A Robotic Mobility Diversity Algorithm With Markovian Trajectory Planners
Daniel Bonilla Licea, Des Mclernon, Mounir Ghogho

In this paper we develop an intelligent algorithm that obtains the optimal trajectory (i.e., a non equally spaced sequence of stopping points) for a robot which tries to find a wireless channel with a minimum predefined channel gain over which to transmit its data. We show that this algorithm can be optimized in two ways: (i) minimum searching time (but suboptimal energy expenditure), or (ii) minimum energy (but not necessarily minimum searching time).
This problem can be viewed as similar to classical RF selection diversity (but with an infinite number of diversity branches). However, it is different in so far as here we seek either highly or poorly correlated branches (i.e., the wireless channels at stopping points) depending upon the realisation at the last stopping point(s). We show that this strategy is superior (both in searching time and energy expenditure) when compared with a classical diversity approach for devising the robot"s trajectory.