CLUSTERED DATA DIFFUSION ROUTING PROTOCOL FOR LARGE-SCALE WIRELESS SENSOR NETWORKS

One of the major challenges in the implementation of WSNs is to prolong the lifetime of the energy source in the sensor nodes. This can be achieved through designing energy-efficient routing protocol. Energy-efficient routing protocol enables WSNs to stay in operation for a long time by managing communication between the sensor nodes and the sink. In addition the routing protocol can handle a large number of sensor nodes in energy-efficient manner utilizing multihop communications among the sensors. Four existing routing protocols performance were analyzed using J-Sim simulator. Results obtained from the performance analysis show that the Directed Diffusion (DD), creates a large amount of overhead when broadcasting a query message to the whole network. This causes huge amount of energy consumption, which reduces WSNs lifetime. Low Energy Adaptive Clustering Hierarchy (LEACH) routing protocol, which assumes one-hop communication range from the sink node is not scalable for large-scale WSNs. To enhance the way the data can be gathered in query-driven data reporting method, a cluster-based data diffusion routing protocol for large-scale wireless sensor networks has been proposed. In the proposed method, the sink node sends the interest message unicastly, only to the cluster heads. In addition, multihop communication between the cluster heads for sending the interest, and receiving the data packets back from the source node has been used. The results obtained have been analyzed and compared with DD and PCDD protocols as well as with other works, in which some enhancements over DD were made using different approaches. The overall results using different metrics have shown that, the proposed protocol outperforms DD and PCDD in saving the energy . The improvement of CDD in saving the energy is between 50% and 63.64% while comparing it to DD and, 8% and 42.8% compared to PCDD, for the fixed density scenario. For the fixed area scenario the improvement is up to 29.4% and at least 15% while comparing CDD to PCDD and up to 63.1% and at least 26.1% while comparing it to DD.
The proposed clustering data diffusion method also was extended to handle a combination of mobile and static nodes, and user’s (sink’s) mobility as well. It improved the coverage of the sensor network and for applications which need some movements of the sink node in order to gather data by sending the query message to the sensor field. A performance analysis for the extended protocol was conducted, and a significant energy saving was achieved especially in denser networks.