Computer networks make it easy to distribute digital media at low cost.  Digital rights management (DRM) systems are designed to limit the access that paying subscribers (and non-paying intruders) have to these digital media.  However, current DRM systems are tied to unicast delivery mechanisms, which do not scale well to very large groups.  In addition, the protection provided by DRM systems is in most cases not persistent, i.e., it does not prevent the legitimate subscriber from re-distributing the digital media after reception.

We have collected the requirements for digital rights management from various sources, and presented them as a set of eleven requirements, associated with five categories.  Several examples of commercial DRM systems are briefly explained and the requirements that they meet are presented in tabular format.  None of the example systems meet all the requirements that we have listed.  The security threats that are faced by DRM systems are briefly discussed.  We have discussed approaches for adapting DRM systems to multicast data transmission.

We have explored and evaluated the security protocols of a unicast distribution model, published by Grimen, et al.\, that provides ``persistent protection''.  We have found two security attacks and have provided the solution to overcome the discovered attacks.  Then we have proposed a more scalable architecture based on the modified model.  We call the resulting architecture persistent protection in multicast content delivery.  We present and formally validate the protocol for control and data exchange among the interacting parties of our proposal.