A strain-sensing based scheme for indoor localization: Analysis, algorithm, and demonstration

In this work, a strain-based positioning problem is analyzed for indoor localization applications. This inverse mapping problem for estimating force-applied location by a few known strain data is formulated and analyzed for a square plate structure. Finite element (FE) simulations are performed to obtain the correlation between the force-applied location and the output index, which is a combination of measurable strain-gauge signals. Due to nonlinear coupling between major axes, the effective zone, which guarantees localization with sufficient accuracy is poor and must be enhanced. A novel iterative scheme is then proposed and implemented, aiming to improve the resolution and the effective zone. By such an effort, the spatial resolution is improved, and the effective zone is enhanced based on the simulation results. Meanwhile, a prototype of localization system is also constructed as the first step toward experimental demonstrations. The test results agree with the theoretical prediction qualitatively and the effective zone indeed can be further improved from 39% to 57% by the iterative algorithm based on a ±5% spatial uncertainty. The lessons and conclusion learned from this work serve as the basis for the 2D smart-floor tile designs currently underway for smart building applications.