Simultaneous Localization and Mapping (SLAM) is the conventional chicken and egg problem where a robot has to map the environment as well as localize itself in the newly created map simultaneously. One of the most used approaches to solving this problem is probabilistic robotics. Some of the most common SLAM solutions using probabilistic methods are Extended Kalman Filter, Sparse Extended Information Filter, and Particle Filter. Since single-robot SLAM is mostly a solved problem, researchers are focusing on multi-robot SLAM to improve the efficiency and speed of map exploration. Multi-robot SLAM can be used in tasks where collaboration can improve performance and create a more accurate map. Applications of multi-robot SLAM includes fire fighting in urban and forest areas, rescue and cleaning operations and underwater and space exploration. Most of the published multi-robot SLAM articles use robots equipped with a sensor to detect range and bearing. Range-only SLAM faces issues because it lacks bearing knowledge, which makes it difficult for the robot to create a transformation matrix needed for the robot to merge maps obtained from other robots. This research explores the problem of multi-robot SLAM using range sensors. A Received Signal Strength Indicator (RSSI) sensor can be used to detect the landmark and another robot. RSSI is a measure of power received from radio signals. Radio beacons can be used as landmarks which helps to remove the data association problem for landmarks as each radio beacons have unique media access control (MAC) address in the network. This thesis proposed an approach to multi-robot range-only SLAM using full particle method.
