Data Mules
From RSN
Contents |
Motivation
Environmental monitoring, at its basic form, involves reliably gathering the measurements from each of the network’s motes which monitor the long term spatial and temporal trends in the environment of interest. A reliable environment monitoring requires low duty cycles of sensor network, which is defined as the percentage of time a network node is active (practically 0.1% or lower) and a sensor network which is deployed in sparse topologies.
Deploying a wireless sensor network over a large area that delivers collected measurements to a central gateway would be uneconomical because this would require a large number of relay points. The reason is that the range of individual wireless links limited (approximately 10 to 50 meters). Even if deploying such a network were feasible, reliably delivering data over it would consume considerable energy due to sending a single packet by each mote on the routing path to the gateway. Furthermore, because wireless links are lossy, the probability of losing a packet increases as the length of the routing path increases.
On the other hand, mobile robots can cover a large geographical area and move close enough to a mote to ensure that the quality of the wireless link will be high. Furthermore, if the robot moves close enough to the mote, the mote can reduce its radio’s transmission power further reducing its energy consumption.
Contributions
This work explores synergies among mobile robots and wireless sensor networks in environmental monitoring. We present a system in which robots act as mules, collecting measurements gathered by sensor network nodes. A proof of concept implementation demonstrates that this approach yields significant improvements in the network’s lifetime by conserving energy that the sensor nodes would otherwise use for communication.
Proof of concept
The proof of concept we designed consists of a network of sensing motes and a number of mobile robots. Each sensing mote is a Tmote Sky from Moteiv. Moreover, each of the Acroname Garcia robots has a Stargate processor board that controls the robots’ motors and sensors. Once the robot approaches a sensing mote, it listens for beacon messages and downloads the requested data after it receives a beacon.
We set the motes’ transmission power to its lowest level to reduce energy consumption. Hence, while planning the mules’ movements, we set the visiting locations to be the motes’ exact locations. This means that the paths mules follow can be pre-calculated as k-TSP tours.
This is joint work with Hopkins interNetworking Research Group.
Related Papers
- O. Tekdas, J.H. Lim, A. Terzis and V. Isler. Using Mobile Robots to Harvest Data from Sensor Fields, IEEE Wireless Communications, Special Issue on Wireless Communications in Networked Robotics, 2008, accepted.
