NCR: A unifying parameter to characterize and evaluate data dissemination scenarios and its analytic study

Nowadays data dissemination often happens in vehicular sensor networks (VSN) and other mobile ad hoc networks in military & surveillance scenarios. The performance of data dissemination depends on many different parameters including speed, motion pattern, node density, topology, data rate, and transmission range. This multitude makes it difficult to accurately evaluate and compare data gathering protocols implemented in different simulation or testbed scenarios. In this paper, we introduce Neighborhood Change Rate (NCR), a unifying measurement for different motion patterns used in epidemic dissemination, a contact-based data dissemination. By its intrinsic property, the NCR measurement is able to describe the spatial and temporal dependencies and well characterize a dissemination / harvesting scenario. We illustrate our approach by applying the NCR concept to MobEyes, a lightweight data gathering protocol. We further analytically study the effective NCR for Markov type motion models, such as Real Track mobility model. A closed-form expression has been derived. From this analytic solution, the NCR can be approximated from the initial scenario settings, such as velocity range, transmission range, and real map/street information. The closed-form formula for NCR can be further employed to evaluate the ED process. The mathematical relationship between the dissemination index and the effective NCR is established and it allows predicting the performance of the ED process in realistic track motion scenarios. The experiment results showed that the analytic expressions for the NCR and for the evaluation of the ED process closely match the discrete-event simulations