Yanyong Zhang, Ph.D. - US grants

This collaborative research proposal is focused on the creation of a large-scale wireless network testbed which will facilitate a broad range of experimental research on next-generation protocols and application concepts. It is recognized that powerful technology and market trends towards portable computing and communication imply an increasingly important role for wireless access in the next-generation Internet. At the same time, new sensor and pervasive computing applications are expected to drive large-scale deployments of embedded computing devices interconnected via new types of short-range wireless networks. The speed of technology innovation in the wireless networking field can be significantly increased with the development of a flexible, open-access wireless network testbed that can be shared by experimental researchers across the networking community.

The proposed ORBIT (Open Access Research Testbed for Next-Generation Wireless Networks) system is a two-tier laboratory emulator/field trial network testbed designed to achieve reproducibility of experimentation, while also supporting evaluation of protocols and applications in real-world settings. In particular, the laboratory-based wireless network emulator will be constructed using a novel approach involving a large two-dimensional grid of static and mobile 802.11x radio nodes which can be dynamically interconnected into specified topologies with reproducible wireless channel models. All radio devices in the system provide open API's that permit end-users to download radio link, MAC and network layer protocols to construct a specific networking scenario. Once the basic protocol or application concepts have been validated on the lab emulator platform, users can migrate their experiments to the field test network which provides a configurable mix of both high-speed cellular (3G) and 802.11x wireless access in a real-world setting. Extensive measurement tools will be provided to support research evaluation, including both network traffic and radio link/spectrum usage aspects.

In addition to the development of the ORBIT wireless testbed infrastructure, this project includes a comprehensive set of "experimental work packages" intended to generate design requirements and serve as end-user application drivers for the system being developed. Specific research topics to be covered during the course of this project are:

1. Ad hoc networking in 802.11x WLAN scenarios [Raychaudhuri, Seskar; Rutgers & Acharya; IBM]
2. Message-based multimedia delivery [Schulzrinne, Columbia; Yates, Rutgers]
3. XML-based content multicasting for mobile information services [Ott, Raychaudhuri; Rutgers]
4. Location-based mobile network services [Schulzrinne; Columbia]
5. Pervasive computing software models for sensor networks [Parashar, Zhang; Rutgers]
6. Security protocols for next-generation wireless networks [Kobayashi; Princeton & Trappe; Rutgers]
7. Intelligent network middleware (INM) for mobile services [Paul; Lucent Bell Labs]
8. Peer-to-peer infrastructure for VoIP and IM [Acharya, Saha; IBM Research]
9. Power/bandwidth efficient media delivery to portable platforms [Ramaswamy, Wang; Thomson R&D]

The project will be conducted as a collaborative effort between several university research groups in the NY/NJ region: Rutgers, Columbia, and Princeton, along with industrial partners Lucent Bell Labs, IBM Research and Thomson. The wireless network testbed will be developed and operated by Rutgers WINLAB, using facilities located at the Rutgers New Brunswick campus and at partner sites in the area. The testbed will be available for remote or on-site access by other research groups nationally, subject to NSF guidelines for use. Additional partners will be sought during the course of the program both for testbed infrastructure development and for research collaboration.

The scientific/technical merits of the proposed project are: advancing the state-of-the-art in design and
implementation of flexible and scalable wireless network testbeds, and experimental investigation of novel
architectures, protocols and service concepts for next-generation wireless networks. Broader impacts are in
acceleration of the R & D cycle for wireless networking by providing the research community with a shared-use experimental platform, and in fostering increased use of experimental methods in both research and teaching.

Abstract:

Providing privacy for sensor networks is an important problem that is complicated by the fact that it is easy for adversaries to observe communications between sensor nodes. A first line of defense for protecting sensor communications is cryptography. However, these methods cannot address the complete spectrum of privacy issues in sensor systems. Specifically, security solutions are inadequate for protecting the privacy of contextual information surrounding a sensor application, such as the source's location, or the time at which a measurement was made, or even the size of sensor data packets.

