TY - GEN
T1 - Block waterfilling with power borrowing for multicarrier communications
AU - Ko, Seung Woo
AU - Kim, Seong Lyun
PY - 2008
Y1 - 2008
N2 - The multicarrier system has received great attention as a solution to transmitting high-rate data over wireless channels with severe inter-symbol interference, where the optimal power allocation scheme is known as the well-known waterfilling [2]. In this paper, the so called block waterfilling (BW) is proposed, which is a hybrid of waterfilling and constant waterfilling [3]. Within BW, there are three channel blocks, being divided from each other by two thresholds: waterfilling block, equal power block and non-power allocation block. For BW, we propose to use the power borrowing, which plays an important role in minimizing the duality gap, releasing us from fine-tuning of the two thresholds. From numerical examples, we have found that BW shows superior performance to the constant waterfilling in terms of throughput enhancement with small amount of additional computational complexity. The main idea behind BW is to parameterize computational burden of the classical waterfilling, which makes us trade the complexity with the solution quality.
AB - The multicarrier system has received great attention as a solution to transmitting high-rate data over wireless channels with severe inter-symbol interference, where the optimal power allocation scheme is known as the well-known waterfilling [2]. In this paper, the so called block waterfilling (BW) is proposed, which is a hybrid of waterfilling and constant waterfilling [3]. Within BW, there are three channel blocks, being divided from each other by two thresholds: waterfilling block, equal power block and non-power allocation block. For BW, we propose to use the power borrowing, which plays an important role in minimizing the duality gap, releasing us from fine-tuning of the two thresholds. From numerical examples, we have found that BW shows superior performance to the constant waterfilling in terms of throughput enhancement with small amount of additional computational complexity. The main idea behind BW is to parameterize computational burden of the classical waterfilling, which makes us trade the complexity with the solution quality.
UR - http://www.scopus.com/inward/record.url?scp=58149108846&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=58149108846&partnerID=8YFLogxK
U2 - 10.1109/VETECF.2008.298
DO - 10.1109/VETECF.2008.298
M3 - Conference contribution
AN - SCOPUS:58149108846
SN - 9781424417223
T3 - IEEE Vehicular Technology Conference
BT - The 68th IEEE Vehicular Technology Conference, VTC 2008-Fall
T2 - 68th Semi-Annual IEEE Vehicular Technology, VTC 2008-Fall, Conference
Y2 - 21 September 2008 through 24 September 2008
ER -