Secure Localization in UWSN using Combined Approach of PSO and GD Methods
Shanthi M. B.1, Dinesh K. Anvekar2
1Shanthi M. B., Dept. of CSE, CMR Institute of Technology, Bengaluru, (Karnataka), India.
2Dinesh K. Anvekar., Dept. of CSE, VVIT, Bengaluru, (Karnataka), India .
Manuscript received on 23 March 2019 | Revised Manuscript received on 30 March 2019 | Manuscript published on 30 March 2019 | PP: 1535-1538 | Volume-7 Issue-6, March 2019 | Retrieval Number: F2560037619/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Particle swarm Optimization (PSO) is a well known global optimization algorithm best suited for solving complex real world problems. But it has a problem of getting stuck in local minima in certain cases. To mitigate this problem, some researchers have applied modification by hybridizing it with best suited local optimization algorithms with it. In this paper, we have put forward our try to combine Gradient-Descent (GD) optimization algorithm with Particle Swarm Optimization (PSO) for localization of sensor nodes in UWSN. Algorithm works in two stages. In the first stage, GD approach is used to identify the local best particle in local neighborhood. In order to enforce the security during localization, GD is combined with Maximum Likelihood (ML) method to identify the suspicious nodes in the swarm. Second stage of the work continues with finding global best solution using PSO. The experiments have shown that combined approach of GD and PSO have better performance over current approaches in terms of security and accuracy
Keywords: Gradient-descent, Maximum-likelihood, Mitigate, Neighborhood, Optimization
Scope of the Article: Probabilistic Models and Methods