Enabling autonomous management of distributed pervasive environments

Current networks are providing plenty of services that users can access and use. These services are more and more pervasive and deployed in different networks distributed across an environment. This raises the problem of managing such environments in order to grant access to services from anywhere and to adapt the environment’s networks to dynamic changes. Also, there is a need of an autonomous behavior to reduce human intervention and assure environment’s consistency. This autonomous and distributed behavior leads to the definition and integration of existing and new technologies to enable autonomous distributed management. This is fulfilled by providing paradigms that bring awareness about the surroundings and enabled tools to manage and adapt the environment’s resources. The main problem is to dynamically provide auto‐configuration of networks to deal with the frequent changes which results from users’ roaming, changing services constraints, and changing services themselves, and adding, upgrading or removing policies. The outcome of these issues is a dynamic system with complex management. Hence, this paper proposes the integration of different techniques to provide an autonomous, distributed, and secure management including auto‐configuration, adaptation, and auto‐protection of pervasive environments. Then, policies control the behavior of the environment, devices configuration and the enforcement of security mechanisms to protect sensitive data. Also, mobile agents are employed to distribute management tasks across the distributed environment. In order to provide auto‐protection capabilities, the autonomous behavior of the environment have to be secured. Actually, this security issue is addressed by defining an agent‐based Public Key Infrastructure (PKI) with X.509 certificates. The agent ensures then that the security functions are applied across all the distributed networks, where specific agents are responsible for conveying necessary information and certificates to local environments. Finally, the paper proposes a semantic‐based privacy management approach using ontologies to decide how privacy information is handled in the environment.