Protocol Integration for Intelligent Monitoring Applications in Wireless Sensor Networks - Foundations on Natural and Artificial Computation - page 512

T U

1

A
C
K
A

A
C
K
A

D
L
Y

D
L
Y

S
Y
N

t listen . . .

t listen t listen

BS

1 2 n

T D

BBS

1

T

U

D
S
C

A
C
K

D
A
T

D
A
T

. . .

D
A
T

Nd1

1

T

D

3 Ts+delay 1

D
S
C

A
C
K

D
A
T

D
A
T

D
A
T

Nd2 . . .

Ts+delay 2

2

A
C
K
A

D
S
C

A
C
K

D
L
Y

S
Y
N

D
A
T

D
A
T

D
A
T

. . .

Nd3

Ts+delay 3

Fig. 1. SA-MAC timing diagram

distributed and self-organized to support topological changes, since these fea-
tures are essential to ensure the ecient scalability of a wireless sensor network
[6]. In order to do this, we integrate SA-MAC with NORIA, which will create
the logical tree in an optimal and intelligent manner, considering the differences
between nodes and always choosing the most adequate ones to be included in
the routes.
In [7] and [8] different routing and self-organizing techniques for WSN are
described. These techniques serve as a basis for future routing algorithms which
are intended to make the network more ecient. After all, we have came up
with the idea of using artificial intelligence techniques to help the decision-
making process in order to get more ecient algorithms. Artificial intelligence
techniques reinforce the eciency and the performance of routing algorithms,
by combining data from nodes and their interactions in order to make de-
cisions oriented to improve global network performance. Decisions related to
information transmission from a source to the sink are one of the most im-
portant aspects in sensor networks. Our approach tries to show how artificial
intelligence techniques, specifically fuzzy logic, can help these decision-making
processes to improve eciency and to extend network lifetime. We are very in-
terested in studies combining fuzzy rules with networking. In [9] the authors
propose two asynchronous MAC protocols, they make a complicated schedule
interval and they design a rescheduling fuzzy logic system to monitor the influ-
ence of accumulative clock-drifts, the variance of trac strength and service
capability on communications. In their experiments, they increase the num-
ber of nodes within a cluster from 5 to 30 and exclude the factors coming
from physical layer and network layer to simplify the analysis. Another in-
teresting approach is taken in [10], where the authors propose FuzzyMAC, a
CSMA/CA-based MAC protocol that utilizes two separate fuzzy logic con-
trollers to optimize both the MAC parameters and a sleeping schedule duty-
cycle. The experiments only show results for 50 nodes and use a proprietary
simulator.