Mobile ad hoc networks (MANETs) are anticipated to play an important role in the future of networking, making the performance analysis of algorithms and protocols for MANETs an important topic.
The mobility of the nodes and the characteristics of these wireless networks, such as a frequently changing topology, unreliable communication links, and a limited transmission range, require the development of algorithms and protocols performing efficiently under such circumstances.
Measuring the performance of algorithms and protocols for MANETs is complicated by the costs associated with conducting a real-world experiment and the conditions of such an experiment not being easily repeatable. Therefore, as with wired networks, network simulation is common practice for evaluating the performance of such a system.
Traditional network simulators designed for simulating wired networks lack the support necessary for mobile networks, like modeling node movement and spatial constraints (e.g. propagation obstacles). In the past, some of these network simulators have been adapted or newly developed to also support wireless networks. However, few of these support the modeling of spatial constraints. Also, state-of-the-art network simulators generally have a high resource consumption and lack user interactivity, modularity, and adaptability. Therefore, we have developed Cubus, a network simulator fully capable of simulating MANETs and supporting the aforementioned additional features.
The development of the system was split across two theses: In the first thesis, the core system was developed. The support of node movement, modeling of spacial constraints, and user interactivity is developed in this thesis.