People nowadays spend a tremendous amount of time in indoor structures, causing them to be known as the “indoor generation.” For example, Americans spend, on average, 90% of their time indoors. In light of this, the demand for accurate indoor navigation and localization systems has been more than ever before. Not only accurate indoor navigation enables emerging applications, e.g., location-based services (LBS), it is vital for public safety, e.g., first responders. Numerous competing approaches have been proposed over the past couple of decades for indoor navigation and localization; however, there is no single technology that has emerged as a clear winner in solving this problem.

Among all approaches, cellular long-term evolution (LTE)-based approaches are particularly attractive as they are infrastructure-free, and if properly exploited, can lead to a practical, affordable, and accurate localization system. This is due to the inherent desirable characteristics of LTE signals: abundance, geometric diversity, high bandwidth (up to 20 MHz), high received power (carrier-to-noise ratio (CNR) ranges between 50 and 80 dB-Hz in different indoor conditions), and some of their downlink signals are free to use. Exploiting LTE-based signals for indoor localization comes with several challenges: (i) specialized receivers to opportunistically extract navigation observables from received LTE signals must be designed, and (ii) the clock biases of LTE base stations (also known as evolved Node Bs or eNodeBs) must be removed or estimated, and (iii) errors resulting from short-delay multipath must be mitigated. This Thesis addresses the aforementioned challenges. The foundational contributions of this thesis are demonstrated in showing meter-level accurate indoor navigation.

The problem of an autonomous surface vehicle (ASV) optimally planning measurement locations to localize a set of pre-deployed acoustic underwater transponders (UTs) is considered. The ASV is assumed to make noisy range measurements to the UTs. A maximum a posteriori estimator is derived to localize the UTs. In addition, a multi-step look-ahead (MSLA) ASV optimal measurement location planning (OMLP) strategy is developed. This planning strategy prescribes future multi-step measurement locations. A physical interpretation of the proposed planner in the single-step, single transponder case is provided. Simulation results demonstrate the trade-off between expected localization performance and computational time associated with various look-ahead horizons and travel distances. Experimental results illustrate the proposed MSLA OMLP strategy's performance in environments containing one and two UTs. The proposed OMLP strategy is able to localize UTs to within 4 meters of their true locations. Additionally, increasing the planning horizon is demonstrated to yield better UT localization at the cost of increased computational burden.