Title: Modeling Container Security Operations at the APM Terminal (Demo only)

Tayfur Altiok, Benjamin Melamed and Abdullah Karaman, Industrial Engineering, Rutgers University

We demonstrate simulation models developed by Rutgers University's Laboratory for Port Security (LPS) team to study the container inspection operation and quantify its performance. We present a detailed computer simulation model, named PORTSIM, of the VACIS operations at a port. The model accurately captures the vessel arrival processes, the current Customs and Border Protection (CBP) container inspection process, and the terminal operations. PORTSIM predicts a number of critical performance measures of the container inspection operation.

Title: Risk Analysis and Modeling of the Vessel Traffic in the Istanbul Strait (Demo only)

Tayfur Altiok and Ozgecan Uluscu, Industrial Engineering, Rutgers University

The Istanbul Strait is identified as one of the six bottleneck points for sea transport by the U.S. Energy Information Administration (EIA) due to its narrow shape and high traffic volume. Approximately ten percent of 50,000 vessels that pass through Istanbul Strait every year are oil or liquefied natural gas tankers, posing risk of serious environmental consequences if an accident were to occur. We demonstrate a simulation model of the transit vessel traffic in the Istanbul Strait. The model was built for risk analysis purposes to study the temporal behaviors of the situational variables and accident risks over time. The ultimate goal of the project is to arrive at operational policies that mitigate risk to the environment, the Istanbul residents, and the economy while maintaining an acceptable level of vessel throughput.

Title: The Lasso with Attribute Partition Search (Poster Presentation)

Suhrid Balakrishnan and David Madigan, Statistics, Rutgers University

Regression and classification problems involving ordered attributes (for example where some input patterns are a set of samples from time series variables) arise in application domains like finance and epidemiology. In such cases, identifying and building models involving predictive runs of the attributes leads to highly interpretable models that may also be very accurate. We present an approach to build such models using a variant of the Group Lasso (Yuan and Lin, 2006).

Session Title: Preparing for Bioemergencies

Dr. Ebright will provide a general overview of biosafety, biosecurity, and dual-use implications of bioweapons-agents research, of the existing regulatory framework, and of gaps in the existing regulatory framework. Dr. Kahn will speak about Princeton's Program on Science and Global Security, and its two primary projects the past four years: the Carnegie Corporation of NY-supported biodefense seminar series and the Josiah Macy, Jr.-supported four state biopreparedness study that has just been completed. Dr. Connell will discuss changes in laboratory environment post 9/11 and how these changes have (or have not) changed the process of academic research. Time permitting, Mr. McCluskey will engage the panel and the audience in a discussion on biosecurity, biopreparedness, and bioemergencies.

Session Speakers and Titles:

Title: Optimization Models for Container Inspection

Endre Boros, Rutgers Center for Operations Research (RUTCOR), Rutgers University

Finding ways to intercept illicit materials, in particular weapons, shipped via the maritime transportation system is an exceedingly difficult task. Until recently, even with increased budget and emphasis, and rapid development of technology, it is still only a very small percentage of ship cargoes are inspected. There is a great need to improve the efficiency in the current inspection processes. Current inspection methods involve a variety of physical sensors, typically employed in an ensemble. Since different sensor technologies have different operating characteristics (inspection cost, detection rate, etc.) it is important to apply sensors in an appropriate sequential manner. Following the model of Stroud and Saeger (2004) we can view an inspection policy as a decision tree, the nodes of which correspond to sensors.

The problem of building an "optimal" decision tree (which e.g., maximizes the overall detection rate subject to inspection cost limits) is a very difficult mathematical problem. We developed recently a polyhedral description of all possible decision trees, and as a consequence, very powerful off the shelves linear optimization packages can be used for the design of optimal inspection policies. This approach leads to the important observations that using multiple thresholds and combinations of several decision trees are improving performance without increasing expenses.

This is joint research with L. Fedzhora, P.B. Kantor, K. Saeger, and P. Stroud.

