MovementAssisted Sensor Deployment Author Guiling Wang Guohong Cao

Movement-Assisted Sensor Deployment Author : Guiling Wang, Guohong Cao, Tom La Porta Presenter : Young-Hwan Kim

Contents Introduction Technical preliminary Movement-assisted sensor deployment protocols Performance evaluations Discussion and future work 2

Sensor Deployment Stationary protocol n Environment is known and under control Dynamic centralized protocol n Environment is unknown and or hostile w Ex) remote harsh fields, disaster areas and battle fields n n A powerful cluster head is need Problem of single point failure Dynamic distributed self-deployment protocol 3

Problem statement n Given the target area, how to maximize the sensor coverage with less time, movement distance and message complexity Processing n n 1) discovering the coverage holes – Voronoi diagram 2) target positions of these sensors, where should move - VEC, VOR, and Minimax protocols Term n n 4 Coverage holes – the area not covered by any sensor Target positions – the points need to sensing

Voronoi Diagram Voronoi polygon n n Voronoi polygon is the set example 5 of of O as Voronoi vertices of O Voronoi edges of O Voronoi neighbors of O

Voronoi Diagram Sensor deployment protocol are based on Voronoi diagrams Each sensor is enclosed by a Voronoi polygon Polygons together cover the target field Each sensor can examine the coverage hole locally Each sensor needs to know its Voronoi neighbors 6

Three Deployment Protocols Based on Voronoi diagram n the location information of Itself and neighbors heuristic n Runs iteratively until it satisfy Distributed Self-deployment protocols Difference n n n 7/31 VEC pushes sensors away from a densely covered area VOR pulls sensors to the sparsely covered area Minimax moves sensors to their local center area Advanced Ubiquitous Computing

VEC(The VECtor-based Algorithm) The attributes of electro-magnetic particles Terms is the distance between two sensors( , ) is the average distance two sensors ( beforehand ) is the distance between a sensor and boundary n n n The virtual force between two sensors ( , ) ( n Case 1. Voronoi polygon not completely w n away from each other Case 2. Voronoi polygon completely (One) w The other sensor will pushed n ) away Case 3. Voronoi polygon completely (Two) w Virtual force is 0 ( Not pushed ) 8/31 Advanced Ubiquitous Computing

VEC The virtual forces between a sensors and boundary( n ) away from boundary Overall virtual force on sensor is the vector summation Algorithm n 9/31 Movement adjustment Advanced Ubiquitous Computing

VEC The execution of VEC n n 10/31 35 sensors / 50 m x 50 m / random deployment Coverage : 75. 7% -> 92. 2% -> 94. 7% Advanced Ubiquitous Computing

VOR(The VORonoi-based Algorithm) Greedy algorithm which tries to fix the largest hole If a sensor detects the existence of coverage holes -> it will move toward its farthest Voronoi vertex n Where is equal to the sensing range Fig. VOR 11/31 Advanced Ubiquitous Computing

VOR Limit n 12/31 The maximum moving distance is half of the communication range Advanced Ubiquitous Computing

VOR Algorithm n 13/31 Oscillation control Advanced Ubiquitous Computing

VOR The execution of VOR n 14/31 Coverage : 75. 7% -> 89. 2% -> 95. 6% Advanced Ubiquitous Computing

Minimax Algorithm Why minimax? n n Distance of the farthest Voronoi vertex is minimized Regular shaped Voronoi polygon Compare with VOR n n 15/31 Similar to VOR, moving closer to the farthest Voronoi vertex Minimax considers more information and it is more conservative Advanced Ubiquitous Computing

Minimax Algorithm 16/31 Advanced Ubiquitous Computing

Minimax Algorithm To find the minimax point, we only need to find all the circumcircles of any two and any three Voronoi vertices 17/31 Advanced Ubiquitous Computing

Minimax Algorithm 18/31 Advanced Ubiquitous Computing

Minimax Algorithm The execution of Minimax n 19/31 Coverage : 75. 7% -> 92. 7% -> 96. 5% Advanced Ubiquitous Computing

Termination & Optimization Termination n 1) the best coverage is obtained 2) reached the specified maximum round 3) a threshold. Defined as the minimum increase in coverage Optimization n When the initial deployment of sensors may form clusters w Coverage low, deployment time prolong n n 20/31 The algorithm ‘explodes’ the cluster to scatter the sensors apart Only runs in the first round Advanced Ubiquitous Computing

Minimax Algorithm 21/31 Advanced Ubiquitous Computing

Performance Evaluations Two aspects : 1)deployment quality, 2)cost n n 1) is determined by the number of rounds needed and the time of each round 2) is determined by the sensor cost and the energy consumption of the deployment Various system parameters n 22/31 Sensor density, field size, topology, communication range, Advanced Ubiquitous Computing

Performance Evaluations Proposed protocol good! Why VEC worst? 23/31 Advanced Ubiquitous Computing

Performance Evaluations 24/31 Advanced Ubiquitous Computing

Performance Evaluations 25/31 Advanced Ubiquitous Computing

Performance Evaluations 26/31 Advanced Ubiquitous Computing

Discussion To maximize the sensing coverage based on Voronoi diagrams Designed three distributed protocols to move mobile sensors form densely deployed areas Simulation results verified the effectiveness of protocols 27/31 Advanced Ubiquitous Computing

Future Work Optimal Movement Communication Sensing Area Extend to Large Sensor Networks 28/31 Advanced Ubiquitous Computing