Two-phase anticipatory system design based on extended species abundance model of biogeography for intelligent battlefield preparation

• Design of extended BBO model with the concept of efforts required in migration.
• Design of a two phase anticipatory system for battlefield preparation.
• Use of ACO, PSO and BBO, which are all natural computation techniques.
• Improvement over traditional approaches which cannot predict the destination.

This paper presents an extended model of biogeography based optimization (BBO) as opposed to the classical BBO wherein the HSI value of a habitat is not solely dependent upon the emigration and immigration rates of species but the HSI value is a function of different combinations of SIVs depending upon the characteristics of the habitat under consideration. The extended model also introduces a new concept of efforts required in migration from a low HSI solution to a high HSI solution for optimization in BBO. Hence, the proposed extended model of BBO presents an advanced optimization technique that was originally proposed by Dan Simon as BBO in December, 2008. Based on the concepts introduced in our extended model of BBO and its mathematics, we design a two – phase anticipatory system architecture for intelligent preparation of the battlefield which is the targeted optimization problem in our case. The proposed anticipatory system serves a dual purpose by predicting the deployment strategies of enemy troops in the battlefield and also finding the shortest and the best feasible path for attack on the enemy base station. Hence, the proposed anticipatory system can be used to improve the traditional approaches, since they lack the ability to predict the destination and can only find a suitable path to the given destination, leading to coordination problems and target misidentification which can lead to severe casualties. The designed system can be of major use for the commanders in the battlefield who have been using traditional decision making techniques of limited accuracy for predicting the destination. Using the above natural computation technique can help in enabling the commanders in the battlefield for intelligent preparation of the battlefield by automating the process of assessing the likely base stations of the enemy and the ways in which these can be attacked, given the environment and the terrain considerations. The results on two natural terrain scenarios that of plain/desert region of Alwar and hilly region of Mussourie are taken to demonstrate the performance of the technique where the proposed technique clearly outperforms the traditional methods and the other EAs like ACO, PSO, SGA, SOFM, FI, GA, etc. that have been used till date for path planning applications on satellite images with the smallest pixel count of 351 and 310 respectively. For location prediction application, the highest prediction efficiencies of the traditional method on Alwar and Mussourie was only 13% and 8% respectively as compared to the proposed method.