Theoretical Foundations of Identity & Trust on Decentralised Networks

We have recently submitted a body of theoretical work for publication that provides explicit descriptions of identity and trust (and the relationship between them), which underpin IdentityFlow Operations. For now, we give a short synopsis of this work here.

The current state of the project

IdentityFlow is currently being updated and bugs are being fixed. Soon we will begin work on some interesting new features:

  • JXTA binding implementation.
  • Hooks for trust (see below).

The current state of the project can be summarised as follows.

Updates on the project

First of all, there is a new sourceforge website, as will be obvious to anyone who was here before. The previous site was a temporary measure until we had a chance to engineer a proper web presence.

The IdentityFlow codebase is undergoing testing ahead of a full update, which is one of the reasons why there hasn't been an update in a while. There will be dramatically more activity over the next while, however, as we fix bugs, refactor and pursue our requirements as part of our OPAALS commitments.

IdentityFlow

IdentityFlow is a meta-model architecture and software implementation that describes and solves Identity Tasks. Identity tasks are tasks that are conducted by protocol driven interactions between between Actors across a network that solve an identity related problem. Identity tasks are modeled and developed as IdentityFlow Operations. Examples of identity tasks, and therefore potential operations, include Single Sign-On (SSO), Single Sign-Off, Attribute Requests and Claim Verification.

IdentityFlow has a layered architecture, which encourages a waterfall design approach, outlined as follows,

  1. High level representation of identity task accessible to domain specialists (may be non-technical experts)
  2. Top level technical representation of naive protocol flow and bindings
  3. Low level operation implementation, including profiles and bindings

IdentityFlow is intended to produce working implementations from the start, where detail and further implementation can be added later. It is intended to embrace specifications and standards, such as Security Assertion Markup Language (SAML) and Shibboleth, but not to drive them. It is also intended to produce the maximum design & code re-use by building operations from logical building blocks. To make the software more readily usable, it is SAML compliant (where possible) by default and a sample SSO operation and Redirect(GET)/POST binding implementations are included.

It is intended that a wide variety of operations, profiles and bindings be available in the future. A main design goal for IdentityFlow is also to integrate 'hooks' for a trust infrastructure into operations, and hence identity task protocol flows.

IdentityFlow is being developed as part of the OPAALS European Framework Programme 6 project as an open source project, available under the new BSD license.

Syndicate content