Faculty Abstracts

“Security Enabled Wireless Network Architecture”

Tao Zhang, Ph.D.,  Associate Professor of Computer Science, NYIT
Farshid Delgosha, Ph.D.,  Assistant Professor of Electrical Engineering, NYIT
Michael Colef, Ph.D., Associate Professor of Electrical Engineering, NYIT

Our research aims to develop a new security enabled network architecture for monitoring patients or elderly individuals in hospitals or nursing homes. In this architecture, wireless sensor networks (WSNs), which are comprised of miniature sensors that measure vital signs, are utilized. In this application scenario, every patient is equipped with a medical WSN. The wireless sensor nodes in each WSN collect medical data from a patient and send it to a sink that is in the communication range of all sensor nodes. Every patient is assigned a single sink that compiles a report on the patient's health status. All sinks construct a multi-hop wireless mesh network (WMN) that forwards medical data to a server. The medical information collected by this network is made available to doctors and nurses for treatment and monitoring purposes.

The key novelties of this new network architecture include: (1) integration of distributed WSNs into the WMNs to provide effective data processing and access for medical diagnosis and patient monitoring applications; (2) construction of a high-throughput backbone in WMN that emphasizes the transmission of a large volume of data; and (3) design and implementation of a set of service-oriented protocols to provide services to users with low energy consumption, low latency, and enabled data security.

Within this new paradigm, we investigate many practical aspects of the network design and its hardware implementation. We expect that the novel design guidelines enabled by this project will inspire and aid engineers and scientists in researching, developing, and deploying more advanced wireless networks that are becoming increasingly important to our society.

 

“Hackers, Spammers, Con artists, Identity Thieves... What is on the cyber criminals’ mind?”

Ziqian Dong, Ph.D.
Assistant Professor of Electrical Engineering, NYIT

  • Setting up scams does not cost much
  • The impact can be huge – from individual user, to large corporation, even a whole country
  • People make money doing it
  • Proof that I’m better than others – to satisfy one’s ego
  • The chance of getting caught for what you’ve done is LOW
  • The chance of getting prosecuted is MINIMAL

Network forensics is a category of forensic science that is different from popular forensics as displayed in the TV series CSI and NCIS. It traces digital fingerprints of cyber criminals. Network forensic scientists are required to understand network protocols, IT infrastructure, underlining technologies behind corporate emails, servers, websites in order to find traces of cyber criminals. They use tools to retrieve and analyze information on IP addresses, domain names, proxies, email headers, event logs and etc. They may also need to have knowledge of psychology to help them build forensic investigation tools .

This research investigates techniques to locate an Internet host based on real-time network delay measurements, avoid DDoS attacks on a network level by implementing differentiated scheduling schemes in network switches and routers and building a test bed to test security and vulnerability of VoIP calls. 

  

“Security and Energy Efficiency of GPU-based Cloud Systems”

Xiaohui (Sean) Cui, Ph.D.
Assistant Professor of Computer Science, NYIT

Enterprise level systems are designed to provide mission critical services to a large number of end users. These systems must respond to enormous numbers of requests to complete relatively small tasks.  Organizations from the government and private sector require that these systems be able to responds quickly, remain secure, be always available, and provide a high energy efficiency. This research shows that techniques which integrate computing on the graphics processor unit (GPU) with a cloud based system represent one way to efficiently meet the high demands of enterprise systems. Special techniques to help further increase the availability, performance, security, and efficiency of such systems are investigated in this paper.

 

“Algebraic Cryptography: A Computationally Efficient Scheme”

Farshid Delgosha, Ph.D.
Assistant Professor of Electrical Engineering, NYIT

Algebraic cryptography is a promising approach to the design of computationally efficient public-key schemes. Cryptosystems designed using this approach are suitable for implementation in devices with limited computational power such as small-size cell phones, personal medical monitoring devices, and smart cards. This research proposes a new technique for the design of public-key cryptosystems and digital signature schemes using paraunitary matrices. The computational complexities of cryptosystems based on this technique are much lower than that of RSA while maintaining the same level of security. A notable feature of paraunitary-based cryptosystems is that they are easily implementable on byte-oriented hardware.

