MobileMAN: integration and experimentation of legacy mobile multihop ad hoc networks

Although research on mobile ad hoc networks has been ongoing for some time, there are relatively few experiences with real ad hoc networks in laboratory testbeds, and users never use multihop ad hoc networks. This seems due to a gap between what end users might find useful, and what research is currently addressing. Indeed, a large portion of research activities concentrate on the development of novel solutions to optimize lower-layer protocols in often unrealistic settings, while little attention is devoted to the quality of service (QoS) these networks may provide to end users in realistic applicative scenarios. The MobileMAN project tried to contribute to reduce this gap by promoting a research plan aimed at combining theoretical research with the integration of developed solutions in prototypes to be used for validating them in realistic small- and medium-scale scenarios (few hops and 10-20 nodes). The aim is to design and integrate a full protocol stack and experimentally quantify the QoS the system is able to provide to the users. This research approach points out that, also in this limited setting, several problems still exist to construct efficient multihop ad hoc networks. In the next article [1] we discuss how cross layering can be exploited to fix some performance problems identified in our analyses.