Evaluating the Performance of QoI Algorithms in Realistic MANETs

Quality of Information (QoI) based solutions for multi-criteria decision-making problems have recently gained attention in tactical networks. These schemes combine application and network layer information to select the best source among multiple available options to meet the specific quality needs of the task. A large part of this research evaluates their performance using simple models of underlying network. In this paper, we study the performance of a QoI scheme in a more realistic network environment. For this purpose, we create a MANET environment with realistic radio propagation, MAC and routing protocols to study the performance of a state-of-the-art QoI scheme for source selection procedure. We cross compare this solution with a simple Quality of Service (QoS) algorithm that selects a source with highest available bandwidth. We evaluate the performance of both schemes in a 100 nodes network. In situations where a single network metric is of significance to the task, simple traffic engineering solution is a better choice with minimal overhead. In contrast, QoI scheme provides an efficient solution when multiple network criteria contribute equally to the decision making, at the cost of increased network overheads and long convergence time.