Investigating multicriteria approaches for the routing problem in ad-hoc wireless networks using QoS-aware metrics

Ad-hoc Wireless Networks have aroused much interest of the scientific and business community in the last two decades. Sensor, vehicular, and mobile networks have evolved from this “ad-hoc” paradigm. Among the various emerging challenges of this research field, the provision of Quality of Service (QoS) is one of the most prominent, since these networks are prone to suffer from instabilities and interference in the wireless medium and frequent topology changes when mobility exists. Depending on application or scenario, the protocol needs to consider two or more QoS criteria when solving the routing problem. In this context, this work proposes to investigate if the use of multiple QoSaware metrics can generate promising compromise solutions considering several network quality indicators in static and mobile Ad-hoc Wireless Networks. For that, a framework that supports several optimization objectives is developed to house the methods. Two new models are proposed - one based on weighted sum method with path size control and another based on compromise method ( -constraint) with pruning mechanism and path size control - and compared with the standard weighted sum method. In order to map a single final solution, a utility function is proposed to choose the parameters (weights and constraints) of each method. In a network simulator, experiments are designed varying mobility, type of application, and packet generation rate. The following network quality indicators are measured: Packet Loss Ratio (PLR), Throughput, Endto-End Delay (E2ED), Network Lifetime (NLT), Normalized Routing Load (NRL), and Packet Error Rate (PER). The results show the proposed methods were more efficient in generating better trade-off solutions and promoting significant improvements in the quality indicators in most scenarios investigated, indicating that these approaches are promising and deserve to be further studied in future works.