Title: Underwater Threats to Our Urban Ports

Michael S. Bruno, Center for Maritime Systems, Stevens Institute of Technology

Stevens Institute of Technology has for several years been operating an integrated ocean observation and forecast system to support maritime operations in the Port of New York and New Jersey. This system was recently enhanced to support the creation of a real-time Maritime Security Laboratory in the Hudson River just east of Castle Point. Ongoing studies supported by the Office of Naval Research have focused on underwater threats, with a view toward developing adaptable, environmentally sound surveillance systems capable of detecting threats in highly complex, urban estuary environments.

Title: Port Security in Praxis

Michael J. Chumer, Information Systems, New Jersey Institute of Technology (NJIT)

HSPD 13 (Homeland Security Presidential Directive 13) outlines the strategic vision for maritime security and begins to define the all encompassing concept of a Maritime Domain Awareness (MDA). HSPD 13 also suggests that work is being done across US ports in the areas of container security, proliferation security, megaports initiative, advance information, Customs Trade Partnership Against Terrorism, joint harbor operations and Operations Safe Commerce. In addition, GAO-06-933T further addresses the growing importance of Information Sharing and states that, "Ports are sprawling enterprises that often cross jurisdictional boundaries; therefore, the need to share information among federal, state, and local agencies is central to effective prevention and response". This extends the areas of importance within maritime domains by addressing the uniqueness of responsibilities of certain ports. For example the Port Authority New York/New Jersey is not only concerned about the maritime domain but also bridges, tunnels, PATH tubes, and four airports.

The talk will focus on how the Port Authority New York/New Jersey is addressing information sharing as well as situational awareness leading to the development of a Common Operational Picture (COP) through their prototype Remote Information Joint Awareness Network (RIJAN) by describing its use in the 2005 Coalition Warrior Interoperability Demonstration.

The talk will conclude by suggesting both social and technical research opportunities in the area of information sharing and Federal, State, and local collaborations.

Title: Emergency Communication in the Public Domain

Eamon Doherty, Fairleigh Dickinson University

There are a variety of organizations that provide limited service regarding emergency data and voice communications without cost to local and state governments in New Jersey. These organizations are Amateur Radio Emergency Communication (ARES) and Radio Amateur Civil Emergency Service (RACES). Many of the members have a proven record in helping with emergency communications and mapping areas in the Katrina Disaster. ARES and RACES members are comprised of academics, scientists, policemen and others with an amateur radio license and an interest in emergency communication. MARS, military amateur radio service can provide text and voice messages between the military and civilians in time of emergency too free of charge. Lastly we will examine the Unisys Mobile CommHub which has been displayed to academics at universities for feedback and allows secure VPN connections to databases in types of an emergency. We will discuss how this device can be used with portable power in a place with severely damaged infrastructure and can be used to help provide limited usage of GIS applications, streaming video, and telephony with the State Emergency Operations Center (EOC).

Title: Location-based Threat Sensing and Emergency Notification (Poster/Demo)

Dennis Egan, Judith Jerkins, K.R. Krishnan, Hanan Luss, Marc Pucci, Arnold Neidhardt, and John Wullert, Telcordia Technologies

This poster summarizes two ongoing research projects at Telcordia Technologies. The Cellular Network Sensing project is sponsored internally by Telcordia. It is devoted to mathematical and modeling issues underlying the use of the cellular network to acquire sensor data to augment information available from a relatively sparse, fixed network of sensors (e.g. SensorNet being developed at Oak Ridge National Laboratory). This work assumes that small, inexpensive sensors capable of sensing chemicals, gases, radioactivity, etc. are widely deployed on ordinary cell phones. The focus of the project is fusing data from such sensors to reduce false alarms and uncertainty that may arise from a sparse sensor network. In addition to data fusion, the project involves work on algorithms that attempt to optimize the selection of mobile sensors to activate and poll.

The "Red Cell" project is sponsored by the Institute for Defense and Homeland Security (IDHS). Its goal is to use the cellular infrastructure to deliver alerts and emergency information to people in a way that is specific to their locations and emergency management roles. The Red Cell system is intended for use in a variety of natural disasters, terrorist attacks, Amber Alerts, and expeditionary force protection scenarios. Telcordia as part of a team headed by the West Virginia High Tech Consortium Foundation has produced a Concept of Operations (CONOPS) for Red Cell.

