On evaluating the trade-offs between broadcasting and multicasting in ad hoc networks

Multicasting in ad hoc networks has received a lot of attention for the important application of disseminating information to multiple recipients. Most multicast protocols require the creation and maintenance of a structure (such as a tree or a mesh) for distributing information to the group members. In contrast, broadcast schemes are simple schemes which aim to distribute the information to all or a fraction of the nodes in the network without having a structural framework. While the creation/maintenance of the structure could potentially be cumbersome and heavyweight, multicast does offer benefits in terms of restricting the number of nodes that perform rebroadcasts. We argue that it is not a given that multicast is a better choice for group communications in all possible scenarios and that there could be circumstances wherein the use of a simple broadcast based technique would be more advantageous. In support of this claim, we study various scenarios to evaluate and quantify the trade-offs between broadcasting and multicasting. In particular, we perform simulation experiments using the on-demand multicast routing protocol and the simple broadcast algorithm as candidate protocols for multicasting and broadcasting, respectively. These protocols have been shown to be the elite protocols in their classes in prior work. Our results demonstrate that multicasting is preferable only under conditions of moderate mobility and with multicast group sizes smaller than 40%.