The PhD investigations aim at exploring new WBAN cooperative localization mechanisms, which could benefit jointly from on-body links, body-to-body links between distinct mobile users or off-body links with respect to the infrastructure. Following a multidisciplinary approach, we have thus addressed theoretical questions related to physical modeling or to algorithmic and cross-layer design. A few more practical aspects have also been dealt with. More specifically, based on WBAN channel measurements, single-link ranging error models are first discussed for more realistic performance assessment. Then a Constrained Distributed Weighted Multi-Dimensional Scaling (CDWMDS) positioning algorithm is put forward for relative MoCap purposes, coping with on-body nodes' asynchronism to reduce system latency and exploiting the presence of constant-length radio links for better accuracy. Subsequently we consider extending this algorithm for larger-scale asbolute MoCap applications within a 2-step localization approach that incorporates additional off-body links in a heterogeneous WBAN framework. Then, both individual and collective kinds of navigation are addressed. In both MoCap and navigation scenarios, low-complexity solutions exploiting on-body deployment diversity enable to combat error propagation and strong range biases due to body shadowing, relying on on-body nodes' dispersion or graph neighborhood to approximate the corrupted distances. Finally, experiments based on real IR-UWB radio platforms validate in part the previous proposals, while showing their practical limitations.
