The demand for high-precision, real-time, and continuous localization in
aerospace and low-altitude unmanned systems is steadily increasing. However,
under weak GNSS availability, obstruction, and degraded ranging geometry,
conventional 3D localization methods often suffer from insufficient observation
constraints and severe accuracy degradation. To address observability
degradation and localization error divergence in single-auxiliary-node systems,
this paper proposes a 3D cooperative localization method based on maneuver
reconfiguration of a single auxiliary node. Within a particle filtering
framework, the method fuses target motion increments with single-shot range
measurements to recursively estimate the target’s 3D position and heading. A
geometric degradation metric and a radial-consistency metric are further
constructed from the line-of-sight distribution and the alignment between target
velocity and line of sight within a sliding window, enabling online detection of
degraded observability. Once degradation is detected, the auxiliary node
performs either lateral maneuvering or circular maneuvering to actively reshape
the target–auxiliary relative geometry and restore observation diversity;
meanwhile, the two metrics can also serve as feedback variables for adaptive
maneuver adjustment. Comparative analyses are conducted under three modes,
namely no reconfiguration, lateral maneuvering, and circular maneuvering, from
the perspectives of 3D trajectory evolution, error time series,
degradation-metric time series, and statistical error performance. Parameter
sensitivity analyses and an additional nonlinear velocity-varying scenario are
also included. Experimental results show that the no-reconfiguration mode
exhibits clear error divergence under degraded conditions, whereas both
maneuver-based strategies effectively reduce localization errors. Lateral
maneuvering is more suitable for rapid recovery in the early degradation stage,
while circular maneuvering achieves better overall accuracy and stronger
post-degradation recovery capability. These results demonstrate the
effectiveness and engineering feasibility of degradation-triggered active
geometric reconfiguration for cooperative localization in low-configuration
unmanned systems.