Extended OLSR and AODV based on multi-criteria decision making method

• Published in: Telecommunication systems Vol. 73; no. 2; pp. 241 - 257
• Main Authors: Kim, Beom-Su, Roh, BongSoo, Ham, Jae-Hyun, Kim, Ki-Il
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.02.2020; Springer Nature B.V
• Subjects: Algorithms; Artificial Intelligence; Business and Management; Computer Communication Networks; Computer simulation; Decision making; IT in Business; Mobile ad hoc networks; Multiple criteria decision making; Multiple criterion; Probability Theory and Stochastic Processes; Telecommunications systems; Topology; Wireless networks; Multi-metric routing protocol; Multi-criteria decision making; Mobile ad-hoc networks
• ISSN: 1018-4864; 1572-9451
• DOI: 10.1007/s11235-019-00609-0

Recently, multi-metric routing protocols have been proposed to enhance the performance of the typical single metric routing protocols in mobile ad-hoc networks. Generally, most of them employ simple value computation to derive a cost value as a way by combining multiple metrics. However, this simple approach fails to define the relationship or priorities between multiple metrics. To overcome this limitation, in this study, we apply the multi-criteria decision making (MCDM) method to determine the weight factors between the metrics. We define criteria to accommodate multiple metric in logical way and decide a better path. For the case studies, we extend the existing proactive and reactive routing protocols, that is, ad-hoc on-demand distance vector (AODV) and optimized link state routing protocol (OLSR). In AODV, we present a strategy for modifying the route request and route reply mechanism to generate a stable path using the MCDM. On the other hand, in OLSR, we propose a modification strategy of MPR selection algorithm to maintain a stable topology using the MCDM. The simulation results show that proposed routing scheme reduces the routing overhead by 15% and 13%, packet loss rate by 12% and 14%, and end-to-end delay by 21% and 19% approximately, compared with other routing schemes such as fixed weighted AODV, OLSR and MCDM based geographical routing protocol.