Home > Events > Oral Candidacy - Afzal Hossain

Oral Candidacy - Afzal Hossain

Start: 8/16/2018 at 2:00PM
End: 8/16/2018 at 5:00PM
Location: 258 Fitzpatrick Hall
Attendees: Faculty and students are welcome to attend the presentation portion of the defense.
Add to calendar:
iCal vCal

Afzal Hossain

Oral Candidacy

   August 16, 2018      2:00 pm      258 Fitzpatrick

Adviser:  Dr. Christian Poellabauer

Committee Members:

               Dr. David Hachen      Dr. Gregory Madey      Dr. Aaron Striegel                

Title: 

“Context-driven and Resource-Efficient Crowdsensing”

Abstract:

Mobile crowdsensing is the technique to extract or infer information about a person or group from smartphone or wearable data collected in opportunistic or participatory manner. This technique is widely used in psychology, social science, and mHealth research studies. Energy is a primary problem in smartphone sensing systems. Therefore, an efficient sensing system needs to maintain reasonable data quality while being resource conscious. Context-awareness plays the critical role in balancing the trade-off between data quality and energy. Moreover, it is very critical to activate the participatory sensing tasks at the right moment (i.e., during the right contexts) to ensure that the data quality is high and the resulting conclusions are meaningful.

However, context-awareness in a system can also lead to increased energy consumption. Therefore, a research challenge is to design a balanced system that is context-aware, energy-aware, while at the same time providing sufficient quality data for analysis. The proposed thesis will investigate techniques to address these challenges. The work will result in the design and implementation of a flexible crowdsensing platform where both opportunistic and participatory sensing activities can be configured in a context-driven way. To do so, it will rely on an energy efficient context engine that accumulates sensor data to provide various contexts to the client system. It will also provide a framework to schedule sensing tasks using complex contexts.