in the down-converted RADAR signal by the active target in response to the RADAR chirp signal. More specifically, this thesis aims to implement a data extraction algorithm from a RADAR signal on a microcontroller for target localization. The data extraction algorithm is based on an envelope detector; by demodulating the received signal from a chirp RADAR, the target range can be calculated starting from the received signal's phase. Then, the target can be precisely localized by using the trilateration method which consists of combining the target ranges calculated from three receivers' signals. The thesis outlines the theory behind this implementation, presents the simulation results and details the system implementation. In particular, in Chapter 2 the trilateration method is described and in Chapter 3 the characteristics of a FMCW RADAR with chirp modulation are presented. In Chapter 4 the block diagram for the data extraction algorithm is introduced and the individual blocks are described, with a particular focus on FIR filters which are at the basis of this implementation. In Chapter 5 the hardware implementation on a TI C2000 F28379D microprocessor is described. In Chapter 6 the algorithm performances are analyzed. Conclusions are drawn in Chapter 7.