Non-Cooperative MUSIC Localization in Indoor Environment
Ehsan Akbari Sekehravani, Angela Dell’Aversano, Raffaele Solimene, Rosa Scapaticci · IEEE Access · 2026
Localizing non-cooperative electromagnetic sources in indoor environments is a challenging inverse problem, primarily due to multipath propagation and the lack of accurate knowledge of the Green’s function. Although subspace-based methods such as Multiple Signal Classification (MUSIC) offer high resolution under ideal modeling assumptions, their performance can severely degrade in realistic indoor scenarios affected by the mismatch between the actual propagation environment and the assumed model. This paper proposes a robust single-frequency localization strategy based on Multiplicative Array Fusion MUSIC (MAF–MUSIC), a distributed processing approach that partitions the receiving array into multiple subarrays, processes each subarray independently, and combines the resulting pseudospectra through multiplicative fusion. The proposed formulation exploits spatial diversity across subarrays to reinforce source-related features that are consistent across independent views, while suppressing spurious peaks arising from multipath and modeling errors. The 2D numerical results demonstrate that MAF–MUSIC effectively mitigates the effects of Green’s function mismatch compared to standard MUSIC, yielding improved localization accuracy and reduced false alarms in complex indoor environments.