Optimal Stochastic Location Updates in Mobile Ad Hoc Networks - 2011

ABSTRACT:

We tend to consider the location service in a very mobile ad-hoc network (MANET), where every node wants to maintain its location data by one) frequently updating its location info within its neighboring region, that is named neighborhood update (NU), and a couple of) sometimes updating its location information to bound distributed location server within the network, which is termed location server update (LSU). The trade off between the operation prices in location updates and also the performance losses of the target application thanks to location inaccuracies (i.e., application prices) imposes a crucial question for nodes to make a decision the optimal strategy to update their location info, where the optimality is in the sense of minimizing the overall costs. In this paper, we develop a stochastic sequential call framework to investigate this downside. Underneath a Markovian mobility model, the situation update call problem is modeled as a Markov Decision Process (MDP). We first investigate the monotonicity properties of optimal NU and LSU operations with respect to location inaccuracies under a general cost setting. Then, given a separable cost structure, we tend to show that the placement update choices of NU and LSU will be independently allotted while not loss of optimality, i.e., a separation property. From the discovered separation property of the problem structure and therefore the monotonicity properties of optimal actions, we have a tendency to notice that one) there continually exists a simple optimal threshold-based update rule for LSU operations; two) for NU operations, an optimal threshold-based mostly update rule exists in a very low-mobility scenario. In the case that no a priori information of the MDP model is obtainable, we additionally introduce a sensible model-free learning approach to search out a near-optimal solution for the matter.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here