A Modification of DYMO Routing Protocol with Knowledge of Nodes’ Position: Proposal and Evaluation

Abstract

Abstract. Knowledge of the physical location of the nodes is known to improve performance in wireless networks. This is especially true in MANETs, where routing protocols face a continuously changing topology. In the past, routing protocols such as Beacon-Less Routing (BLR) used the location information of the nodes to build the forwarding path in a distributed manner. In this work, we borrow the forwarding approach in BLR and apply it in the route discovery process of DYMO. Under the assumption of nodes knowing their own location, the receiving nodes will compute a delay. The node with lower delay will resend the RREQ first. The rest of forwarding nodes will drop the RREQ once they receive this first RREQ. Thus the best forwarding node is selected in a distributed manner. This modification is expected to reduce the amount of RREQs circulating in the network, lessening the routing overhead.

DYMO Self-Forwarding: A Simple Way for Reducing the Routing Overhead in MANETs

“…To this end, two basic scenarios have been used for testing purposes with 13 and 25 nodes, respectively. In [8] similar results are presented for cases with poor connectivity considering 9 nodes within the same simulation area, where the strategy is expected to perform poorly as also stated in [17]. The geographic area simulated here corresponds to a square shape of 150 × 150 m meters.…”

“…The main goal of this work is to assess the feasibility of the proposed modification in scenarios with different node densities, where the impact of the routing overhead due to the flooding in DYMO may be critical. The main original contributions that differ and improve our previous study in [8] are as follows: (1) a discussion on the Dynamic Forward Delay (DFD) function is presented along with the formulation developed for our routing proposal; (2) the scenarios proposed here include scenarios with better connectivity (higher density of nodes), which are aimed at better modeling of the conditions in which VANET and WSN often perform;…”

“…The full description of the proposed protocol, called DYMOselfwd (DYMO with selective forwarding) was presented in [8]. The proposed modification applies to the retransmission of the RREQ in DYMO and is aimed at reducing the routing overhead in the network.…”

“…Although the Euclidean distance is not always the best choice, as it does not guarantee converging to the destination, we recall that this work is focused on scenarios with high density of the nodes. The performance of DYMOselfwd in low-density scenarios has been already covered in [8], where it has been shown that flooding techniques (e.g., DYMO) perform better. As future work, it is intended to address again scenarios with poor connectivity where, in the case of route failures, we may switch from one routing protocol (e.g., DYMOselfwd) to the other (e.g., DYMO).…”

“…This paper presents a modification of the DYnamic MANET On-demand (DYMO) routing protocol [7] that includes the position of the node in the routing strategy. This modification was previously introduced in [8] where simulations were also included showing improvements for some simple cases. Here, further details are shown about the implementation and results of the proposed modification.…”