Delay-Minimized Route Design for Wireless Sensor-Actuator Networks

Edith C.-H. Ngai1, Jiangchuan Liu2, and Michael R. Lyu1

1Department of Computer Science and Engineering, The Chinese University of Hong Kong 2School of Computer Science, Simon Fraser University, Vancouver, BC, Canada

IEEE Wireless Communication & Networking Conference 2007

Outline Introduction Related Work Route Design Problem (RDP) Formulation MST-Based Route Design Algorithm Performance Evaluation Conclusion and Future Work

WSN Distributed and large-scale like the Internet A group of static sensors

resource constrained wireless communications

WSAN Collection of sensors and actuators Sensors

numerous resource-limited and static devices monitor the physical world

Actuators resource-rich devices equipped with more energy, stronger

computation power, longer transmission range, and usually mobile

make decisions and actuate adaptively in response to the sensor measurements

Applications

Motivation Given

Each static sensor has a limited buffer Non-uniform data generation rates among the sensors Sensor stores locally sensed data and uploads the data

until some actuator approaches

Strategy Actuator visits locations with higher importance (i.e.

higher data rate) more frequently

Question How to minimize the inter-arrival time from the actuator

to the static sensors???

=> Route Design Problem (RDP)

Related Work Mobile elements to carry data in wireless networks

Architecture using moving entities (Data Mules) to collect sensor data [Shah et. al. SNPA’03]

Mobile sinks with predictable and controllable moving pattern [Chakrabarti et al. IPSN’03, Kansal et al. Mobisys’04]

Mobile sinks can find the optimal time schedule to stay at appropriate sojourn points [Wang et al. HICC’05]

Message ferry (MF) approach to address the network partition problem in sparse ad hoc network [Zhao et al. Mobihoc’04]

Related Work (cont.) Joint mobility and routing algorithm with mobile relays to

prolong the network lifetime [Luo et al. Infocom’05] Partitioning-based algorithm to schedule the movement of

mobile element (ME) to avoid buffer overflow and reduce min. required ME speed [Gu et al. Secon’05]

Vehicle routing problem (VRP) Considers scheduling vehicles stationed at a central facility to

support customers with known demands Minimize the total distance traveled Variations

Capacitated VRP (CVRP) VRP with time windows (VRPTW)

Problem Formulation WSAN consists of multiple actuators and a set of static sensors

Actuators move in the sensing field along independent routes Each static sensor has a limited buffer to accommodate locally sensed data When an actuator approaches, the sensor can upload the data to the actuator and free the

buffer Sensors are assigned with different weights Wj according to their data rate, type, or

importance

System Parameter

Route Design Problem (RDP)

Characteristics1. The sensors are of different weights, according to their

data generation rates and importance. Sensor locations with higher weights will achieve lower

average actuator inter-arrival times.

2. Sensors upload data to actuators through wireless communications Data transmission is possible only when the distance between

the sensor and actuator is within a communication range Rs.

3. It is not necessary for each route to pass through the depot (or the base station) Actuators generally can interact with the base station by

wireless communications.

Definition and Property

Route Design Algorithm Design independent routes for multiple actuators Utilize multiple minimum spanning trees (MSTs) Construct M routes with equal period where highly

weighted sensors will be visited more frequently A sensor location with weight Wi will be visited by Wi*M

actuators (routes) E.g. Wi = 0.75, M=4 => Ni = 3 If all routes have the same period T, from property (2), the

average inter-arrival time Aavg will be T/3

(1) Clustering with MSTs Ni = ceil (Wi * M) Locations with the same Ni belong to the set Si Our algorithm builds M spanning trees Tk, where

k = 1, …, M Locations with highest Ni=M will be included in all

trees Then, the locations with the next highest Ni will be

assigned to Ni trees with lowest costs The process repeats until there is no remaining

locations

Example

(2) Form a TSP Solution The M spanning trees result in M groups of nodes to be

walked through by distinct actuators The route design problem can be reduced to traveling

salesman problem (TSP) for each group of nodes In literature, several algorithms to calculate the TSP paths

are provided, such as the nearest neighbor, LKH, and some polynomial approximation schemes

We adopt the Approx-TSP-Tour algorithm here, which use MST to create a tour and perform a preorder traversal on the tree to obtain a Hamiltonian cycle

(3) Determine the Locations of Actuators It is more efficient for a sensor to have short waiting time Maximum inter-arrival time Amax may also be an

important consideration other than Aavg We focus on the sensor locations with the highest Wi and

select it as reference point pr Each actuator k will be assigned to the point after

travelling for time T*k/M from pr on its own route Encourage more even inter-arrival time of the actuators

Performance Evaluation Parameters

Average Inter-arrival Distance under Uniform Random Sensor Distribution

N=50, M=5 N=100, M=7

Average Inter-arrival Distance underNon-uniform Sensor Distribution

N=50, M=5 N=100, M=7

Distribution of Average Inter-arrival Distance

N=50, M=5 N=100, M=7

Conclusion and Future Work We focused on WSN with multiple actuators and their route design We demonstrated the problem is NP-hard and proposed an effective

MST-based approximation algorithm It aims at minimizing the overall inter-arrival time of the actuators It differentiates the visiting frequency to sensor locations with

different weights Simulation results suggested that the algorithm remarkably reduces

the average inter-arrival time Future work: Improve the performance of the route design algorithm

and consider the cooperation among the actuators

Thank you!