Title: A Formal Approach To Practical Network Security Management (Poster Presentation)

Sudhakar Govindavajhala, Computer Science, Princeton University

When a system administrator configures a network so it is secure, he understands very well the users, data, and most importantly the intent---what he is trying to do. However, he has a limited understanding of the mechanisms by which components interact and the details of each component. He could easily misconfigure the network so a hacker could steal confidential data. In addition to this complexity, about one hundred new security vulnerabilities are found each week, which makes it even more difficult to manage the security of a network installation---because of the large number of program vulnerabilities and challenging time constraints. Even professional administrators find this a difficult (impossible) task. How does one enable the system administrator to securely configure the network with a limited understanding of its components, program bugs and their interactions?

The solution is a security analysis framework that modularizes information flow between the system administrator, security expert and the bug expert. The administrator specifies what he is trying to do, the security expert specifies component behavior, the bug expert specifies known bugs. We developed a rule based framework---Multihost, Multistage, Vulnerability Analysis (MulVAL)---to perform end-to-end, automatic analysis of multi-host, multi-stage attacks on a large network where hosts run on different operating systems. The MulVAL framework has been demonstrated to be modular, flexible, scalable and efficient. We used the framework to find serious configuration vulnerabilities in software from several major vendors for the Windows XP platform.

Title: Developing Treatments for Chemical Weapon Attacks: the UMDNJ-Rutgers University CounterACT Research Center of Excellence (Poster Presentation)

Joshua P. Gray (1), Jeffrey D. Laskin (2), Donald R. Gerecke (1), Marion K. Gordon (1), Patrick J. Sinko (1), Ned Heindel (3), Diane E. Heck (1), (1) Pharmacology and Toxicology, Rutgers University; (2) Environmental & Occupational Medicine, UMDNJ Robert Wood Johnson Medical School; (3) Department of Chemistry, Lehigh University

The National Institutes of Health (NIH) has recently awarded a five-year, $19.2 million grant to UMDNJ-Robert Wood Johnson Medical School and Rutgers University to support the creation of a Center of Excellence focused on the development of new and improved medical countermeasures against high priority chemical threats. This Center will focus on developing drugs to treat sulfur mustard poisoning, a potent chemical warfare vesicant. Although it has been studied for more than 80 years, the mechanisms mediating its actions as a vesicant remain unknown; moreover, to date, there are no effective medical countermeasures for exposure to warfare vesicants. In collaborative studies with Battelle Memorial Institute, members of the Center are optimizing lead compounds to determine pharmacophores active against sulfur mustard. Studies are being initiated to evaluate the efficacy of these potential countermeasures in model systems of sulfur mustard toxicity. In addition, Research and Development Projects are underway to identify specific mechanisms of action of sulfur mustard and potential new targets for therapeutic intervention in three major vesicant targets: the eye, the skin and the lung. Investigators on these projects work closely with a Pharmacology and Drug Development Core and a Medicinal Chemistry and Pharmaceutics Core with considerable expertise in drug development, providing insights to facilitate the development of sulfur mustard countermeasures. The research laboratories of each Core group are members of the Environmental and Occupational Health Sciences Institute (EOHSI), a facility jointly sponsored by UMDNJ-Robert Wood Johnson Medical School and the School of Pharmacy at Rutgers University. A Training and Education Program directed at health care providers at Rutgers University, UMDNJ-School of Public Health, UMDNJ-Robert Wood Johnson Medical School and the Health Sciences Program at Lehigh University has been established. The principal investigators include Dr. Jeffrey Laskin as the overall director of the Center; Dr. Donald Gerecke at Rutgers as co-director of the Center; Drs. Marion Gordon, Joshua Gray, Diane Heck, Debra Laskin, and Patrick Sinko, also at Rutgers; and Dr. Ned Heindel at Lehigh University. Investigators at Robert Wood Johnson Medical School and Rutgers are members of the Environmental Occupational and Health Sciences Institute.

