Multipath Exploitation and Knowledge Based Urban Radar Imaging Using Compressive Sensing
Abstract
The primary objective of this effort is to develop appropriate signal models and algorithms for behind-the-wall stationary and moving target localization and building layout mapping within the CS framework. A secondary objective is to enable enhanced imaging and detection of low-signature buried targets in forward-looking ground penetrating radar applications. To this end, we have made the following main contributions: 1) Mitigation of clutter and stationary targets through conversion of populated scenes to sparse scenes based on Moving Target Indication techniques; 2) Effective front wall clutter suppression under reduced data volume for detection of stationary targets behind walls; 3) Exploitation of the rich multipath nature of the indoor environment in conjunction with compressive sensing for improved stationary target detection and localization in sparse scene scenarios; 4) Enhanced detection and tracking of moving targets through multipath exploitation; 5) Utilization of prior information of building construction practices for determining the building layout under compressive sensing; 6) Robust multipath exploitation based target localization approaches through dictionary learning in the presence of inaccuracies in knowledge of building layout; 7) Effective reconstruction of target scene using a distributed network of through-the-wall radar units in the presence of multipath; 8) Exploitation of target spatial extent for high-resolution through-the-wall radar imaging under the sparse reconstruction framework; 9) Multi-view target detection schemes for forward-looking ground penetrating radar operating at close-to grazing angles; and 10) Coherence factor based rough surface clutter suppression in forward-looking ground penetrating radar applications.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 23, 2016
- Accession Number
- AD1058562
Entities
People
- Fauzia Ahmad
- Moeness Amin
Organizations
- Villanova University