Stochastic Modeling of Mobile Ad Hoc Networks

Ad Hoc Networks are computer networks that do not require any infrastructure (e.g. Access Points) for its devices to exchange information, relying on a cooperative, distributed information relaying to deliver the message from the source device to the destination device. If the wireless devices in this network are mobile, we talk about a Mobile Ad Hoc Network (MANET) and if the devices' mobility is constrained or very fast (for example when the wireless devices are installed on vehicles), we talk about Vehicular Ad Hoc Network (VANET).

The main issues with MANETs and VANETs is that their topology is highly dynamic and every node generates and relays network traffic simultaneously, which makes it difficult to model how the information is exchanged between two devices in the network.

Having accurate models to predict network performance can improve the efficiency with which the MANET is utilized, and can enable use cases with strict performance requirements in terms of throughput and delay (e.g. road safety and health services dispatching).

Several studies aim at modeling such networks, with various limitations in terms of model generality or realism. The goal of a thesis on this topic is to advance the understanding of MANET performance in a variety of scenarios and to update the constantly advancing landscape of proposed models.

Related Literature:
[1] On Performance Modeling for MANETs Under General Limited Buffer Constraint
[2] Effect of Mobility Model and Packet Size on Throughput in MANET's
[3] Towards empirical study based mathematical modeling for throughput of MANETs
[4] Modeling path duration distributions in MANETs and their impact on reactive routing protocols
[5] On the Throughput and Delay in Ad Hoc Networks With Human Mobility