My Research

I am a Faculty Research Associate in the Dept of Computer Science, University of Maryland College Park.

Through my research, I have explored various aspects of computer networking and distributed systems. My research interests are in protocols, services, and architectures for mobile/wireless networks and ubiquitous computing environments. My work covers different layers of the protocol stack from the physical layer up to the application layer. Specific research projects target location determination systems, sensor networks, protocol modeling and analysis, peer-to-peer systems, network measurements, and security.

List of Publications

The following is a list of the different projects that I have worked on:

The Cyclone Time Technology

With Ashok Agrawala, Professor, Dept of Computer Science University of Maryland, College Park
A project of the MIND Lab

Cyclone Time Technology enables heterogeneous systems that include clocks of various inherent precision, resolution and stability to synchronize. Cyclone offers several advantages over the current master-slave based techniques by avoiding a single point of failure and achieving accuracy that does not depend on the actual local clock drift rates. Further, the use of only local information makes the scheme highly scalable.

The proposed scheme works by first assuming a local clock model at each node that takes the clock offset and drift rate into account. Timestamps are exchanged between neighboring nodes which permit each node to derive a common time base using only its local information. The mathematical basis of the approach comes from linear algebra, in which the principal eigenvector of a stochastic matrix can be calculated by repeated multiplications of matrices.

Cyclone also defines the notion of a common Virtual Clock. Cyclone allows all nodes to calculate the parameters, drift rate and offset, for a virtual clock such that any node can convert its clock readings to the reading of the common virtual clock at any time instant, even through different nodes’ local clock reading may be widely different.

Our initial results show that we can achieve a stable and high degree of clock synchronization. The degree of synchronization achieved is affected by the perturbations in both (i) the transit time and (ii) the clock drift rate, but does not depend on the clock drift rate.


PinPoint: An Asynchronous Time-Based Location Determination System

With Ashok Agrawala, Professor, Dept of Computer Science University of Maryland, College Park
A project of the MIND Lab

PinPoint is a distributed algorithm that enables a set of n nodes to determine the RF propagation delays between every pair of nodes, from which the inter-node distances and hence the spatial topology can be readily determined. PinPoint does not require any calibration of the area of interest and thus is rapidly deployable. Unlike existing time-of-arrival techniques, PinPoint does not require an infrastructure of accurate clocks (e.g., GPS) nor does it incur the o(n^2) message exchanges of “echoing” techniques. PinPoint can work with nodes having inexpensive crystal oscillator clocks, and incurs a constant number of message exchanges per node to determine the location of n nodes. Each node’s clock is assumed to run reliably but asynchronously with respect to the other nodes, i.e., they can run at slightly different rates because of hardware (oscillator) inaccuracies. PinPoint provides a mathematical way to compensate for these clock differences in order to arrive at a very precise timestamp recovery that in turn leads to a precise distance determination. Moreover, each node is able to determine the clock characteristics of other nodes in its neighborhood allowing network synchronization. Evaluation of the prototype in typical indoor and outdoor environments shows that PinPoint gives an average accuracy of four to six feet, in different environments, allowing PinPoint to support accurate rapidly deployable localization scenarios.


Rover II: An Information Dynamics-Based Context Aware Platform

With Ashok Agrawala, Professor, Dept of Computer Science University of Maryland, College Park
A project of the MIND Lab

The Rover II Technology uses the Information Dynamics paradigm developed at the MIND Lab to provide a context-aware integration platform that is platform independent. The Rover infrastructure focuses on two major tasks: to expose as much contextual information as possible and to create services which need to communicate with each other and external sources, such as the Internet. We provide an easy to use application programming interface (API) that allows developers to create Rover-enabled applications with communication and messaging in mind. This project is an extension to the Rover project below.

Horus: A WLAN Location Determination System

With Ashok Agrawala, Professor, Dept of Computer Science University of Maryland, College Park
A project of the MIND Lab

My thesis work defines Horus, which is a location determination system based on the 802.11 networks.

