Application of inverse problem algorithms for integrated circuit temperature estimation

This article presents an approach to the application of inverse problem algorithms for integrated circuit temperature estimation. In order to estimate circuit temperature from remote sensor measurements, it is necessary to solve an inverse heat conduction problem (IHCP). As the IHCPs are extremely sensitive to input data errors, special algorithms must be used to obtain robust estimates of circuit temperature. Computer simulations have been performed applying different algorithms to produce circuit temperature estimates in case of multiple heat sources. Simulation results are presented. Additionally, the possibility of digital filter algorithm implementation is discussed.