Networks and computers often hold a company's most precious and costly comm
odities-its intellectual properties and proprietary data. If the network an
d computer vulnerabilities of any large network-such as the Lucent Technolo
gies intranet-are not identified and mitigated they could enable an intrude
r to seriously compromise the security of a company's network, computers, a
nd data. Given the voluminous nodes and hosts in Lucent's intranet, it is n
ot operationally feasible to scan the entire network to search for vulnerab
ilities. In this paper, we describe a methodology for statistical sampling
and analysis, combined with a network and host security discipline for deve
loping Lucent's cyber security profile in an effective, efficient manner. W
e have also developed a methodology for correlating vulnerabilities in and
among the network and operating systems. We found the distribution of high-
risk vulnerabilities to be very concentrated. Through root cause analyses,
we developed a focused plan for mitigating vulnerabilities effectively and
efficiently. These patent-pending methodologies will enable cyber security
management in a large networked environment and, if properly deployed, will
be to network security what quality control charts are to manufacturing pr
ocesses.