Our research focuses on identifying the noisy characteristics of the wireless channel and developing techniques to overcome them to obtain accurate positioning. It is advantageous to run the location determination algorithm on the client devices to achieve privacy and decentralized implementation. Since these devices are usually energy constrained, it is important to reduce the computation requirements for location determination algorithms. We have developed location-clustering techniques based on the signal strength received from the access points to reduce the computational requirements of the location determination algorithm and allow the system to scale to large areas.

The Horus system has been used in other research projects such as the Rover system and the Location Based Authentication  project. The Horus system has been tested in areas as large as 20,000 square feet and the accuracy is 2.5 feet on the average for different testbeds.

As part of our work, we developed device drivers to query the wireless card and API's to make the Horus system independent of the underlying device driver/card. Our software has been used by other wireless researchers around the world.

Currently, we are looking into techniques to ensure user privacy even if the user is communicating with the system and the signal strength from his device can be recorded by the system. As part of our ongoing work we are experimenting with different clustering techniques, automating the radio-map generation process, developing applications and services over Horus, dynamically changing the system parameters, and finding the optimal placement of access points to achieve maximum accuracy.


Rover: Location-Aware Computing for Wireless Environments

With Ashok Agrawala and A. Udaya Shankar, Professors, Dept of Computer Science University of Maryland, College Park
A project of the MIND Lab

Rover enables location, time and context-aware applications for wireless devices that scale to very large user populations. Users interact with the Rover system through client devices (Rover-clients) that typically are small handheld units with a wireless communication interface. Rover-clients can have great heterogeneity in capabilities in terms of processing, memory and storage, graphics and display and network interfaces.

The Rover server interacts with the clients to provide and manage the different service requests from the Rover-clients. To scale the Rover server operations to a very large client set, we have defined a new Action model which allows fine-grained, real-time scheduling of server operations.

Apart from system design, I have also been involved in project with a team of other students in implementing different aspects of the system.

We have implemented and demonstrated both outdoors and indoors version of Rover. In the outdoor case, we used a GPS unit attached the the clients to provide location service. It had an accuracy of less than 3 meters. For the indoor case, the location was based on an earlier version of the Horus system.


802.11-based Research

With Ashok Agrawala, Professor, Dept of Computer Science University of Maryland, College Park
A project of the MIND Lab

In this project, we explore different aspects of the 802.11 protocol ranging from characterizing the wireless traffic, enhancing the security model of 802.11 based networks, and building new applications that makes use of the widespread of 802.11 networks.

Many studies on measurement and characterization of wireless LANs have been performed recently. Most of these
measurements have been conducted from the wired portion of the network based on wired monitoring or SNMP statistics. In
the wireless traffic characterization project, we argue that traffic measurements from a wireless
vantage point in the network are more appropriate than wired measurements or SNMP statistics, to expose the wireless medium
characteristics and their impact on the traffic patterns. While it is easier to make consistent measurements in the wired part of
a network, such measurements can not observe the significant vagaries present in the wireless medium itself. As a consequence
constructing an accurate measurement system from a wireless vantage point is important but usually quite difficult due to the
noisy wireless channel.

In our work we have explored the various issues in implementing such a system to monitor traffic in an
IEEE 802.11 based wireless network. Our analysis reveals rich information about the PHY/MAC layers of the
IEEE 802.11 protocol such as the typical traffic mix of different frame types, their temporal characteristics, correlation with
the user activities and the error characteristics of the wireless medium. Moreover, we identify anomalies in the operation of
the IEEE 802.11 MAC protocol.

Another technology that we implemented to enhance the 802.11 security model is Koolspan. In Koolspan, user authentication is performed through smartcard-based physical tokens. The Koolspan Client Key secures wireless traffic by connecting to Koolspan SecurEdge Unilock installed behind the access point. A major goal for Koolspan is to be transparent to the standard 802.11 access points and clients, allowing easy integration with the current installed networks.


