Transcript Slide2

後卓越計畫進度
中央大學 許健平 教授
淡江大學 張志勇 教授
報告 張育嘉
Location Estimation for Wireless
Sensor Networks

Location estimation algorithms can be
categorized as

Range-based schemes


TOA, TDOA, AOA, RSSI
Range-free schemes


DV-based scheme
Convex Position Estimation (CPE)
Convex Position Estimation (CPE)
Distributed Location Estimation
Algorithm



A few nodes have known position – equipped with
GPS or placed deliberately (Beacon node)
The remainder nodes estimate position from
knowledge about communication links
The beacon node has the ability of modifying the
power level
Distributed Location Estimation
Algorithm
P
O2 (x2,y2)
O2 (x2,y2)
r
O1 (x1,y1)
O3 (x3,y3)
O1 (x1,y1)
Ｑ
O4 (x4,y4)
O2 (x2,y2)
O1 (x1,y1)
O3 (x3,y3)
O2 (x2,y2)
O1 (x1,y1)
O3 (x3,y3)
Reduce the Range of ER

If a normal node have farther neighbor beacon, the ER
can be reduced to smaller one
O4 (x4,y4)
O4 (x4,y4)
O2 (x2,y2)
O2 (x2,y2)
O1 (x1,y1)
O1 (x1,y1)
O3 (x3,y3)
O3 (x3,y3)
Reduce the ER: Rule 1

Rule 1: The cut point is based on the intersection point
of circle and the borders of estimative rectangle
Reduce the ER: Rule 2

Rule 2: The cut point is based on the midpoint of
intersection arc
Reduce the Computation Complexity


We need to use some trigonometric functions, such as
sin, cos, asin and acos to find the midpoint of an arc
A line segment is used to instead of the arc
How to Cut the ER?

We need to define complete rules to decide what
border must be cut
(a)
(b)
(c)
Slope of Line Segment

We can divide the slope range of a line segment into
four regions as the figure show.
tan 112.5° = -2.4142
tan 67.5° = 2.4142
tan 22.5° = 0.4142
tan -22.5° = -0.4142
tan -67.5° = -2.4142
Real networks environment