Degree Type


Date of Award


Degree Name

Doctor of Philosophy


Computer Science

First Advisor

Ying Cai

Second Advisor

Wallapak Tavanapong


Mobile devices have brought new applications into our daily life. However, ecient man-

agement of these objects to support new applications is challenging due to the distributed

nature and mobility of mobile objects. This dissertation describes a new type of mobile peer-

to-peer (M-P2P) computing, namely geotasking, and presents ecient management of mobile

objects to support geotasking. Geotasking mimics human interaction with the physical world.

Humans generate information using sensing ability and store information to geographical lo-

cations. Humans also retrieve this information from the physical locations. For instance, an

installation of a new stop sign at some intersection in town is analogous to an insertion of a

new data item into the database. Instead of processing regular data as in traditional data

management systems, geotasking manages a collection of geotasks, each dened as a computer

program bound to a geographical region. The hardware platform for geotasking consists of

popular networked position-aware mobile devices such as cell phones, personal digital assis-

tants, and laptops. We design and implement novel system software to facilitate programming

and ecient management of geotasks. Such management includes inserts, deletes, updates,

retrieval and execution of a geotask triggered by mobile object correlations, geotask mobil-

ity, and geotask dependency. Geotasking enables useful applications ranging from warning of

dangerous areas for military and search-and-rescue missions to monitoring the population in

a certain area for trac management to informing tourists of exciting events in an area and

other such applications. Geotasking provides a distributed and unied solution for supporting

various types of applications.


Copyright Owner

Kihwan Kim



Date Available


File Format


File Size

102 pages