This project investigates the development of a framework for providing three critical types of contextual privacy to sensor communications: source location privacy, temporal privacy, and traffic privacy. The project takes the viewpoint that the existing network stack can be modified to protect privacy while maintaining desirable levels of resource-efficiency. This investigation will enhance the privacy levels achieved through the development of new routing protocols involving the use of directed random walk techniques to obfuscate the data source, the modification of the structure of sensor messages to prevent traffic analysis attacks, the introduction of delay in the delivery of messages to reduce temporal correlation attacks, and the introduction of modifications to physical layer communications and the sensor topology to prevent the localization of a communication source. Through dissemination of the research results in both archival publications and new curricula, this project will advance the development of sensor applications by addressing critical privacy issues before sensor systems become a communal asset.

Large scale parallel systems are critical to our computational infrastructure to take on the challenges imposed by applications whose scale and demands exceed the capabilities of machines available in the market today. Pushing the limits of hardware and software technologies to extract the maximum performance, in turn, exacerbates other problems. Notable amongst these problems is the susceptibility to failures, which arises as a consequence of growing hardware transient errors, hardware device failures, software complexity, and the complex hardware/software inter-dependencies between the nodes of a parallel system. These failures can have substantial consequences on system performance, in addition to impacting the costs of maintenance/operation, thereby putting at risk the very motivation behind deploying these large scale systems.

This research is expected to make three broad contributions towards developing a runtime infrastructure, called PROGNOSIS, for failure data collection and online analysis. The first set of contributions will be on collecting and analyzing system events and failure data from an actual BlueGene/L system over an extended period of time. In addition to presenting the raw system events, the research will be developing filtering techniques to remove unimportant information and identifying stationary intervals, together with defining the attributes for logging and their frequency. The second set of contributions will be models for online analysis and prediction of evolving failure data by exploiting correlations between system events over time, across the nodes, and with respect to external factors such as imposed workload and operating temperature. The third set of contributions will be on demonstrating the uses of PROGNOSIS. Tools such as PROGNOSIS can help substantially in the development of self-healing systems, which has been noted to be an important goal in the emerging area of Autonomic Computing by several computer vendors.

Large scale parallel systems are critical to take on the challenges
imposed by highly demanding applications of critical importance. Pushing the
limits of hardware and software technologies to extract the maximum
performance can increase their susceptibility to failures. This arises as a
consequence of growing hardware transient errors, hardware device failures,
and software complexity. These failures can have substantial consequences
on system performance, and add to the costs of maintenance/operation,
thereby putting at risk the very motivation behind deploying these large
scale systems. Rather than treat failures as an exception and take
reactive remedies, this project intends to anticipate their occurrence
and take pro-active runtime measures to hide their impact.

This research is expected to make three broad contributions towards
developing a runtime fault-tolerance infrastructure.
The first set of contributions is on collecting and analyzing
system events from an actual BlueGene/L system over an
extended period of time. The second set of contributions are models for
online analysis and prediction of evolving failure data.
The third set of contributions are on failure-aware parallel job
scheduling and checkpointing. On the educational front, in addition to
enhancing graduate curriculum and research, this project intends to involve
undergraduate students and women. The tools developed in this project and the
related results will be made available in public domain and published in
leading journals/conferences. In addition, the PIs will also push these
tools to be incorporated on actual systems, to enhance their fault-tolerance
abilities.

While sensor applications promise profound scientific and economic impacts on our society, their success is nonetheless determined by whether the sensor networks can provide a steady stream of correct data. This need for continuous data provisioning over a significant time period, however, imposes great challenges to the underlying system. Specifically, wireless sensor networks are prone to a large array of hazards, such as frequent node failures, congestion, and sensing errors. These challenges are further complicated by the fact that sensor systems are usually seriously energy constrained.

This project has three main components: DADA, TARA, and MARA. DADA, a 2-Dimensional Adaptive Scheduling Framework, rapidly repairs network coverage and connectivity by cleverly waking up the redundant nodes that are needed for repairing holes created by the node failure. TARA, a Topology-Aware Resource Adaptation Framework, strives to increase the resource provisioning by bringing more sensor nodes online to accommodate those packets that contain valuable data, potentially regarding the source of the congestion. MARA, a Measurement Assurance and Robust Aggregation Framework, provides a set of data classification and cleansing tools that validate the data before they flow into the network. This project attempts to deliver the guarantee that sensor systems must gracefully recover themselves in the presence of network exceptions, in order to satisfy the application needs as well as lower the network management costs. In addition to technical papers that report the research results, this project will also produce a suite of software tools that will be made available to the community.

