A Case Study of Bio-Optimization Techniques for Wireless Sensor Network in Node Location Awareness

Background: In wireless sensor networks the sensors are deployed randomly in the sensing field, therefore the location awareness of the deployed nodes is challenging. The objective is to estimate the location of the deployed sensor nodes through bio optimized algorithms. Methods: This paper compares performance of three best bio-optimization algorithms available: Particle Swarm Optimization (PSO), Shuffled Frog Leaping (SFLA) and Firefly Algorithms (FFA) in estimating the optimal location of randomly deployed sensors. The optimum solution helps to ensure the maximum coverage of sensing capability and Quality of Service (QoS) of the network. The simulation is done using LabVIEW to understand the performance of these algorithms. Findings: The objective function and fitness value is calculated for the algorithms and based on those the error value is determined for 50 nodes and 10 beacons in a 100x100 sensor field dimension. To evaluate the performance under noisy environment a noise of 2% and 5% is added and simulated. Also the transmission radius of beacons changed from 25 to 20m to analyse the error value for optimum location estimation, however when changing the transmission radius it is found that the number of localized nodes are reduced. The performance is analysed based on localization error, computing time and memory. SFLA offers better in localization error however its computation time is more. The PSO takes less computing time and memory compared to FFA. The average error Vs sensor nodes depicts SFLA performance is superior. The methods adapted in this paper throw more lights on performance of these algorithms under noisy environment and effect of localization error under various transmission radiuses. Improvements: These algorithms can be further analysed for centralised localization and also location awareness for mobile nodes.