Solving inverse source problems for linear PDEs using sparse sensor measurements

Many physical phenomena across several applications can be described by partial differential equations (PDEs). In these applications, sensors collect sparse samples of the resulting phenomena with the aim of detecting its cause/source, using some intelligent data analysis tools on the samples. These problems are commonly referred to as inverse source problems. This work presents a novel framework for solving such inverse source problem for linear PDEs by drawing from certain recent results in modern sampling theory. Under the new framework, we study the well-known diffusion PDE and present numerical results that highlight the validity and robustness of the approach.