Most commercial wireless devices do not make lower-layer properties (e.g., raw waveform-level samples from an analog-to-digital converter) accessible to users. Recently, however, the research community has directed its attention towards the development of cognitive radios that will expose the lower-layers of the protocol stack to researchers and developers. Although the promise of such a flexible platform is great, there are also some serious potential security drawbacks. It is easily conceivable that cognitive radios could become an ideal platform for abuse since the lowest layers of the protocol stack will be accessible to programmers in an open-source manner. The proposed project addresses these concerns by focusing on two important building blocks needed in constructing a holistic solution to ensuring the trustworthy operation of software radios: first, the investigating team plans to develop tools to quantify the degree to which spectrum etiquette policies are abused in a network of cognitive radios and, second, the team plans to investigate methods for identifying such spectrum abuse, which is necessary in order to drive anomaly detection and response mechanisms. Overall, the broader impact of the effort is centered around the fact that cognitive radios represent an emerging technology that requires security mechanisms to be developed before these highly-programmable radios reach the public market.

Regardless of whether sensors are used just for observations or as the basis for actual responses, the trustworthiness of sensor data becomes critical when important decisions are based upon these data. Unlike traditional networked systems, sensor networks also take measurements of physical phenomena. Traditional information assurance, which focuses on the integrity of data, does not cover all of the sources of errors that might arise in sensing data. In fact, sensor data can be corrupted at the environmental level, whether through a natural loss of calibration or through a deliberate perturbation of the measurement environment by an adversary. These sources of errors, which affect the process of measurement (PoM), are unique to sensor networks and cannot be addressed through the usual network-centric methods. Hence, to complement traditional information assurance services, defense mechanisms are needed to protect the sensor network from PoM errors. Only when a trust wrapper surrounds sensor measurements should applications make decisions or take actions with important implications.

The proposed research consists of two aspects: corruption research and assurance research. For the former, the team will conduct a thorough threat analysis by cataloging the set of PoM errors that may be introduced for a variety of different sensors. For the latter, the team will augment the existing programming stack for sensor networks by developing a suite of measurement assurance tools.


Developing methods that give sensor networks a natural immunity to PoM errors will allow sensor applications to have a greater chance for success and wider spread use.

TC:Large: Collaborative Research: AUSTIN?An Initiative to Assure Software Radios have Trusted Interactions (CNS-0910557)

Software and cognitive radios will greatly improve the capabilities of wireless devices to adapt their protocols and improve communication. Unfortunately, the benefits that such technology will bring are coupled with the ability to easily reprogram the protocol stack. Thus it is possible to bypass protections that have generally been locked within firmware. If security mechanisms are not developed to prevent the abuse of software radios, adversaries may exploit these programmable radios at the expense of the greater good.
Regulating software radios requires a holistic approach, as addressing threats separately will be ineffective against adversaries that can acquire, and reprogram these devices. The AUSTIN project involves a multidisciplinary team from the Wireless Information Network Laboratory (WINLAB) at Rutgers University, the Wireless@Virginia Tech University group, and the University of Massachusetts. AUSTIN will identify the threats facing software radios, and will address these threats across the various interacting elements related to cognitive radio networks. Specifically, AUSTIN will examine: (1) the theoretical underpinnings related to distributed system regulation for software radios; (2) the development of an architecture that includes trusted components and a security management plane for enhanced regulation; (3) onboard defense mechanisms that involve hardware and software-based security; and (4) a algorithms that conduct policy regulation, anomaly detection/punishment, and secure accounting of resources.
Developing solutions that ensure the trustworthy operation of software radios is critical to supporting the next generation of wireless technology. AUSTIN will provide a holistic system view that will result in a deeper understanding of security for highly-programmable wireless devices.

