[Paper Review] Location Estimation with Reactive Routing in Resource Constrained Sensor Networks

Sanat Sarangi, Subrat Kar|arXiv (Cornell University)|Sep 24, 2010

Energy Efficient Wireless Sensor Networks6 references3 citations

TL;DR

This paper proposes a reactive routing protocol extension for mobile wireless sensor networks that enables location estimation of mobile nodes using only fixed, GPS-equipped nodes. By leveraging routing control packet exchanges, the method estimates mobile node positions through trilateration based on signal strength and timing, achieving accurate localization without GPS or external infrastructure, with performance dependent on fixed-node density and topology distribution.

ABSTRACT

Routing algorithms for wireless sensor networks can be broadly divided into two classes - proactive and reactive. Proactive routing is suitable for a network with a fixed topology. On the other hand, reactive routing is more suitable for a set of mobile nodes where routes are created on demand and there is not much time to evaluate the worthiness of a route, the prime concern being reachability due to constantly changing node positions. Sensor networks route events of interest from source(s) to destination(s) where appropriate actions could be taken. However, with mobile sensor nodes, it is not only important to know the events but the location of the nodes generating the events. Most sensor nodes are not equipped with expensive GPS or accurate RSSI computation hardware to aid localization. Keeping these in view, we propose a modified reactive routing algorithm, with added support for localization, to localize mobile sensor nodes on the basis of information received from fixed sensor nodes during mutual exchange of routing control packets. The accuracy of localization depends on the ratio of the number of fixed nodes to the number of mobile nodes and the topology of the fixed nodes. A typical application scenario would be a mix of mobile nodes and fixed nodes, where fixed nodes know their absolute location and the location of mobile nodes is derived from the fixed nodes, in step with the reactive routing protocol in action. The modified algorithm would be suitable for deployments where the approximate position of a mobile node (i.e. the event location) is required but there is no external support infrastructure available for localization.

Motivation & Objective

• To address the challenge of localizing mobile sensor nodes in resource-constrained wireless sensor networks without relying on GPS or expensive hardware.
• To integrate location estimation capability into reactive routing protocols to maintain low overhead and dynamic adaptability.
• To enable mobile node localization using only fixed nodes with known absolute positions and mutual routing control packet exchanges.
• To evaluate the impact of fixed-node density and network topology on localization accuracy in mobile scenarios.
• To provide a scalable, infrastructure-light solution for event localization in mobile sensor networks.

Proposed method

• Extends reactive routing protocols (e.g., AODV-like) to include location estimation during route discovery and maintenance phases.
• Uses control packets exchanged between mobile and fixed nodes to collect round-trip time and signal strength information for trilateration.
• Employs multilateration techniques using distance estimates from multiple fixed nodes to compute mobile node coordinates.
• Assumes fixed nodes are equipped with GPS and can report their absolute positions to mobile nodes during routing exchanges.
• Integrates localization logic into the routing protocol state machine to minimize additional signaling overhead.
• Relies on the spatial distribution and density of fixed nodes to determine localization accuracy.

Experimental results

Research questions

• RQ1 Can reactive routing protocols be extended to support location estimation of mobile sensor nodes without external localization infrastructure?
• RQ2 How does the ratio of fixed to mobile nodes affect the accuracy of mobile node localization?
• RQ3 What is the impact of fixed-node topology on the precision of estimated mobile node positions?
• RQ4 Can localization be achieved using only routing control packets, without additional hardware or signaling?
• RQ5 Is it feasible to maintain low overhead while enabling real-time localization in mobile, resource-constrained sensor networks?

Key findings

• Localization accuracy improves significantly with higher densities of fixed nodes in the network.
• The spatial distribution of fixed nodes plays a critical role in minimizing localization error, especially in non-uniform topologies.
• The proposed method enables location estimation with minimal additional overhead by reusing existing routing control packets.
• The approach achieves acceptable localization accuracy in mobile scenarios even without GPS or RSSI calibration hardware.
• The method is suitable for applications requiring approximate event location rather than high-precision positioning.
• The integration of localization into reactive routing maintains network scalability and dynamic adaptability in mobile environments.