Академический Документы
Профессиональный Документы
Культура Документы
transmission by running a linear program. This algorithm is spanning tree or the same an aggregation tree rooted at BS
able to maximum lifetime of network with certain location is formed. In this paper, the network lifetime is credited
of each node and base station. until all nodes are active.
One year later, Dasgupta et al. [3] extend MLDA with
applying cluster based heuristic algorithm called CMLDA 4. Genetic Algorithm
where nodes are grouped into several clusters with pre-
In this study, GA is applied in order to obtain balanced and
defined size. The energy summation of cluster member
energy efficient spanning trees. Every chromosome
nodes is their cluster’s energy. The distance between
represents a tree where gene index indicates node and
clusters is computed by maximum distance between every
contained value point out corresponding parent. Using
pair of nodes of two clusters. From cluster formation on
standard GA, the optimum minimum spanning tree will be
then, MLDA is applied.
resulted.
Ten at al. [4] study two spanning tresses so that aggregate
scheme and data gathering have been applied to extend 4.1 Gene and Chromosome
network lifetime. In this paper, generally, two methods are A chromosome is the collection of genes or nodes which
considered to manage power among nodes. The first, power have fixed length according to the number of former nodes.
aware version (PEDAP) which have an attempt to extend
lifetime, balancing the energy consumption among nodes, 3 0 5 1 2 0
unlike second method, PEDAP, non power aware version
which minimizing the total energy consumed from the 1
system in each data gathering round [1]. This method 5
extends the lifetime of the last node. The edge cost is 0
2
calculated in different ways. In PEDAP, edge cost is the
4 3
same summation of energy mounts for transmission and
receiving while PEDAPPA, dividing PEDAP edge cost with
transmitter residual energy results asymmetric Figure 1. Chromosome and corresponding Tree Example
communication costs. A node with higher cost is considered 4.2 Crossover
later in the tree as it has few incoming. After determining of The main step to produce new generation is crossover or
edge costs, Prime’s minimum spanning tree rooted at the reproduction process. In fact, it is a simulation of the sexual
BS, will be formed for routing of packets. This calculation is reproductive process that the inheritance characteristics
computed per every 100 iterations. Being active for all nodes naturally are transferred into the new population. To
and awareness of node locations at the BS also are their generate new offspring, crossover selects a pair of
assumptions. individuals as parents from the collection formed by
Jin et al. [5] utilize GA to fulfill energy consumption selection process for breeding. This process will continue
reduction. This algorithm gets a primary number of pre- until the certain size of new population is obtained. In
defined independent clusters and then biases them toward general, there are various crossover operations which have
optimal solution with minimum communication distance by been developed for different aims. The simplest method is
the iterations of generation. They come to conclusion that single-point in which a random point is chosen whereby two
the number of cluster heads is reduced about 10 percentage parents exchange their characteristics. Table 1 shows an
of the total number of nodes. They also show cluster based example of mating of two chromosomes in single point way.
methods decrease 80 percentage of communication distance
as compare to direct transmission distance. Table 1: Single point method at random point 6
In 2005, Ferentinos et al. [6] improve the proposed Jin et al.
First Second
algorithm with extended fitness parameter. They investigate
Parents 101101’01101101 011110’10001011
energy consumption optimization and uniformity
Offsprin 101101’10001011 011110’01101101
measurement point, using a fitness function involved status
g
of sensor nodes, network clustering with the suitable cluster
heads, and also the selection between two signal ranges 4.3 Fitness Function
from normal sensor nodes. Fitness function is a procedure which scores any
chromosome. This value helps us to compare the whole ones
3. Problem Statement to each other to survival or death. Below, we propose a
In this study, we suppose that every node firstly has pre- fitness function where N is the number of nodes and setup
defined energy for receiving multi data packets as it is able energy is considered for calculating electrical power.
to monitor environment in addition to transmit children Echildren is required energy to sending data packet received
packets as well as sending single one to parent or BS. This from children.
task periodically will be continued as long as possible.
In our algorithm, at first, all nodes send a sample certain
packet to the BS once they are at ready. Just then minimum
(IJCNS) International Journal of Computer and Network Security, 89
Vol. 2, No. 9, September 2010