Enforcing policies in pervasive environments

Abstract

This paper presents an architecture and a proof of concept implementation of a security infrastructure for mobile devices in an infrastructure based pervasive environment. The security infrastructure primarily consists of two parts, the policy engine and the policy enforcement mechanism. Each mobile device within a pervasive environment is equipped with its own policy enforcement mechanism and is responsible for protecting its resources. A mobile device consults the nearest policy server, notifies its current state including its present user, network presence, other accessible devices and location information if available. Using this information the policy server queries the "Rei" engine to dynamically create a policy certificate and issues it to the requesting device. The system wide policy is described in a semantic language "Rei", a lightweight and extensible language which is able to express comprehensive policies using domain specific information. The "Rei"policy engine is able to dynamically decide what rights, prohibitions, obligations, dispensations an actor has on the domain actions. A policy certificate is created and issued to the device. The policy certificate contains a set of granted permissions and a validity period and scope within which the permissions are valid. The policy certificate can be revoked by the policy enforcer based on expiration of the validity period or a combination of timeout, loss of contact with an assigned network.

Preserving User Privacy and Security in Context-Aware Mobile Platforms

“…1) User policy editor: As described by [20], policies may be enforced indefinitely or for a certain time period based on a policy certificate validity period or a combination of timeout or loss of contact with an assigned network. However, the user has the option of modifying or adding rules to the policy through the interface shown in Figure 11.…”

Authentic Refinement of Semantically Enhanced Policies in Pervasive Systems

“…Examples for policy frameworks based on semantic policy representation are KAoS [18] and Rein(n) [7,5]. Rein(n) is based on rules written in the RDF-based rule language Notation3 1 and has been used in [11] to implement a policy-controlled pervasive system. KAoS represents even the policy structure itself in description logic (i. e. in DAML, the antecedent of OWL) and makes use of the Java Theorem Prover 2 in order to integrate rules and variables which are not supported by plain DAML / OWL.…”

Towards a smart world and ubiquitous intelligence: A walkthrough from smart things to smart hyperspaces and UbicKids

“…Since proactive computing has put emphasis on humansupervised automatic operations rather than human-computer interactions, it is therefore rational and necessary to have a specific part in the UbicKids architecture to manage the corresponding supervision information that may include task, plan, policy, rule and background knowledge [3], [18]. Such information is cataloged as some kind of contexts by some researchers.…”