This proposal is targeted at realizing the vision of a "green'' Internet of Things (IoT) and green communication using transmit-only devices. It is predicted that by 2020, there will be 50 billion embedded devices deployed in our ambient environment, most of which will be reporting data to the cloud through wireless communication. Thus, it is imperative to design green communication technologies in which power consumption is minimized and bandwidth utilization is optimized. Existing communication protocols, however, are not optimized for power consumption and bandwidth utilization because they were designed to facilitate reliable two-way exchange of information between communicating parties -- their requirements are completely different from those of IoT applications. The needs of emerging IoT applications, such as unidirectional communication flow, dense deployment, small packet sizes, are all opportunities for significantly simplifying network design. This project aims to simplify the protocol design by removing receiving functions from the embedded devices, such that they only spend radio resources on sending application data, thus minimizing power consumption and maximizing bandwidth utilization.

This project will study a set of algorithms that can achieve high throughput for wireless networks using transmit-only devices. The biggest challenge for a transmit-only network is the handling of packet collisions as the transmitters do not have any means of knowing whether others are transmitting at the same time. Transmit-only can be thought of as a single-input-multiple-output multiple access (SIMO-MAC) channel, but there is a fundamental difference between transmit-only and previous work in SIMO-MAC from the information theory community - in almost all of the previous studies, while transmitters do not communicate among themselves, they do rely on feedback from receivers to make transmission decisions related to encoding and/or scheduling. Transmit only, on the other hand, assumes once the network is deployed and in operation, each transmitter does not have any feedback from other transmitters or receivers. In this case, to reduce packet collisions, and to further ensure packet collisions do not lead to packet loss, this project proposes a set of strategies to pro-actively control the network topology as well as transmission schedules before network deployment. First, an optimal receiver placement strategy and a network dynamics based transmitter placement strategy are proposed to minimize the packet loss during a collision by exploiting the fact the stronger signal can be decoded at the receiver. Second, a transmission scheduling algorithm is proposed to overlap transmissions that can be decoded together (by different receivers) to minimize the collisions. The proposed scheduling algorithm also takes into consideration transmitter mobility to minimize their negative impact on the network throughput.

The Next-Phase MobilityFirst (MF) project aims to have a major impact on the architecture of the future Internet by re-architecting it to address the needs of emerging mobile platforms and applications. Adoption of technologies arising from this project may be expected to provide improved efficiency, security and robustness that would benefit both network operators and end-users of the Internet. This project, originally funded as a collaborative research effort under the NSF Future Internet Architecture (FIA) program (2010-13) in which the MF architecture was designed over the past 3 years, is centered on a new name-based service layer which serves as the narrow-waist of the protocol; this name-based services layer makes it possible to build advanced mobility-centric services in a flexible manner while also improving security and privacy properties. The architecture incorporates novel storage-aware routing techniques which provide significant improvements in mobile network capacity and functionality. The next phase of the MobilityFirst project is aimed at making the transition from early-stage architecture and prototyping to advanced real-world services and trial network deployments. The research and experimental trials agenda is aimed at validating and refining the core name service, routing, security and management components of the MF architecture, while also responding to emerging trends in network technology and services such as the cellular mobile data explosion, the growth of content, the emergence of cloud computing, and software-defined network (SDN) technology.

Intellectual Merit: This project includes several research thrusts aimed at transitioning the MobilityFirst architecture to advanced services and field deployable technology. These include: (1) advanced name-based network services and development of enhanced global name service (GNS) technology; (2) network security and privacy designs and enhancements; (3) design of advanced content services; (4) application of MobilityFirst protocols to next-generation mobile cloud computing; (5) design of advanced context-aware services; (6) technical and economic study of cellular-Internet convergence; (7) software-defined network (SDN) ready protocol design; and (8) technology platforms, router implementation and deployment strategies. These research thrusts will be informed by three distinct real-world network environment trials: a "mobile data services" trial with a wireless ISP (5Nines) in Madison; WI; a "content production and delivery network" trial involving several public broadcasting stations in Pennsylvania connected by a greenfield optical network called PennREN; and a "context-aware public service" weather emergency notification system (CASA) with end-users in the Dallas/Fort Worth area. These network environment trials are the centerpiece of the proposed project, and are expected to provide a firm basis for validation of the MobilityFirst protocol stack and its usefulness for developing advanced mobile, content, context and cloud applications, while also advancing the technology to the field-deployment stage. Expected outcomes from the project include research results on security, privacy, content/context/cloud services and SDN; MobilityFirst protocol stack software revisions; router technology implementations; multiple real-world trial deployments of the technology; and experimentally supported evaluations of the architecture. This project is a collaborative effort involving Rutgers, UMass, MIT, Duke, U Michigan, U Wisconsin, and U Nebraska with the participation of several industrial research and network environment trial partners.