Energy Efficient Wireless (Sensor) Networks

With Mohammed Younis, Professor, Dept of Computer Science University of Maryland, Baltimore County   
and  Dr. Khaled Arisha, Senior Research Scientist, Honeywell Advanced Systems Group, Columbia, MD
Work as intern at Honeywell Advanced Systems Group, Columbia, MD

Battery power is a scarce resource in wireless devices in general and in sensor nodes in particular. Therefore, this power needs to be conserved. In this project, we have defined energy efficient link layer and network layer protocols for sensor networks. Existing protocols for minimum energy routing chooses end-to-end paths depending on the battery capacity and transmission costs of the nodes on the path. However, they ignore other performance metrics such as end-to-end delay and throughput which are crucial to some applications. 

Our work focuses on balancing different performance metrics. By changing system parameters, different systems can achieve different performance objectives depending on the mission assigned to the sensor network.  

We designed a routing protocol that selectively turns sensor nodes on or off based on the individual nodes' energy state and on the the global system performance requirements. The routing decisions are changes based on events in the system such as target movement or a drop of the energy of a node below a threshold. We showed that our routing protocol achieve significant improvement is performance over the current energy efficient protocols without sacrificing energy efficiency.

We also experimented with different energy-efficient TDMA MAC layer protocols and designed techniques for assigning slots to obtain better throughput and less changes in the state of the wireless card circuitry.  

Current extensions to our work include designing clustering algorithms for sensor networks and implementing the mechanisms in a real world multi-hop wireless network. .


Instance-Based Network

with Liviu Iftode, Associate Professor, Dept of Computer Science, Rutgers University.

In this work we consider the design principles of the Instance-Based Network (IBN), an extended version of a generic Content-Based Network (CBN). IBN acts as an overlay communication platform over which end-point entities, called contents, communicate independently from their physical locations while providing the flexibility of having different instances of the same content. The semantics of different instances are assigned by the application using the IBN. Routing in the IBN is instance-based; the IBN can route a message to a specific content instance or to the closest instance, if no exact match is found for the destination content instance. Moreover, the IBN replicates the stored contents in order to provide fault tolerance. 

Possible applications for the IBN applications include:

We have developed an implementation prototype based on Pastry as the underlying peer-to-peer lookup service.

Currently, we are working on evaluating the performance of the IBN, implementing applications over it, and experimenting with different underlying infrastructures.


Autonomous Transport Protocol

with Liviu Iftode, Associate Professor, Dept of Computer Science, Rutgers University.

The basic service provided by the Autonomous Transport Protocol (ATP) is a reliable transport connection between two endpoints, identified by content identifiers, independent of their physical location. Autonomy allows dynamic endpoints relocation on different end hosts without disrupting the transport connection between them. ATP depends on the existence of an underlying Instance Based Network (IBN) to achieve its goals. ATP layers at the intermediate nodes can actively participate in the connection. Data is transferred by a combination of active and passive operations, where the ATP layer of a node can decide whether to actively push the data to the destination or to passively wait for the destination endpoint to pull the data. The decision to whether to use the active or passive modes can be taken by a local policy on the node running the ATP protocol.

Current research directions include designing and evaluating different policies for the pull/push decision, designing and implementing applications over ATP, and enhancing the protocols to increase system security.


Location-Based Authentication

with Bill Arbaugh, Assistant Professor, Dept of Computer Science University of Maryland, College Park.

The idea behind location-based authentication is how to authenticate a user based on his position. Once the user position has been authenticated, the user is authorized to access resources based on his position.

We solve the location determination and authentication problem using the Horus system. Results from actual wireless experiments show the feasibility of this scheme. 


Analysis of Network Protocols

with Raymond Miller, Professor, Dept of Computer Science University of Maryland, College Park

For this branch of my networking research, we have been working on formally modeling different networking protocols and analyzing their properties.

We modeled the IEEE 802.11 protocol using the systems of communicating machine approach. We also analyzed the model and confirmed that it is free from deadlocks, unspecified receptions, non-executable transitions. Moreover, we showed that the model have some desirable liveness properties.

We are currently working on the passive testing problem where a system can be tested passively in order to detect, identify and locate faults.


Last updated on Tuesday, 24. October 2006
-- Moustafa