18.02.2014 Views

wireless ad hoc networking

wireless ad hoc networking

wireless ad hoc networking

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Coverage and Connectivity of Wireless Sensor Networks 9<br />

Theorem 1, if all other sensors are at least 2-perimeter-covered, the sensing<br />

field is 2-covered.<br />

This algorithm can be easily translated to a distributed protocol, where<br />

each sensor only needs to collect local information to make its decision.<br />

The computational cost for each sensor is dominated by sorting the intersection<br />

angles. This incurs a time complexity of O(d log d), where d is the<br />

maximum number of sensors that are neighboring to another sensor. Compared<br />

to the algorithm in Ref. [2], which requires O(d 2 ) for each sensor<br />

to observe the intersection points within its coverage, this algorithm incurs<br />

lower computational complexity.<br />

Note that when a sensor s i is k-perimeter-covered, it is not necessary that<br />

the sensing range of s i is k-covered. In Figure 1.4, sensor s 1 is 2-perimetercovered<br />

but its sensing range is not 2-covered since the sh<strong>ad</strong>ow region<br />

contained in s 1 is only 1-covered. Since this algorithm looks at perimeter<br />

coverage, this can be explained by observing that s 2 is not 2-perimetercovered<br />

(the dashed segment of s 2 ’s perimeter is only 1-perimeter-covered).<br />

Although this property differs from the algorithm in Ref. [2], when gathering<br />

all sensors’ local decisions, the coverage level of the whole field can be<br />

correctly determined.<br />

Ref. [3] further proposes a new algorithm for determining the coverage<br />

level of a three-dimensional (3D) space. Given a set of sensors in a 3D sensing<br />

field, the goal is to determine if this field is sufficiently α-covered, where<br />

α is a given integer, in the sense that every point in the field is covered by<br />

at least α sensors. The sensing range of each sensor is modeled by a 3D<br />

ball. At the first glance, the 3D coverage problem seems very difficult since<br />

even determining the subspaces divided by the spheres of sensors’ sensing<br />

ranges is very complicated. However, the authors show that tackling<br />

1<br />

s 1<br />

s 2<br />

Figure 1.4 An example of the difference between 2-perimeter-coverage and<br />

2-coverage. The sensor s 1 is 2-perimeter-covered but not 2-covered.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!