Skip to Main Content
The notion of attribution is usually tied to identity: the ability to associate the originator of something with that data. This notion is both a simplification of the myriad aspects of attribution, and a masking of many different types of attribution under the rubric of “identity.” Current efforts at attribution of data in the network focus either on traceback of packets or signing data at various network layers, and management of the keys involved. As the Internet evolves into the next generation network, and people implement testbeds to facilitate that evolution, they can design support mechanisms for other forms of attribution. What types of attribution should those mechanisms support? This paper presents requirements for attribution that will be useful both in the next generation infrastructure and in the data it manages.