Low-Complexity Source Localization Based on Quaternion Analysis in Smart Ocean

With the proliferation of marine activities, underwater Internet of Things (UIoT), which integrates various techniques for supporting smart ocean, has attracted more research interest. The trend is that one desires to use computationally efficient and widely applicable algorithms for source localization. For fulfilling the above requirements, this paper proposes a novel unitary quaternion (UQ) model, which is applied to widespread centro-symmetric arrays. The estimation and decomposition of the corresponding covariance matrix can be executed in the real number field, thus benefiting from low complexity. Moreover, we analyze the physical implication of the proposed model and associate it with the emerging quaternion-based attitude estimation and control, which reveals the potential advantages of the UQ model in UIoT. In the simulations, we test the algorithm performance in several realistic underwater scenarios, demonstrating the flexibility and applicability of the UQ model.