Jana: Ensuring Secure, Private and Flexible Data Access

Anand Sarwate received a grant through the DARPA Brandeis Program. The program was written up in the NY Time Bits blog:


The main program is here: http://www.darpa.mil/program/brandeis

"The Brandeis program seeks to develop the technical means to protect the private and proprietary information of individuals and enterprises. The vision of the Brandeis program is to break the tension between: (a) maintaining privacy and (b) being able to tap into the huge value of data. Rather than having to balance between them, Brandeis aims to build a third option – enabling safe and predictable sharing of data in which privacy is preserved.”

Anand's project is a subaward from Galois, Inc. to DIMACS with Rebecca Wright as PI and David Cash (CS) and Anand Sarwate as co-PIs. The project is to develop a private “data as a service” platform which merges secure multiparty computation and differential privacy to allow flexible secure and private sharing of sensitive data. They will be partnering with other teams funded through the DARPA program to integrate our technology and algorithms into an application system that is currently being negotiated. The Rutgers team will develop algorithms for cryptography and differential privacy and find ways to integrate these two frameworks.

Jana: Ensuring Secure, Private and Flexible Data Access September 15, 2015 — March 14, 2020

Models and Algorithms for Human Activity Recognition

Waheed Bajwa received a one year, $110K subaward from General Dynamics as part of the ARL Robotics CTA Program for the following project: "Models and Algorithms for Human Activity Recognition." The abstract is as follows:

"This research deals with the challenges of human activity recognition using either intensity or both intensity and depth data. It addresses the problems of unsupervised learning of human action attributes from video data as well as provides means for semantic understanding of these action attributes."

Collaborative Situation Aware PNT (CSAP) Solution

Wade Trappe, in collaboration with Mayflower Inc, has received a Phase-1 STTR from the US Air Force for a project titled, "Collaborative Situation Aware PNT (CSAP) Solution". This contract is a 6-month award, with Rutgers University receiving $45K.

The objective behind the proposed effort is to develop and demonstrate cognitive radio (CR) technology to mitigate limited licensed spectrum by accessing efficiently unlicensed spectrum bands in support of military space-based positioning and navigation technology (PNT) needs. In particular, one objective is to overcome signal blockage conditions due to intentional and unintentional impairments faced by PNT signals. It is envisioned that the next generation satellites will be able to support PNT signal transmissions on non-traditional frequency bands and allow cognitive radio based GPS receivers to acquire PNT signals in alternative frequency bands (i.e., S-band and C-band instead of L-band). The CSAP solution represents an enabling technology that allows tactical units to maneuver through the GPS RF challenged environment without any impact on their PNT capability.

Differentially Private Anomaly Detection

Anand Sarwate received an award from the Department of Homeland Security (DHS) as as supplement to the CCICADA center with Rebecca Wright and Anand as PIs. The main award is

2009-ST-061-CCI002-07: Center of Excellence for Command, Control, and Interoperability

The project is:

Differentially Private Anomaly Detection July 1, 2015 — June 30, 2016

The goal is to develop algorithms for screening and anomaly detection in private data using a combination of techniques from group testing, active learning, and sequential hypothesis testing. The objective is to evaluate how well and in what contexts differentially private algorithms can reliably detect anomalies while preserving the privacy non-anomalous data/individuals.

Cyber Resilient Energy Delivery Consortium

Saman Zonouz received a grant from the Department of Energy for the project Cyber Resilient Energy Delivery Consortium. This is a joint grant with MIT, OSU, UIUC, TSU, WSU, Dartmouth, ANL, PNNL, ASU. The total budget is $22M for 5 years (Rutgers share: $500K)

The abstract is given below.

The threat to the power and oil & gas industries is real, and is growing. To get past the "react and patch" mode of responding to cyber intrusions, we need to find a way to transform cutting-edge research in EDS cyber security and resilience into technology in which industry chooses to invest that protects, detects, manages, and helps in recovery from cyber attacks. The overarching project objective is to identify and perform research and development to measurably increase the cyber-resiliency of energy delivery systems, particularly those for power, oil, and gas. We will accomplish those objectives, delivering both foundational and translational results that will enable the nation to achieve the goal of resilient energy delivery systems.

