Abstract— The conflict between achieving good performance in terms of time etc. and achieving high quality of security protection introduces new challenges in security critical grid scheduling. Extensive study indicates that the scheduling performance is affected by the heterogeneities of security and computational power of resources. Different tasks may have varied security requirement and even for the same security requirement, security overhead may vary for different node at which the task is scheduled. In this paper, a security driven scheduling using genetic algorithm (SDSG) is proposed which aims at maximizing security while restricting security overhead under a certain limit. Extensive simulation results over dynamically created heterogeneous grid environment reveal that SDSG achieves better security and exhibit less security overhead and makespan in comparison to other such algorithms viz. MinMin MaxMin, SPMinMin and SPMaxMin. Index Terms— Grid computing, Security-aware grid scheduling, Cipher suite, Makespan, Response time I. INTRODUCTION A computational grid is a collection of geographically dispersed heterogeneous computing resources, giving the image of a single large virtual computing system to users [1][2][3]. Scheduling on such platform is an important aspect more so, being this a heterogeneous system. The main challenge of task scheduling in grids is its highly dynamic environment, where the computing resources have their own access policies, security, availability etc. At the same time, resources are of greater heterogeneity ranging from desktop PCs to supercomputers. Grid computing has often extensively supported collaborative projects on the internet. Most of these projects have stringent security requirements. The application itself imbibes security up to some extent, but more usually it is to be supported and ensured by the grid environment. The dynamic and multi-institutional nature of the grid introduces challenging security threats warranting the development of the new technical approaches towards this problem. In a security aware environment, responsibility is delegated to the scheduler for allocating the task on those resources that gives best possible security, while understanding the computational and security heterogeneity of the resources. R. Kashyap is with Lal Bahadur Shastri Institute of management, Delhi, India (phone: 91-11-25307700; fax: 91-11-24522474; e-mail: rekhakashyap@lbsim.ac.in). D. P. Vidyarthi is with School of Computer Science, Jawaharlal Nehru University, Delhi, India (dpv@jnu.ac.in ). Task scheduling in grid is an NP-hard optimization problem, so many heuristic and meta-heuristics algorithms are in use towards finding suboptimal solutions, i.e., solutions whose optimality cannot be guaranteed. Meta- heuristics like Simulated Annealing (SA) [4], Genetic Algorithm (GA) [5], Ant Colony Optimization (ACO) [6], particle Swarm Optimization (PSO) [7], etc. are also used for grid scheduling as they generally produce higher quality results than simple heuristics, although may take a bit longer as they have to generate and evaluate many solutions rather than just one. These nature based meta-heuristics follow the Darwin’s natural selection law i.e. only the fittest can survive. GA a population-based meta-heuristic, was created by John Holland [5] and produces the next generation with the techniques inspired by evolutionary biology, such as inheritance, mutation, crossover, and selection. GA considers a solution as an organism, thus better the quality of the solution higher is the survival probability, through crossover (also called recombination) and mutation. GA can escape from the local optimal to search for the global optimal. In this paper, we propose a genetic algorithm for job scheduling to address the heterogeneity of security mechanism in a computational grid. The proposed Security Driven Scheduling using Genetic algorithm (SDSG) improves the security of the heterogeneous grid while restricting the security overhead within a limiting range. Next section discusses some related work in this field. Scheduling strategy is described in section 3. Section 4 briefs the security model used in this work. Proposed SDSG is analyzed in Section 5. Experimental results and observations for SDSG and the compared heuristics are presented in Section 6 while making the conclusion in the last Section. II. RELATED WORK To achieve the promising potential of underlying distributed resources in the grid, effective scheduling algorithms are fundamentally important. Scheduling, an NP- hard problem, is rather complex one as the Grid being a geographically dispersed heterogeneous multiprocessing environment. Consequent to this is the emergence of many heuristic and evolutionary approaches towards this problem [8] [9] [10] [11] [12] [13]. Some well known heuristic based grid scheduling algorithms, proposed in the literature, are as follows. Casanova et al. [14] proposed an adaptive grid scheduling algorithm for parameter sweep applications, where tasks can share input files and also extended Sufferage heuristics as XSufferage. DFPLTF (Dynamic Security-Driven Scheduling Model for Computational Grid using Genetic Algorithm R. Kashyap, D.P. Vidyarthi Proceedings of the World Congress on Engineering and Computer Science 2011 Vol I WCECS 2011, October 19-21, 2011, San Francisco, USA ISBN: 978-988-18210-9-6 ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online) WCECS 2011