|
For Full-Text PDF, please login, if you are a member of IEICE,
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
|
Performance Analysis of Power Saving Class of Type I for Voice Service in Two-Way Communication in IEEE 802.16e
Eunju HWANG Kyung Jae KIM Bong Dae CHOI
Publication
IEICE TRANSACTIONS on Communications
Vol.E95-B
No.3
pp.845-856 Publication Date: 2012/03/01 Online ISSN: 1745-1345
DOI: 10.1587/transcom.E95.B.845 Print ISSN: 0916-8516 Type of Manuscript: PAPER Category: Network Keyword: IEEE 802.16e, power consumption, sleep mode, VoIP,
Full Text: PDF>>
Summary:
In IEEE 802.16e, power saving is one of the important issues for battery-powered mobile stations (MSs). We present a performance analysis of power saving class (PSC) of type I in IEEE 802.16e standard for voice over Internet protocol (VoIP) service with silence suppression in two-way communication. On-off pattern of a voice user in two-way communication is characterized by the modified Brady model, which includes short silence gaps less than 200 ms and talkspurt periods shorter than 15 ms, and so differs from the Brady model. Our analysis of PSC I follows the standard-based procedure for the deactivation of the sleep mode, where a uplink packet arrival during a mutual silence period wakes up the MS immediately while a downlink packet arrival waits to be served until the next listening window. We derive the delay distribution of the first downlink packet arriving during a mutual silence period, and find the dropping probability of downlink packets since a voice packet drops if it is not transmitted within maximum delay constraint. In addition, we calculate the average power consumption under the modified Brady model. Analysis and simulation results show that the sleep mode operation for the MS with VoIP service yields 32 39% reduction in the power consumption of the MS. Finally we obtain the optimal initial/final-sleep windows that yield the minimum average power consumption while satisfying QoS constraints on the packet dropping probability and the maximum delay.
|
|