EARS: Machine Learning and Social Protocols for Enhancing Spectrum Access for Wireless Communications

Janne Lindqvist received an NSF award for $300,000.00 from the NSF Enhancing Access to the Radio Spectrum (EARS) program. Richard Howard of WINLAB is senior personnel on the grant.

Janne would like to acknowledge the help of Shridatt “James” Sugrim for helping with parts of the proposal.

The abstract is given below.

Title: EARS: Machine Learning and Social Protocols for Enhancing Spectrum Access for Wireless Communications

PI: Janne Lindqvist. Senior Personnel: Richard Howard

This project designs, implements and studies novel protocols for enhancing access to radio spectrum by taking advantage of the fundamentals of human behavior.It produces technologies, theories, and guidelines for protocols that are expected to significantly improve the efficiency of spectrum access. This may lead to substantial societal impact by allowing more work to be done with the same resources. Success will also benefit the environment by limiting the infrastructure needed for the required data traffic; both energy and infrastructure investment can be minimized. Further, the project creates a model of human behavior in a specific technical context that can serve as a basis for similar projects in other technology areas involving resource optimization. This interdisciplinary project applies and develops expertise from areas of social computing, machine learning, wireless technology, security engineering, physical analogs, mobile systems, and user-centered design. The project implements and deploys a system that enables efficient bandwidth sharing with machine learning and social protocols that goes beyond what is possible with technology alone. Social protocols are cooperative yet discretionary methods that allow to users distribute access more fairly using inherently natural decision-making processes as opposed to externally imposed ones. The project consists of three major activities: (1) study the main approaches to social protocols, including persuasive computing, clinical behavior change theories, and micro-tasks; (2) develop a system that observes users trying to access the network and facilitates control rules that allow maximum value to users in the most transparent way; and (3) deploy these social protocols and the system in live networks to evaluate the approaches in real-life settings.

EAGER: Renewables: Collaborative Research: Foundations of Prosumer-Centric Grid Energy Management

Narayan Mandayam and Arnold Glass (Psychology) received an NSF EAGER Award to explore new directions in prosumer-centric smart grid management. This is a collaborative proposal between Rutgers, Princeton and Virginia tech. The award was 300K and our share is 100K. Please see details below.

EAGER: Renewables: Collaborative Research: Foundations of Prosumer-Centric Grid Energy Management

PI: Narayan Mandayam, Co-PI: Arnold Glass (Psychology)

The realization of the vision of a smart power grid in which a significant portion of energy stems from renewable sources and other prosumer-owned devices (a prosumer is a consumer who can take the dual role of seller and a buyer of electricity) is contingent on large-scale, active prosumer participation in energy management. However, just because such participation can yield significant technological and societal benefits, it cannot be assumed that prosumers will actually become fully involved in the smart grid. Empirical data shows that, despite its exciting prospects, the widespread adoption of the smart grid has been hindered by modest user participation. Motivated by emerging grid scenarios, this project employs the mathematical framework of prospect theory, a seminal contribution to behavioral economics that won the Nobel Prize, to study grid energy management, as well as understanding and overcoming barriers to user participation. Prospect theory provides a methodology for understanding people’s economic choices based on their actual behavior and their assessment of potential gains and losses versus the assessed levels of risk. The interdisciplinary research team of engineers and a cognitive psychologist will provide a fundamentally new understanding of the role of prosumers in the smart grid and study: 1) new fundamental results on the impact of prosumer behavior on energy management, 2) prosumer-centric, sustainable energy management schemes and associated pricing mechanisms that optimize grid operation by tightly integrating the effect of prosumer behavior and subjective utility perceptions via novel prospect theory notions, 3) grid-aware energy management that accounts for grid dynamics and uncertainty due to factors such as renewables, and 4) real-life cognitive psychology experiments that will yield new, realistic behavioral models for smart grid prosumers. The results of the project will advance multiple disciplines including power systems, game theory, economics, and cognitive psychology. The unique marriage of smart grid design and cognitive psychology will offer an innovative educational opportunity to involve students from both engineering and psychology via participation in behavioral experiments.

