Data Transmission Reduction Schemes In WSNS For Efficient IOT Systems

Abstract

Wireless Sensor Networks (WSNs) play a crucial
role in collecting the data from the nodes in the
Internet of Things (IoT) systems. However, as
transmission consumes dominant factor of
energy consumption, the number of data
transmissions need to be reduced. To address
this issue, this project evaluates a two-tier data
reduction framework by utilizing data
compression and prediction schemes. The data
prediction (DP) scheme is implemented by
using neural networks and Long Short-Term
Memory (LSTM) to minimize the data
transmission between sensor nodes and cluster
nodes by predicting the data locally and
transmitting only when the deviation exceeds
the threshold. The data compression (DC)
scheme employs Principal Component Analysis
(PCA) to reduce data volume between cluster
heads (CH) and gateway nodes. The
performance of the two-tier scheme is
evaluated in terms of percentage reduction in
data transmissions.