Michael Nowatkowski

Certificate revocation list distribution in VANETs using Most Pieces Broadcast

This paper describes a novel idea for distributing certificate revocation lists (CRLs) in a vehicular ad hoc network (VANET) scenario. The idea, Most Pieces Broadcast (MPB), takes advantage of the two distinct channel types in VANETs while reducing contention for the wireless medium. Broadcast methods that reduce wireless medium contention in VANETs are highly desirable to assist in keeping the medium available for transmission of time-critical safety messages. MPB scales remarkably well using raptor coding to generate redundant file pieces.

Unified pseudonym distribution in VANETs

VANETs continue to mature and their installation is becoming a reality. Many ideas have been exchanged on how best to balance privacy and security. The use of pseudonyms has been almost universally accepted as a critical part of this equation. Simulated results, using ns-3, demonstrate the need for more than a single road side unit (RSU) contact for pseudonym refill due to the limited number of certificates that can be issued in that transit. This paper provides a universal protocol using multiple RSUs and multiple service channels for the distribution of pseudonyms for refill, intra-regional, and inter-regional purposes.

Cooperative Certificate Revocation List Distribution Methods in VANETs

This paper discusses two new methods for distributing certificate revocation lists (CRL) in a vehicular ad hoc network environment using cooperative methods. The main purpose for using cooperative methods is to attempt to reduce the number of collisions in the dedicated short range communication (DSRC) broadcast environment. The reduced number of collisions will increase the effective throughput of the medium. The first method uses a polling scheme to determine which nodes possess the most number of CRL file pieces. The second method takes advantage of the multiple service channels available in DSRC. Both methods use a form of coding that reduces the impact of the piece problem. Both methods are compared to the Code Torrent method of file distribution.

Key terrain in cyberspace: Seeking the high ground

In military doctrine, key terrain refers to areas which, if seized, afford an advantage to an attacker or defender. When applied to geographic terrain, this definition is clear. Key terrain might include a hill that overlooks a valley an enemy wants to control or a crossing point over a river that must be traversed before launching an attack. By definition, dominance of key terrain is likely to decide the overall outcome of a battle. While cyber key terrain is similar to geographic key terrain in some ways, there are also significant and often counterintuitive differences. Some consider cyber terrain to be tied to a physical location and to be represented in cyberspace by routers, switches, cables, and other devices. We will argue that key terrain in cyberspace exists at all of the cyberspace planes, which include the geographic, physical, logical, cyber persona, and supervisory planes [1]. In many cases, features of cyber terrain will not be tied to a specific location, or the geographic location will be irrelevant. In this paper we deconstruct and analyze cyber key terrain, provide a generalized framework for critical analysis, and draw parallels between cyber and physical key terrain while providing examples of key terrain in cyber operations. During a cyber operation, an analysis of key terrain will aid in the strategy and tactics of both the offense and the defense. During peacetime, an understanding of cyber key terrain can be employed broadly, ranging from helping a system administrator focus scarce resources to defend his network all the way to allowing nation-state militaries to develop long-lasting and effective doctrine.

Framework to Support Per Second Shifts of Pseudonyms in Regional VANETs

Vehicular ad hoc networks (VANETs) provide the means to add convenience, services, and safety to the road. This paper introduces a means to adapt the concepts of Public Key Infrastructure for a VANET environment under the worst case scenario of a pseudonym shift every second. Regions are used to scale down the size of certificate revocation lists (CRLs), administrative overhead, and the search space to link a message to its originator. Regions also provide a means for expansion of the geographical area covered and provides the ability to balance geographic mobility with privacy. The framework outlined below can then be modified to the decided pseudonym shift frequency and/or extended for other purposes.

The effects of quality of service on vehicular network pseudonym distribution in congested urban environments

Vehicular networks are meant to exist wherever the road will take them. This includes small towns, rural highways, suburbs, downtown urban centers, and urban highways. The density of vehicles varies greatly across these environments. This work looks at the effects of implementing Quality of Service (QoS) as well as a relatively similar method that inserts stochastic delays in pseudonym (PN) transmission in the two urban settings of a downtown grid and an urban highway, both under heavily congested conditions. The simulated results (using ns-3) are compared to previous work that examined communication suppression (as opposed to priority as in this work). Four metrics are used for method comparison: average overall background data throughput, average overall PNs distributed, maximum number of PNs distributed, and the distribution of the PNs across the vehicles as a function of need. While some of the results obtained were expected, the overall conclusion that implementing quality of service, or even a simplistic imitation, can significantly improve the overall data throughput and provide more PNs is an interesting result.

Simulation analysis of the United States Military Academy Reception Day

Each June approximately 1200 new cadets (NCs) are welcomed to the United States Military Academy (USMA) during Reception Day (R-Day). Amid the mass shuffling of bodies and the yelling of the upper-class cadet cadre, all 1200 NCs must process completely. Changes are made to the in processing system in attempt to make it more efficient. However, the effect of these changes can only be gauged once a year during the following R-Day. In an attempt to expedite the refinement process, the R-Day administrators approached our design team to create a simulation model that could be used to analyze the effect of proposed changes to the system prior to R-Day 2002. Using ProModel/spl reg/ simulation software, our team created a simulation of the in processing system and conducted a statistical analysis of the results in order to recommend improvements to the structure of the system.