Dynamic Context-Aware Data Protection Through Virtual Micro Security-Perimeters in Smartphones and Wearable Devices

Saman Zonouz and Dario Pompili received a 2.5 year, $576K grant from the Department of Homeland Security for their project "Dynamic Context-Aware Data Protection Through Virtual Micro Security-Perimeters in Smartphones and Wearable Devices".

The abstract follows.

Abstract: Smartphones are quickly becoming the dominant platform over which cloud services and content are consumed. However, security on current mobile platforms leaves much to be desired. Multi-user support on mobile operating systems (OSes) is just starting to be offered, and no mobile OSes offer first class facilities for separating sensitive data (e.g., a user's work files) from mixing with non-sensitive data or leaking to untrusted endpoints. An example is how multiple e-mail accounts are handled in today's mobile email clients. It is common for emails from one account to be accidentally forwarded over another account where they may be stored on untrusted servers, or leak to unintended recipients. The outcomes will provide fine-grained policy based data-protection as a first class primitive in the mobile OS itself so that the user doesn't have to maintain completely different environments. The proposed solution will allow individual data and applications to be contained in micro-security perimeters or capsules. These capsules can be securely installed on a phone, and are subject to a data security policy defined by the capsule owner. The OS will then track the flow of data on a per-capsule basis as it is used by applications on the phone, and enforce the security policies associated with it. The PIs will be working with AT&T Labs through their established collaborations and connections to realize the proposed solution as a real-world open-source software package for public use.

Runtime Optimal Semantic Gap-Filling System Security Monitoring via Offline Automated Executable Profile Generation and Dynamic Sensor Deployment

Saman Zonouz received a 3 year, $572K grant from ONR as a sole PI for the project "Runtime Optimal Semantic Gap-Filling System Security Monitoring via Offline Automated Executable Profile Generation and Dynamic Sensor Deployment".

The abstract follows.

Abstract: Secure operation of next generation critical cyber infrastructures requires not only intrusion preventative security hardening solutions, e.g., stack canaries, to prevent attackers from intruding the critical computer systems but also online optimal system security monitoring to provide semantic security status reports about low-level malicious activities within the systems. The objective of this research is to develop the foundations and working practical algorithms to provide adaptive and semantic gap-filling system security monitoring capabilities in complex and critical computing assets. The proposed automated self-aware and scalable cyber security solution will enable computing systems to survive cyber attack scenarios with minimum manual intervention and will provide the security admins with meaningful and concise reports about security incidents in real-time.

Active data screening for efficient feature learning

Waheed Bajwa and Anand Sarwate received a one year $160,000 grant from NSF CIF for the project entitled "Active data screening for efficient feature learning".

The abstract is given below.

Active data screening for efficient feature learning

Advances in sensing and data acquisition technologies make it easy to generate vast quantities of data that must be stored, communicated, processed, and understood. This data may be of variable quality and the nature of the data may vary over time -- this variability can cause difficulties for existing approaches to efficiently represent the data. Current methods use economical representations of the data in terms of a smaller number of properties, or features, of the raw data. Standard feature representations such as Fourier and wavelet representations may not be efficient at representing the data from these new acquisition technologies. One paradigm to overcome this mismatch is the data-driven approach, in which an algorithm processes the data to learn a novel and efficient feature representation for the given data. While these are more useful, such approaches may not scale well to massive data sets.

This work designs new methods for data-driven feature learning that are scalable and robust to noisy, time-varying data.

It proposes an "active screening" approach to learning new features; the data processing algorithm uses a low-complexity criterion to screen for useful and informative points for feature learning. Advantages of active screening include a reduction in the computational and storage overhead as well as the ability to reject outliers or other spurious and misleading data. The investigators develop active screening methods for consistent estimation under generative models for the data, analyze the tradeoff between representation and classification in active screening for discriminative dictionary learning, and extend the active screening analysis to distributed settings for distributed dictionary learning. They investigate the promise of these methods on two large-scale electrocardiography (ECG) datasets of 170+ patients. This work combines ideas from statistics (feature screening) and machine learning (active learning and selective sampling) to design efficient representations of complex signals from massive data sets and may inform the design of new data acquisition technologies by incorporating screening ideas into the technologies themselves.sarwate-activescreening.jpg

            Illustration of the benefits of active screening: selecting useful samples leads to better signal representations.


Subscribe to Rutgers University, Electrical & Computer Engineering RSS