Can we solve finite-horizon Markov decision processes (FHMDPs) while raising low memory requirements? Such models find application in many cases where a decision-making agent needs to act in a probabilistic environment, from resource management to medicine to service provisioning. However, computing optimal policies such an agent should follow by dynamic programming value iteration raises either prohibitive space complexity, or, in reverse, non-scalable time complexity requirements. This scalability question has been largely neglected. In this paper, we propose SIFTER (Space Efficient Finite Horizon MDPs), a suite of algorithms that achieve a golden middle between space and time requirements. Our former algorithm raises space complexity growing with the square root of the horizon's length without a time-complexity overhead, while the latter's space requirements depend only logarithmically in horizon length with a corresponding logarithmic time complexity overhead. A thorough experimental study under diverse settings confirms that SIFTER algorithms achieve the predicted gains, while approximation techniques do not achieve the same combination of time efficiency, space efficiency, and result quality.

The paper presents a new two-step cellular location determination (CLD) method based on signal strength and wave scattering models. The received signal level (RSL) method is first used in combination with maximum likelihood estimation (MLE) and triangulation to obtain an estimate of the location of the mobile. Due to non line of sight (NLOS) conditions and multipath propagation, this estimate lacks acceptable accuracy and consistency for demanding services, as numerical simulations reveal. Thus, the wave scattering 3D multipath channel model of Aulin is employed together with extended Kalman filtering (EKF) to obtain improved location estimates with high accuracy. The EKF is initialized at the MLE obtained from the RSL method, which is proved to be highly appropriate. Numerical simulations under urban, suburban and rural environments were utilized to evaluate the accuracy and consistency of the proposed two-step enhanced RSL method; the results of the worst-case rural environment are presented.