The Random Waypoint Model
We have seen in the previous chapter that a major shortcoming of random walk models is the lack of a fundamental aspect of intentional mobility, namely, the notion of trajectory of a movement. Indeed, random walks are aimed at mimicking completely random movement patterns, and as such they can be considered as relatively accurate in describing mobility patterns where mobile entities display such highly random behavior.
Are there many next generation wireless network application scenarios where node mobility can be classified as “highly random?” Unfortunately (or luckily, depending on the viewpoint), the answer is no: apart from a few wireless sensor network scenarios (e.g., animal tracking in open environments, in which mobility can be faithfully modeled by Lévy flights), in the vast majority of next generation wireless network scenarios node mobility is not random, but obeys some (often strict) rules.
The random waypoint model (RWP), first defined by Johnson and Maltz (1996) to study the performance of a routing protocol for MANETs, can be considered as a first attempt to define a simple, synthetic mobility model aimed at modeling intentional human movement. Thus, from the viewpoint of modeling accuracy, and despite its several shortcomings as described in this chapter, the RWP model can be considered a breakthrough in next generation wireless network mobility modeling compared to random walks. This, combined with the popularity of the Dynamic ...