Title: A Hierarchical Modeling and Planning Support System for Responding to Emergency Events

Sastry Isukapalli and Panos Georgopoulos, Environmental and Occupational Health Sciences Institute

A hierarchical modeling and planning support system is under ongoing development, that provides estimates of potential impact due to the release of chemical, biological, or nuclear agents in emergency events of various types (industrial accidents, terrorist acts, etc.) This system provides contaminant concentration estimates at various levels of detail: (1) rapid estimates through the use of simplified atmospheric dispersion models such as ALOHA, HPAC, CALPUFF, etc., (2) complex temporal and spatial profiles of the agents in the atmosphere through a highly detailed assessment using Computational Fluid Dynamics (CFD) models, and (3) an intermediate level assessment that provides estimates of not only the mean values of concentrations at a location but also the essential statistical properties of fluctuations of these concentrations, thus providing information for estimating risk in a probabilistic framework. The system is GIS-based, and enables comparative evaluation of outcomes from multiple models in a consistent manner, and also provides linkages to multiple supporting databases. Current enhancements to the system are focused on linking contaminant concentration estimates with population distribution, demographics, and human activity patterns in order to estimate exposures and doses, and associated risks. Examples of case studies include biological agents such as anthrax, chemical warfare agents such as sarin, and industrial chemicals such as chlorine.

Title: Securing Future Wireless Networks: Challenges and Strategies

Pandurang Kamat, Computer Science and WINLAB, Rutgers University

There has been considerable advancement in wireless technologies over the past five to ten years. We have witnessed the rapid commercial adoption of WiFi technologies, an explosion in research activity in ad hoc and sensor networks, and the development of the next generation of highly-programmable radio platforms. As these systems mature, they will become the target for misuse?not only will they be the target for attacks, but they will also be the very means for launching new forms of security attacks. In this talk, we will survey different wireless technologies that are being developed and identify a set of security challenges for each of these technologies. We will then outline a collection of techniques that can be used to cope with these risks?ranging from improved protocol designs to employing trusted computing techniques to ensure trustworthy operation of wireless devices.

Title: Needs and Opportunities in Exposure Characterization for Homeland Security Events

Paul Lioy and Panos Georgopoulos, Environmental and Occupational Medicine, UMDNJ-Robert Wood Johnson Medical School

The exposure to segments of the general population and worker populations to highly toxic agents during a terrorist event or catastrophic accident will vary in intensity and resultant potential victims. Preparedness for either of the above needs to include more accurate definition of the types and patterns of activities that alter contact with toxicants among these populations. Research needed to characterize the above will be described, including examples of current studies.

Title: Comprehensive Training Assessment and Gap Analysis related to the CBRNE threat in New Jersey (Poster Presentation)

Brendan McCluskey, Mitchell Gayer, Jason Emmel, James Smith, Kathleen Wioland, Mary Mickelsen, Robert Matticola, UMDNJ

The New Jersey CBRNE Center for Training and Research at UMDNJ (CBRNE Center) conducted an assessment and analysis of CBRNE-related training that is available currently to New Jersey's emergency responders. This analysis was intended to identify strengths and weaknesses in the systems in place to provide the required training. In order to accomplish the goals of the project the following steps were undertaken:

The project culminated with a comprehensive report detailing each of the above subject areas as well as a comprehensive catalog of all CBRNE-related training that is currently available in New Jersey.

Title: Applications of Mid IR Technologies for Environmental Monitoring and Homeland Security (Poster/Demo)

Joe Montemarano and Claire Gmachl, PRISM/ NSF-ERC on Mid-InfraRed Technologies for Health and the Environment (MIRTHE)

Quantum Cascade Lasers have been used in telecom applications (1.5 micron) for more than a decade. By shifting to the Mid IR (3-30 micron portion of the spectra) a much broader range of chemical opportunities can be detected with several orders of magnitide sensitivity improvement. When combined with improved detectors and signal processing, these new mid IR systems offer impressive sensitivity (ppb and lower), small size, real-time monitoring and low cost. Initial applications target environmental monitoring of indoor and public spaces, especially for volatile organics, and a wider range of compounds and particulates of interest in homeland security applications will be pursued. Princeton leads this newly established NSF ERC in partnership with JHU, UMBC, CUNY, Rice and TAMU, and a wide range of industry members.

