Randomized AB-Face-AB Routing Algorithms in Mobile Ad Hoc Networks

One common design for routing protocols in mobile ad hoc networks is to use positioning information. We combine the class of randomized position-based routing strategies called AB (Above-Below) algorithms with face routing to form AB:Face2:AB routing algorithms, a new class of hybrid routing algorithms in mobile ad hoc networks. Our experiments on unit disk graphs, and their associated Yao sub-graphs and Gabriel sub-graphs, show that the delivery rates of the AB:Face2:AB algorithms are significantly better than either class of routing algorithms alone when routing is subject to a threshold count beyond which the packet is dropped. The best delivery rates were obtained on the Yao subgraph. With the appropriate choice of threshold, on non-planar graphs, the delivery rates are equivalent to those of face routing (with no threshold) while, on average, discovering paths to their destinations that are several times shorter.