TY - GEN
T1 - Energy and performance considerations in work partitioning for mobile spatial queries
AU - Gurumurthi, Sudhanva
AU - An, Ning
AU - Sivasubramaniam, Anand
AU - Vijaykrishnan, N.
AU - Kandemir, Mahmut
AU - Irwin, Mary Jane
N1 - Publisher Copyright:
© 2003 IEEE.
PY - 2003
Y1 - 2003
N2 - A seamless infrastructure for information access and data processing is the backbone for the successful development and deployment of the envisioned ubiquitous/mobile applications of the near future. The development of such an infrastructure is a challenge due to the resource-constrained nature of the mobile devices, in terms of the computational power, storage capacities, wireless connectivity and battery energy. With spatial data and location-aware applications widely recognized as being significant beneficiaries of mobile computing, this paper examines an important topic with respect to spatial query processing from the resource-constrained perspective. Specifically, when faced with the task of answering different location-based queries on spatial data from a mobile device, this paper investigates the benefits of partitioning the work between the resource-constrained mobile device (client) and a resource-rich server, that are connected by a wireless network, for energy and performance savings. This study considers two different scenarios, one where all the spatial data and associated index can fit in client memory and the other where client memory is insufficient. For each of these scenarios, several work partitioning schemes are identified. It is found that work partitioning is a good choice from both energy and performance perspectives in several situations, and these perspectives can have differential effects on the relative benefits of work-partitioning techniques.
AB - A seamless infrastructure for information access and data processing is the backbone for the successful development and deployment of the envisioned ubiquitous/mobile applications of the near future. The development of such an infrastructure is a challenge due to the resource-constrained nature of the mobile devices, in terms of the computational power, storage capacities, wireless connectivity and battery energy. With spatial data and location-aware applications widely recognized as being significant beneficiaries of mobile computing, this paper examines an important topic with respect to spatial query processing from the resource-constrained perspective. Specifically, when faced with the task of answering different location-based queries on spatial data from a mobile device, this paper investigates the benefits of partitioning the work between the resource-constrained mobile device (client) and a resource-rich server, that are connected by a wireless network, for energy and performance savings. This study considers two different scenarios, one where all the spatial data and associated index can fit in client memory and the other where client memory is insufficient. For each of these scenarios, several work partitioning schemes are identified. It is found that work partitioning is a good choice from both energy and performance perspectives in several situations, and these perspectives can have differential effects on the relative benefits of work-partitioning techniques.
UR - http://www.scopus.com/inward/record.url?scp=33645230709&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33645230709&partnerID=8YFLogxK
U2 - 10.1109/IPDPS.2003.1213115
DO - 10.1109/IPDPS.2003.1213115
M3 - Conference contribution
AN - SCOPUS:33645230709
T3 - Proceedings - International Parallel and Distributed Processing Symposium, IPDPS 2003
SP - 8
BT - Proceedings - International Parallel and Distributed Processing Symposium, IPDPS 2003
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - International Parallel and Distributed Processing Symposium, IPDPS 2003
Y2 - 22 April 2003 through 26 April 2003
ER -