Title: Modeling of Transportation Evacuation Problems for Better Planning of Disaster Response Operations

Kaan Ozbay and Mustafa Anil Yazici, Civil and Environmental Engineering, Rutgers University

In this presentation, we will first provide a brief review of the previous work dealing with transportation evacuation as a result of natural or man-made emergency conditions such as hurricanes, earthquakes, or sabotage. Then, we will present major issues related to the modeling of these transportation evacuation problems (Ozbay et al. [2,3]). Finally, we will briefly describe a network-optimization model that we have developed to study the less frequently addressed problem of the determination of the optimal location and capacity of shelters, under hurricane conditions. Our model is a stochastic extension of the Cell Transmission (CTM) system optimal dynamic traffic assignment (SO DTA) model first proposed by Ziliaskopoulos et al. [4]. The major source of the uncertainties in our model is the probabilistic nature of link capacities due to the impacts of events that can completely or partially damage highway links. Our results show that introducing probabilistic link capacities can considerably alter network-wide flow and shelter utilization patterns [1]. Thus, if the planners decide to consider the predictions of a deterministic model only, they may face the risk of underestimating evacuation times and having inadequate levels of emergency supplies in some shelters.

[1] Ozbay K., Yazici M.A. Determination of Hurricane Evacuation Shelter Capacities and Locations with Probabilistic Road Capacity Constraints, Accepted for Presentation at the 86th Annual Meeting of the Transportation Research Board, Washington, D.C., 2007.

[2] Ozbay K., Yazici M.A. and Chien S. I-Jy. Study Of The Network-Wide Impact Of Various Demand Generation Methods Under Hurricane Evacuation Conditions. Proceedings of the 85th Annual Meeting of the Transportation Research Board, Washington, D.C., 2006a.

[3] Ozbay K. and Yazici M.A., "Analysis of Network-wide Impacts of Behavioral Response Curves for Evacuation Conditions", Proceedings of the IEEE ITSC 2006 Conference, 2006b.

[4] Athanasios K. Ziliaskopoulos, A Linear Programming Model for the Single Destination System Optimum Dynamic Traffic Assignment Problem, Transportation Science, Vol.34, No.1, 2000, pp 37-49.

Title: From Stochastic Optimization to Housing Permits: The Navigating the Pot-holes to Implementation in the Real-world

Warren Powell, Operations Research and Financial Engineering, Princeton University

One dimension of planning for emergency response is to allocate resources (FEMA trailers, fuel, water, medical personnel, vaccines) so that we can quickly respond to whatever emergencies arise (flu outbreaks, attacks on ports or the power grid, hurricanes). The academic community has developed a rich array of modeling and algorithmic strategies to "solve" these problems, but the path to real-world implementation can involve surprising pot-holes. Perhaps one of the most notorious examples is the saga of the FEMA trailers. We will use this and other examples to highlight some of the challenges we face in having an impact on actual emergencies.

Title: Plans for New DHS Center based at Rutgers University

Fred Roberts, Director of DIMACS and of New Center, Rutgers University

The US Department of Homeland Security has established an Institute for Discrete Sciences based at Lawrence Livermore National Laboratory and has awarded to Rutgers a university-affiliated center award to collaborate with IDS and to lead a team of new university centers. The new DHS center at Rutgers is called the Homeland Security Center for Dynamic Data Analysis (DyDAn). It is a partnership of Rutgers, Princeton, RPI, Texas Southern, Texas State - San Marcos, AT&T Labs, and Bell Labs, and will lead four new centers, with others based at University of Pittsburgh, University of Illinois, and University of Southern California. This talk with describe the plans for DyDAn research and educational programs.

Title: Research at Stevens Maritime Security Lab (Poster Presentation/Demo)

Hady R. Salloum, Stevens Institute of Technology

The poster will display information and research activities that are currently conducted at the MSL lab and how it relates to Homeland Security.

Title: Emergency Planning as a Continuous Multiplayer Game

Michael Chumer and Murray Turoff, Information Systems, New Jersey Institute of Technology (NJIT)

Currently there are serious problems with multi-organizational abilities to plan reliable responses to emergencies. The response to any wide scale disaster involves the necessity for many individuals from different organizations or organizational units to come together to form a virtual organization that has the characteristics of a High Reliability Organization (HRO). HROs exhibit cooperation, trust, and mutual respect for expertise. In addition, there is the recognition of the need to uncover and immediately deal with mistakes and errors in processes without degrading the cohesiveness of the resulting team.

