Open architecture networks provide applications with fine-grained control over network elements. With this control comes the risk of misuse and new challenges to security beyond those present in conventional networks. One particular security requirement is the ability of applications to protect the secrecy and integrity of transmitted data while still allowing trusted active elements within the network to operate on that data. This paper describes mechanisms for identifying trusted nodes within a network and securely deploying adaptation instructions to those nodes while protecting application data from unauthorized access and modification. Promising experimental results of our implementation within the Conductor adaptation framework will also be presented, suggesting that such features can be incorporated into real networks.
Jun Li, Mark Yarvis, Peter L. Reiher