Broader Impacts: The MobilityFirst project will have impact as a new approach to a future Internet that by design addresses mobility and mobile platforms, and as an enabler of new mobile Internet applications of social value such as context-aware emergency notification services. The release of open source protocol software may be expected to help to stimulate further experimental research on future Internet architectures across the networking community. The project also contributes to education and training in the key areas of Internet and mobile network technology.

This project leverages prior work on virtual mobile network technology at NICT and the MobilityFirst future Internet architecture at WINLAB, Rutgers University, to develop a comprehensive services and networking solution for high-performance cyber physical systems that scales to the "trillion object" level targeted by the JUNO program. The aim is to develop a virtual mobile cloud network (vMCN) which provides seamless and low latency services to real-time mobile users and applications.

Major research themes addressed by this project include the design of a new virtual networking framework using NICT's BYON mobile cloud technology integrated with MobilityFirst's globally unique identifier (GUID) based protocol stack; design of virtual network services for efficient support of cloud services; exploiting locality to speed up global name resolution; and dynamic migration of cloud services across networks. The project will adopt a top-down application driven methodology to validate and benchmark the performance of the proposed virtual mobile cloud network for a specific advanced CPS application. A proof-of-concept prototype of the proposed vMCN system will be developed using JGN-X and GENI testbeds in Japan and US respectively.

Technologies resulting from this project are expected to enable a wide range of commercial and government applications involving real-time cloud services for mobile devices. The project will also provide guidance for the development of future virtual network technologies of increasing interest to the networking and computer industries. The proposed collaboration will also help to strengthen research ties between US and Japan specifically in the field of future Internet architecture.

The development of radio technologies that support efficient and reliable spectrum sharing is an enabler for utilizing the spectrum being made available through the National Broadband Plan. Software defined radios represent a promising technology that supports spectrum sharing as evidenced by the large amount of algorithms and protocols that allow for cognitive radio networks (CRNs) to be deployed. Unfortunately, the economic promise of dynamic spectrum access is easily undermined if cognitive radio users act dishonestly or maliciously, thereby subverting protocols that are founded on the cooperation of users. It is therefore important that mechanisms are developed that ensure the trustworthy operation of CRNs in the presence of potentially malicious or malfunctioning wireless nodes. The objective behind the project's research activities is to develop technological solutions that ensure that cognitive radios operate in trustworthy manner in spite of potential security threats. As a result of this research effort, it is possible for radio spectrum to be more reliably utilized, thereby ensuring that the economic opportunities associated with the radio spectrum are fairly utilized by everyone. The educational impact of the work comes from its multi-disciplinary foundation, broadening student views of wireless system design, and guiding the next generation of wireless engineer to include security and reliability in the design process.

Wireless technologies are an enabler for economic growth in the United States, and cognitive radio networks are an emerging form of wireless system that make spectrum access more available to the broader population. Unfortunately, cognitive radio systems are susceptible to threats that undermine the correct operation of their algorithms and protocols, and thus solutions that support the secure operation of cognitive radio networks are needed. This project ensures the trustworthy operation of cognitive radio networks by: 1) developing algorithms that ensure the correct operation of spectrum sensing procedures upon which spectrum access protocols rely; 2) developing traffic monitoring tools that identify improper communication activity by cognitive radio devices; and 3) developing new forms of interference-resistant communications that ensure that cognitive radio communication continues reliably in the face of interference. The research effort is inter-disciplinary, pulling from statistical tools to network traffic analysis to communications theory to support the secure operation of cognitive radio networks. The algorithms and protocols developed in this project are complemented by a systems prototyping and experimentation effort aimed at guaranteeing that the technologies developed are suitable for deployment in real wireless systems.