Page 39 - 2024-Vol20-Issue2

35 |
at the level of packet. Before the source node transmits any
packets, firstly, it will issue a session request. Then, relays are sign steps to make the design of any product less sensitive to
specified to either accept whole relay requests to the session
or not to accept them at all. The nodes reject session requests
either if they have promoted sessions more than the optimal
Pareto value or if the rate at which they tolerate sessions (less
than the generosity score, e) is major than the rate at which
other nodes comply with their session requests. In another
expression, if the value of generosity was greater, the nodes
would be more willing to receive the session request even
though they did not receive a similar amount of assistance
in response. Next, nodes are divided based on their strength
and limitations. The power constraint is defined as the ratio
between the initial allocated powers of a node with its pre-
dictable life (i.e., the slots number) [1, 7]. The equivalent
power (energy) constraint linked with the present session can
then be fine from a node with lower energy (power) shrink-
age. This is due to a node with more power constraints has
no incentive to work more prolifically in advancing packets
because it distinguishes that other nodes with less constraints
of power are unable to exchange in type. For a specified pe-
riod, a node could be a source, a relay, or a destination. A
node consumes energy as a source and relays it to others. The
entire amount of energy that a node expends as a source and
relay will be retained by its power constraints [1, 10]. This is
explained in (1) [1].
k (1) St = T12 + T22 + T32
? es p j + er p j = P class (p)
j=1
Where: es j is the average energy consumed for each slot as
a source, er j is the average energy spent for each slot as a
relay, K stands for the total number of sessions, j is the current
session and class (p) stands for the power class p constraint.
GTFT has the advantages of the vital (dynamic) algorithm.
Ratio of the packets forwarded to other nodes does not depend
only on the settings of initial power but also on the state of the
network. If the rates of another nodes accepting relay requests
are lower than the similar node that will receive them, the
node will be modified and thus forward fewer data to others
at some point [1, 11].
IV. TAGUCHI METHOD
This article aims to present the problem of improving GTFT
node behaviour as a series of experiment design tests based
on Taguchi’s method for designing experiments. According
to optimization theory, an experiment is a series of tests in
which the inputs are changed according to a certain procedure
to determine the causes of the output response changes [12].
Dr. Taguchi Genichi improved the quality control called
”Taguchi Design of Experiments”. The Taguchi method was
Jumaa, Mohamad, Allawy & Mohammed
used to arrive at the best values of the factors affecting the de-
the influence of noise. The results of using the Taguchi method
were optimal in improving product quality [13]. The princi-
ple of orthogonality is the mathematical basis of Taguchi’s
method for designing an experiment, as it is used to measure
the influence of controllable factors. Taguchi matrices contain
columns of independent coefficients, and the size of orthogo-
nal matrices in Taguchi depends on the number and levels of
designer variables. Details could be founded in [14].
The number of experiments affecting the design can be found
based on several factors, which are the number of Taguchi pa-
rameters (M) and the number of levels for each parameter (G),
as these factors are related to the relationship (M*f+1), where
f = (G-1) [15]. Three levels, four factors can be controlled
with nine experiments (L9 (3)4). The orthogonal Taguchi
group can be found in [16], [17], [18]. While other applica-
tions of Taguchi method could be seen in [19], [20], [21], [22].
The fitness function and SNR are used to evaluate the exper-
iment. The SNR is used to study the influence of factors to
find the optimal solution. To calculate the SNR, (2, 3, 4, 5,
and 6) are needed [13].
Sm = (T 1 + T 2 + T 3)2 (2)
N
(3)
Se = St - Sm (4)
Se (5)
Ve = (N - 1)
(Sm -Ve) (6)
SNR = 10Log (NxVe)
N is the number of results equal to 3 and T (T1 to T3) is the
results values. In this article, T1, T2, and T3 are the measured
throughput, the measured ?, and the measured ? respectively
(? and ? are defined in sections 5, 6). Meanwhile, SNR is
the signal to noise ratio.
V. GTFT COOPERATION STRATEGY BASED
AD-HOC NETWORK SIMULATION
In this article, a single node (n1) behaviour has been con-
sidered which is a part of a population contains N nodes