Abstract
This thesis explores magnetic anomaly detection schemes for two-dimensional fixed arrays of magnetic sensors employed to discern the presence of mobile objects. Large ferrous bodies generate a magnetic dipole, which causes an anomaly within the Earth's magnetic field. The large number of sensor signals from a fixed array can be combined in order to improve detection performance. Included here is a presentation of signal processing schemes for combining such measurements for the purpose of drawing inferences about the presence of a magnetic dipole. The Neyman-Pearson lemma is used to construct a detector for a mobile dipole as a likelihood ratio test under a constant false alarm rate criteria. Receiver operating characteristic curves summarize the performance of the detector system. Presented is a framework for illuminating the impact of the number of sensors comprising the system, the spacing between adjacent sensors and the ambient magnetic noise power present within the surrounding environment.