Enhancement Of Clustering Algorithm Using 3D Euclidean Distance To Improve Network Connectivity In Wireless Sensor Networks For Correlated Node Behaviours
Node behaviour plays an important role for network clustering to increase performance in wireless sensor networks. Clustering is one of the most important techniques used in wireless sensor networks for energy consumption reduction to prolong node lifetime. The main idea of clustering is that, inste...
Saved in:
| Summary: | Node behaviour plays an important role for network clustering to increase performance in wireless sensor networks. Clustering is one of the most important techniques used in wireless sensor networks for energy consumption reduction to prolong node lifetime. The main idea of clustering is that, instead of transmitting with the maximal power, nodes collaboratively determine their neighboring node by forming the proper neighborhood relation under certain criteria, with the purpose of maintaining connectivity. The connectivity of nodes relies on cooperative node where energy in each node plays an important role to keep the node in cooperative state longer. Most of the research in clustering minimizes energy consumption per node by adjusting nodes transmission power and balancing energy consumption to prolong network lifetime. However, energy consumption in ad hoc network mostly affected by extra activities performed by misbehave node in which previous research failed to address. Misbehave activities such as selfish and malicious node tends to perform correlated behaviour which is capable to partition the network. When partition occurs, network connectivity will be loss and degrade the entire network performance. Thus, this research proposes enhancement on clustering algorithm to mitigate the impact of correlated node behaviour on network performances. The objective of this research is to enhance the existing clustering algorithm to improve network connectivity under the event of correlated node behaviour. This research will formulate 3D Euclidean distance to measure the correlated degree. Correlated degree will measure the proposed parameters to optimize energy consumption to prevent partitioning. The 3D Euclidean Distance with a correlated degree will contribute to the connectivity of the neighboring nodes and it is formulated based on three-point distance within a correlation region. Then, the enhancement clustering algorithm will be constructed based on correlated degree which is selected as a cluster head to serves as a link connectivity between individual node and its neighbor to form network clustering. This research uses an experimental based simulation using NS-2 and C++ programming. The experiment uses four different scenarios namely, cooperatives node, selfish node, malicious node and failure node. The clustering algorithm will be compared against LEACH, EEC, R-HEED, PEGASIS, and LCA algorithms to evaluate the network performance of WSN. The results based on different scenarios behaviour node show that the enhance clustering algorithm built by 3D Euclidean Distance using correlated degree provides better network performance compared with LEACH, EEC, R-HEED, PEGASIS, and LCA algorithms. When compared with other algorithms, the connectivity of network using the clustering algorithm increased by 11% for ECA, 17% for LEACH, 20% for EEC, 23% for R-HEED, 14% for PEGASIS, and 15% for LCA. This finding shows that the enhancement of the existing clustering algorithm prolongs its network lifetime and the node may communicate with neighboring node efficiently. It is also shown that the enhancement for the clustering algorithm may help correlation region to change its cluster formation dynamically to achieve the required network connectivity and increase network performance. |
|---|