Authors : Khudhair Abbas Mohammed, Muamer N. Mohammed, Mazlina Abdul Majidmohd and Mohd Sharifuddin Ahmad
Abstract: Security aspects are serious issues for multi-agent systems that should be taken into consideration when designing the systems. The significant effects and outcomes of the software agent technology make its application more and more popular in several fields. The rapid development of the agent technology promotes its ability to be applied in many different areas such as transportation, e-Business and healthcare. However as with any other systems, the agent technology is also vulnerable to security threats and consequently, there is a need to equip the system’s functionalities with some security defense system. A number of methodologies are providing methods and techniques for constructing security for multi-agent systems. A survey of the current state-of-the-art of security issues for MAS is the aim of this study. Techniques and models for securing multi-agent systems are presented in this study which is categorized according to the concepts and models regarding agent’s and system’s functionalities. Additionally, possible threats and attacks on multi-agent systems are defined and considered with agent’s roles and communications concerns. Moreover, up to date security solutions are elaborated at both levels: agent and the system levels. Finally, an analysis of existing work problems and the challenges of future research for the security of multi-agents are presented.
Khudhair Abbas Mohammed, Muamer N. Mohammed, Mazlina Abdul Majidmohd and Mohd Sharifuddin Ahmad, 2018. DA Conceptual Security Sub-Model for Enhancing Human-Agent Collaboration Based on Generic Nodal Approach. Journal of Engineering and Applied Sciences, 13: 4905-4912.