Application of inverse problem algorithms for integrated circuit temperature estimation

This article presents an approach to the application of inverse problem alg orithms for integrated circuit temperature estimation. In order to estimate circuit temperature from remote sensor measurements, it is necessary to so lve an inverse heat conduction problem (IHCP). As the IHCPs are extremely s ensitive to input data errors, special algorithms must be used to obtain ro bust estimates of circuit temperature. Computer simulations have been perfo rmed applying different algorithms to produce circuit temperature estimates in case of multiple heat sources. Simulation results are presented. Additi onally, the possibility of digital filter algorithm implementation is discu ssed. (C) 1999 Elsevier Science Ltd. All rights reserved.