Research on Indoor Positioning Algorithm based on Hybrid Sparrow Algorithm Improved Back Propagation Neural Network

Indoor positioning using Bluetooth’s received signal strength indicator has gained significant attention and widespread usage due to its affordable cost, low power usage, and easy deployment. Nonetheless, conventional indoor positioning algorithms commonly experience issues, including building occlusion and multipath effects in functional contexts, leading to inadequate indoor positioning results. To enhance positioning accuracy, this paper presents an Hybrid Sparrow Algorithm(HSSA), which optimises the indoor positioning algorithm of Back Propagation(BP) neural network. The algorithm generates the initial population of sparrows via Circle chaotic mapping and incorporates the Levy flight strategy, expanding the search space exploration range to improve global optimisation. The optimised algorithm is employed to establish the HSSA-BP localization model, enabling position prediction The results of the simulation illustrate that the HSSA-BP algorithm generates an average absolute error of merely 0.02m, which is notably better than the BP algorithm’s 0.12m and the SSA-BP algorithm’s 0.10m. The positioning accuracy has therefore been improved considerably.