2. Problem de?nition
In a densely deployed WSN, the nodes have overlapping communication
and detecting ranges. If there should be an occurrence of an event, multiple
leaf nodes tend to report an event producing high tra?c and devour the majority of their resources
simultaneously. Besides, the information got at the sink is redundant and and
frequently inadequate because of packet loss during transmission. Techniques
such as sleep scheduling have been successfully in-corporate in WSN to minimize
the in-network communication yet it is at the cost of accuracy of event
detection. Most of the existing work considers a trade-o? between e?cient event monitoring and minimum in-network communication.
Strangely, demonstrating the sensor ?eld as a MRF and utilizing the Voronoi
neighbors to ?nd the minimum energy- consuming path to the sink, would be
bene?cial to minimize the overall communication within the network. However,
two major issues restrict modeling the sensor ?eld as a Markov random ?eld.
2.1. Network performance
The targets of enhancing the operations of a WSN, such as power conservation
and resource optimization, are in turn linked
to the function of energy. An example of such a case is
the issue of power conservation
and control. For the ith node, in such a network, the power required for transmission is
given by ‘?i’ such
that ?i is a set of random variables. The corresponding cost function is linked to the problem of constraint optimization.
2.2. Evaluating model’s
By considering the cost function of a given model, the joint
probability ‘P (?)’, can be easily computed given the network’s routing
protocol and the random parameters for example transmission energy and so on are
known. The joint probability accordingly got can be utilized to predict
the protocol’s e?ciency with
respect to the transmission energy.
Considering that every node
has equal probability of data transmission, the transmission energy of every
node is optimized while guaranteeing the
full connectivity of the deployed network isn’t lost.