Given the diversity of organizations (government at all levels, humanitarian, and private), the most feasible way to accomplish this is through an ongoing dialectic multi-team asynchronous game to play out alternative defense and offense scenarios on a continuous basis. This would involve a total of approximately half a day a week of time for all those who would make up the command and control team for an actual disaster. It would allow the primary responders in different organizations to be the actual source of a detailed and evolving set of plans for the most likely threats.

With such simulations and the evolving discussions it is far more likely that such plans will be far more realistic than one time "paper plans" devised by higher level committees. The built-in competitive process will serve to uncover mistakes and errors that can then be use for analysis and plan improvement. Our current and prior work in this area at NJIT lends support to the above observation.

Title: Security of Data in the Wireless Environment

Patrick White, Stevens Institute of Technology

The wireless communication channel, unlike the fixed wire network, is generally more exposed to eavesdroppers. While encryption can reduce this limitation, the result is often an increase in overhead, which can be a significant problem for the typical band-limited wireless channel. We offer several alternatives that have low overhead. They include using high-frequency, mm-wave signals that have limited propagation to using multi-network radios where frequency bands, coding/modulation, etc., can be modified with each packet.

Title: Collaborative Scenario Creation System - Making Scenario Creation Public (Poster Presentation)

Xiang Yao, New Jersey Institute of Technology

The poster introduces a new method and Group Support System (GSS) which allow disaster planners and managers in different locations to discuss and create disaster scenarios collaboratively and asynchronously. The objective of this approch, using a shared collabortaive database, is to increase scenario creativity, credibility, and realism by allowing larger and diverse groups of individuals to be involved in the planning process. By spending short periods of time every week the planning can be treated as a continuous operation and aid the formation of a virtual team that can come together and act as a team in a real emergency. This project is directed by Professor Murray Turoff.

Title: Decentralized Trust Management based on the Reputation of Information Sources

Meng Yu and Wanyu Zang, Computer Science, Monmouth University

In a distributed system, classical PGP or PKI methods provide means to establish trust based on credentials. The problem is that the credentials are not restricted to specific operations. Furthermore, the structure of Certification Authority (CA) is centralized, which is vulnerable to be a single point failure even if backups are present. In this paper, we propose a decentralized method to establish trust based on the reputation of information sources. We focus on quantified trust levels of information pieces, in addition to the trust of information sources. We use transfer functions to describe the lost of information during transfer. We also use a summary function to simulate the summary of information from multiple information sources. Finally, we studied how to address coalition problem in a distributed system. The effectiveness of our method is evaluated in this paper.

Title: Homeland Security Instructions for Nuclear Medicine Patients at the New Jersey Medical School, UMDNJ (Poster Presentation/Demo)

Lionel Zuckier, V.K. Lanka, L.C. Thelin and F. Grippo, New Jersey Medical School, Office of Radiation Safety Services

Following 9/11, concern regarding the possibility of terrorism involving radioactive materials led to the distribution of 10,000 personal radiation detectors to homeland safety personnel by the Department of Homeland Security. Inevitably, reports appeared of patients undergoing diagnostic or therapeutic nuclear medicine procedures triggering homeland security detectors, especially in the greater New York region. For this reason, on February 10, 2003 the Society of Nuclear Medicine (SNM) recommended that patients be provided a letter which contains details regarding their procedure and a contact person who can be reached for verification. On December 9, 2003, the Nuclear Regulatory Commission (NRC) advised licensees to inform patients of this potential problem, and to provide them with written information for law enforcement use, stating that the patient poses no danger to the public, and that the administration is allowed by NRC medical use regulations.

University Hospital, UMDNJ is the center of referral for many of the State's most advanced medical services and specialty care programs. The Department of Nuclear Medicine at University Hospital performs some 6,000 diagnostic and therapeutic procedures annually. Patients administered longer-lived diagnostic and therapeutic radionuclides are given documentation regarding their medical administration of radioactive materials, conforming to the suggestions of the NRC and SNM. Experimental measurements and modeling of excretion have allowed us to define the length of time during which patients may alarm detectors, thereby providing guidance regarding the period of time that patients should retain their documentation. Most recently, these calculations have been ported to an educational web site which contains an on-line calculator to refine estimates of the period of time patients may trigger an alarm and which can be used to print a letter of documentation.