Public-key support for group collaboration

This paper characterizes the security of group collaboration as being a product not merely of cryptographic algorithms and coding practices, but also of the man-machine process of group creation. We show that traditional security mechanisms do not properly address the needs of a secured collaboration and present a research prototype, called NGC (next generation collaboration), that was designed to meet those needs. NGC distinguishes itself in the care with which the man-machine process was analyzed and shaped to improve the security of the whole process. We include a detailed analysis of the problem of binding a name to a key, traditionally thought to be the province of PKI, but we show that the SDSI local name concept produces a result with superior security to that produced by standard PKI.