Skip to Main Content
Mobility has a significant impact on the ad hoc network protocol performance. Mainly, the protocol performance has been evaluated in simulations and using synthetic mobility models, which have two main drawbacks: (a) they assume that wireless devices start and remain in the simulation for a user defined simulation time; and, (b) they are unrealistic. Real mobility models that are derived from real user traces challenge the assumption that wireless devices start and remain in the simulation for the entire user defined simulation time, but they rather show that wireless nodes posses dynamic membership (nodes join and leave the simulation dynamically based on some random variable). In this paper, we evaluate the maximum node degree mobility metric for real mobility models, which has got little attention due to the assumption of static connectivity graph on the number of nodes. The contributions of this paper are two-fold. First, we introduce the algorithm that computes the maximum node degree mobility metric, which in turn provides the upper bound on the number of neighbors for a given node. Second, we show that the upper bound can be used to improve the algorithm complexity by introducing a new algorithm metric, namely efficiency. Its usefulness is shown through a case study for evaluating the gains in the algorithm complexity of incentive protocols.