۱- Introduction

۲- Literature Survey

۳- Proposed Algorithm

۴- Results Analysis

۵- Conclusions and Future Work

References

Abstract

In modern era WSN, data aggregation technique is the challenging area for researchers from long time. Numbers of researchers have proposed neural network (NN) and fuzzy logic based data aggregation methods in Wireless Environment. The main objective of this paper is to analyse the existing work on artificial intelligence (AI) based data aggregation techniques in WSNs. An attempt has been made to identify the strength and weakness of AI based techniques.In addition to this, a modified protocol is designed and developed.And its implementation also compared with other existing approaches ACO and PSO. Proposed approach is better in terms of network lifetime and throughput of the networks. In future an attempt can be made to overcome the existing challenges during data aggregation in WSN using different AI and Meta heuristic based techniques.

Introduction

Artificial intelligence (AI) includes number of techniques like Particle Swarm Optimization (PSO), Neural Networks (NN), Genetics Algorithms (GA)Ant Colony Optimization (ACO) etc. Artificial intelligence based techniques help in improvement of network lifetime & throughput. ACO may be used for the better routing to wireless sensor networks and PSOis good solution for clustering to select the best cluster head in the network. These approaches may be used in wireless sensor networks at different stages. Basically there are four energy dissipation ways in wireless sensor networks; First one is retransmitting the data, second one is overhearing means “when a particular node catch unwanted data”, third one is idle listening and fourth one is overhead of the data [14]. When data aggregation operation performed in WSN, data conflicting arises in the networks.To overcome thisissue of data conflicting in WSN environment, secure data aggregation may be considered as one of the solution. There are number of data aggregation approaches in WSNssuch as;(i) Centralized approach, (ii) In network aggregation, (iii) Cluster based approach and (iv) Tree based approach [3]. Some researcher used particle swarm optimization technique for cluster head selection. Number of cluster head in wireless sensor networks, iscalculating by equation