Projects

Selected Projects (PhD level)

(1) e-ESAS: A Mobile based Remote Monitoring System for Breast Cancer Patients in Rural Bangladesh
(Funded by International Breast Cancer Research Foundation)

Outline: 90 percent of the estimated 30,000 women diagnosed with breast cancer in Bangladesh die from it. Breast cancer (BC) patients need traditional treatment as well as long term monitoring through an adaptive feedback-oriented treatment mechanism. Based on our 31-week long field study, we have designed e-ESAS – the first mobile-based remote symptom monitoring system (RSMS) developed for rural BC patients where patients are the prime users rather than just the source of data collection at some point of time. e-ESAS demonstrates the potential to positively impact the cancer care by (1) helping the doctors with graphical charts of long symptom history, (2) facilitating timely interventions through alert generation and (3) improving three way communications (doctor-patient-attendant) for a better decision making process and thereby improving the quality of life of BC patients.

Time: Aug 2010-Aug 2012

Current status: e-ESAS is currently being used by 10 cancer patients in Bangladesh.

 

(2) IRENE: Context Aware Mood Sharing for Social Network

Outline: Social networking sites like Facebook, twitter, and myspace are becoming overwhelming powerful media in today’s world. Facebook has 500 million active users and twitter has 190 million visitors per month and increasing each second. On the other hand number of smart phone users has crossed 45 millions. Now we are focusing on building an application that will connect these two revolutionary spheres of modern science that has huge potential in different sectors. IRENE a facial expression based mood detection model has been developed by capturing images while the users use webcam supported laptops or mobile phones. This image will be analyzed to classify one of several moods. This mood information will be shared in the user profile of Facebook according to privacy settings of the user. Several activities and events will also be generated based on the identified mood.

Time: Aug 2010-Dec 2012

Current status: The first prototype of IRENE has been developed.

 

Selected Projects (MS Level)

(1) Wellness Assistant: A Virtual Wellness Assistant

Outline: The number of people over age 65 will almost double by 2030 and as they age, they generally prefer to stay at home to go to nursing home [http://www.mobiquitous.org/2005/challenges.html]. Again people often need to travel to see doctors for routine checkups (in case of diabetes, obesity etc.).It is time to break through the physical boundaries of hospitals, and bring the hospital information to the homes of the people than bringing them to the hospital.With WA people can easily get the advantage of health care without traveling to doctors for routine monitoring. The goal of the project is to develop a set of middleware solutions (i.e., Application Programmable Interface (API)) and a WA application using the developed APIs so that others can develop applications on mobile devices (PDAs, smart phones). In this project, WA and middleware solutions will be developed at MarquetteUniversity and will be validated at Medical College of Wisconsin (MCW).

Time: May 2005-May 2007

Current status: The first prototype of Wellness Assistant (PDA portion) has been developed. A survey on the application has been done and the results are published.

 

(2) Healthcare Aide: (Towards a Virtual Assistant for Doctors)

Outline: The advancement of available, portable, low cost handheld devices (PDAs, cell phones, etc.), Bluetooth, and WiFi has resulted in the users’ mobility at unprecedented levels. As these devices can communicate with one another, the combined capabilities can be leveraged to form a useful new set of tools. ‘Healthcare Aide’ has been designed to provide not only an easier and more efficient mode of communication among healthcare professionals, but also a smooth pathway for real-time decision making. Special attention has been given on the ‘duty handover/sign out’ mechanism. Our pervasive middleware MARKS (Middleware Adaptability for Resource Discovery, Knowledge Usability and Self-healing) provides the underlying support for ‘Healthcare Aide’ in a completely transparent manner.

Time: May 2005-Dec 2005

Current status: The first prototype of Healthcare Aide has been developed. It also has been presented in the Forward Thinking Poster Session, Marquette University. A survey on the application has been done and the results are published.

 

(3) SHANTI : Secure Resource Discovery in Pervasive Health Care

Outline: Pervasive healthcare computing aims at seamlessly integrating into and facilitates existing healthcare delivery models with handheld devices and wireless network. Thus it allows the healthcare professionals and patients to overcome constraints of place, time, and personnel availability. To maximize the utility of pervasive healthcare applications, we need an efficient service discovery mechanism that incorporates several necessary critical and complex security, privacy, trust and risk issues. In this project, we present different challenges of secure service discovery. We are also working on implementing a behavioral model which will be used to indicate malevolent attitude of a node. Output of the behavior model will play an important role in calculating dynamic trust. This model will be fitted in our already developed trust model. Dynamic service integration feature will be used to dynamically replace a failed resource with an alternative one.

Time: Jan 2006-May 2007

Current status: We have implemented this project on a test bed of wirelessly connected Dell Axim 50v PDAs. We also simulated the proposed model using OMNeT++.

 

(4) SMARKS (Secure Middleware Adaptability for Resource Discovery, Knowledge Usability, and Self-healing)

Outline: There are several middleware dealing with several aspects of security. But we are still waiting for one which will inherently provide all aspects of security for all the middleware services. SMARKS features include validating devices, discovering resources, modeling trust, handling malicious recommendations, and avoiding privacy violation. At each of these steps we have ensured security. All these features will be available as middleware services in the form of API (Application Programmable Interface). As a result any application which has been developed using these middleware services will not have the burden to consider security.

Time: Jan 2006-May 2007

Current status: We have developed the following modules of SMARKS.

 

I) An Impregnable Lightweight Device Discovery (ILDD): In order to ensure security and privacy in pervasive computing scenario, we need a mechanism to maintain a list of valid devices which will also prevent malicious devices from participating in any task. As an attempt to address this issue, we have modified the well known H-C (Human-Computer) authentication protocol. It is based on the theme LPN (Learning Parity with Noise. Along with two separate models for both large and small networks, the model presents several possible attack scenarios with their solutions.

 

II) A Formalized Trust Framework (FTF): As pervasive computing applications largely depend on the mutual collaboration of devices present in the vicinity of the ad hoc network, the trust in pervasive computing becomes fundamental. Here we present a context specific, reputation based trust model that satisfies a wide variety of trust properties. To the best our knowledge, our model is the first formal trust model for pervasive computing which includes: 1. A recommendation protocol with multi hop recommendation addressing capability. 2. A flexible behavioral model with ability to handle interactions represented as vectors.

 

III) Malicious Recommendation Protector (MalPro): Though the issue of pervasive trust has been addressed with due importance in many models, importance of malicious recommendation, an integral part of trust, has been undermined. The performance of any reputation based trust model is directly related with its capability in restricting malicious recommendations. Here we have developed a lightweight statistical method for identifying malicious recommendation and hence weeding out malicious devices.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: