Virtual Coordinates with Backtracking for Void Traversal in Geographic Routing

Abstract

Geographical routing protocols have several desirable features for use in ad hoc and sensor networks but are susceptible to voids and localization errors. Virtual coordinate systems are an alternative solution to geographically based routing protocols that works by overlaying a coordinate system on the sensors relative to well chosen reference points. VC is resilient to localization errors; however, we show that it is vulnerable to different forms of the void problem and have no viable complementary approach to overcome them. Specifically, we show that there are instances when packets reach nodes with no viable next hop nodes in the forwarding set. In addition, it is possible for nodes with the same coordinates to arise at different points in the network in the presence of voids. This paper identifies and analyzes these problems. It also compares several existing routing protocols based on Virtual Coordinate systems. Finally, it presents a new routing algorithm that uses backtracking to overcome voids to achieve high connectivity in the greedy phase, higher overall path quality and more resilience to localization errors. We show these properties using extensive simulation analysis.

Aligned Virtual Coordinates for Greedy Routing in WSNs

“…The paper advocates the use of 3 reference points to assign the virtual coordinates, constructing a 3-dimensional VCS. We showed that this 3D VCS may not sufficient to map the network effectively [23]. VCap, even with 4 coordinates, is worse than GPSR both in delivery ratio (node pair reach-ability) and path quality.…”

“…In contrast, high density may not resolve the routing anomalies in VCS [23]; it may even exacerbate them due to the quantization noise problems such as VCS forwarding void presented in section III. The complementary solutions for a void avoidance are not effective as described in section II.…”

“…Consider a 4D VCS, set up according to the VCAP scheme [15], [13], [23]. Further, consider a set of nodes A, B and C with virtual coordinates as V(A), V(B) and V(C), where A and B are neighbors, and B and C are also neighbors.…”

“…Most existing research work in geometric routing protocols concentrates on optimizing the complementary routing algorithm such as perimeter routing [3], [2], [6], or backtracking in VCS [12], [14], [13], [16], [23]. The dominant part of the geographical routing, greedy forwarding, is largely ignored.…”

On Certain Connected Resolving Parameters of Hypercube Networks

“…Manuel et al [9] have shown that the problem remains NP-complete for bipartite graphs. There are many applications of resolving sets to problems of network discovery and verification [1], pattern recognition, image processing and robot navigation [2], geometrical routing protocols [10], connected joins in graphs [11] and coin weighing problems [12]. This problem has been studied for trees, multi-dimensional grids [2], Petersen graphs [13], torus networks [14], Benes networks [9], honeycomb networks [15], enhanced hypercubes [16] and Illiac networks [17].…”