 

“Medical Terminologies and EHRs”

Huanying (Helen) Gu, Ph.D.
Associate Professor of Computer Science, NYIT

Terminologies to encode clinical and administrative healthcare data are ubiquitous in today’s health information technology environment. Systematized Nomenclature of Medicine - Clinical Terms (SCT)  has been endorsed as a premier clinical terminology by many international organizations. It is slated to play a significant role in the HITECH initiative to adopt Electronic Health Records (EHRs) by providing standardized encodings of healthcare data. SCT will also help in achieving the long-term goals of improved quality of care and reduced costs by enabling secondary use such as decision-support, data mining, etc. SCT is an integral component of standardization in health information technology, which can facilitate sharing and integration EHRs.

However, there are indications that at the moment SCT is not optimally structured for its intended use in EHRs by practicing healthcare professionals. Aspects of SCT’s coverage, representation, and organization may interfere with its success. Given SNOMED's expanding content and attendant complexity, its quality assurance is a critical task facing SNOMED's maintenance personnel. There is assuredly a need to provide automated and semi-automated methodologies for aiding editors in this endeavor. We have developed different innovative algorithms to rapidly and efficiently identify—with high probability— deficiencies in terminologies to warranting expert review, which will enable better delivery and utilization of healthcare resources—now and in the future.  We also investigate security and privacy issues in EHR sharing and integration.  

 

“Sink Shift – An Easy Approach to Extend WSN Lifetime”

Wei Ding, Ph.D.
Assistant Professor of Computer Science, NYIT

For untethered and unattended low-end Wireless Sensor Networks (WSNs), especially those deployed in hostile environments, it is impossible to replace the battery of any sensor node. A considerable research effort has been allocated to address energy saving issues to prolong the lifetime of WSNs. The current tree-based WSNs suffer from uneven consumption of energy and corresponding early exhaustion of batteries of nodes closer to the sink. In this research, an attempt at solving this problem is done by adding sinks located away from the current sink shifting active sink after exhaustion of nodes at closer layers.

In tree-based WSNs, the network is represented by a routing tree whose root is the current sink. When the tree stops, its normal data collection after early exhaustion will be switched to another sink and a new tree will be initialized, hopefully with a different set of nodes at closer layers nodes to the new sink. Hence the new tree can continue to work. This approach can be combined with low duty cycle techniques, which put nodes into sleep according to various scheduling algorithms, to further extend a network’s lifespan. During the network initialization, all sinks exchange location information. A minimum circle is defined as the smallest circle that covers the current sink and all leaf nodes of the tree. The central angle is calculated as the angle from the center of the circle to the current sink and each backup sink. The next sink will be the one with the maximal central angle.

 

“A Non-Zero Effort Attack on Cyber-behavioral Biometric User Verification”

Kiran S Balagani, Ph.D.
Assistant Professor of Computer Science, NYIT

Previous efforts in cyber-behavioral verification considered only zero-effort impostor attacks in their evaluations (i.e., the genuine samples of one user have been used as impostor attempts for another and an impostor as such did not mimic a victim’s cyber-behavior). Taking continuous verification with keystroke dynamics as a case in point, we show that a non-zero effort forgery attack, created from a victim’s stolen keystroke timing information, has an alarmingly high success rate. Our attack is executed in three steps: 1) steal a victim’s keystroke timing information using a rogue keylogger; 2) create a forged verification attempt using the victim’s keystroke timing information; and 3) replay the forged attempt to mock victim’s typing behavior. In our experiments, with as little as 50 to 200 stolen keystrokes (roughly, less than two lines of text typed in an email), we were able to create attacks that had as high as 87.75 percent success rates against verifier configurations that showed less than 10 percent “zero-effort” impostor pass rates. Our results suggest that there is a wide gap between the performances of cyber-behavioral keystroke authentication systems evaluated against a “zero-effort” impostor model and a “non-zero effort” attack model.