A practical method for authentication of remote computer systems

The preservation of physical security for computer systems has long been used as a foundation for computational trust. This requirement has come as a result of the use of shared secrets as host identifiers since there has been no empirical means of ascertaining the trustworthiness of a host in an absolute sense. Recent technological trends erode the usefulness of physical security as the basis for trust as well as open the possibility for a new trust model. This new model is not based on the traditional human qualities of trust such as reputation and accountability, but on capability, timeliness and determinism. A software-based procedure is demonstrated for establishing the trustworthiness of a remote computer system in the context of this new model. Deficiencies of this method are examined and used as a motivation for the development of a hardware implementation that offers greater durability of trust as well as ease of use.