A system for anchor-less self-positioning and simultaneous 2D estimation of rectangular room dimensions is proposed. The system is based on narrowband transceiver capable of simultaneous transmission and reception. The transmission is generated at omnidirectional antenna, while eight sectorized antennas with 45deg apertures in horizontal are used for reception. The full duplex property enables measurements of spatially discriminated channel impulse response (CIR) at the same place where the signal is originated. Hence, by using radar approach, the presence of walls and corners can be evaluated based on round trip time (RTT) measurements. The novel contribution of this paper includes position estimation confidence indicator defined as instantaneous received power of the CIR??s strongest element. Additionally, it is assumed that a transceiver is a part of an ??intelligent?? mobile robot platform capable of moving, updating its position, and keeping track of past positions along with the corresponding measurement data. We propose a new mobility model, focused on efficient exploration of an unknown environment by moving towards the walls according to the corresponding confidence indicators values. Simulations inside narrowband synthetic test environment are performed and the performances of localization and room dimension estimates, as a function of number of observations, are evaluated. Index Terms ?? full duplex antennas, channel impulse response, indoor, mapping, mobility model, radar approach estimation, ray tracing, rectangular rooms, round trip time, simultaneous localization and mapping, sectorized narrowband antennas, selfpositioning, spatial CIR discrimination, transceiver localization
展开▼
