
For FullText 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.

LowLoadingDelay Parallel Adaptive Loading for Reducing Transmit Power of OFDM System
Ming LEI Hiroshi HARADA
Publication
IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Vol.E90A
No.7
pp.13001311 Publication Date: 2007/07/01
Online ISSN: 17451337
DOI: 10.1093/ietfec/e90a.7.1300
Print ISSN: 09168508 Type of Manuscript: Special Section PAPER (Special Section on Multidimensional Mobile Information Networks) Category: Keyword: orthogonal frequency division multiplexing (OFDM), adaptive loading,
Full Text: PDF>>
Summary:
The conventional successive adaptive loading algorithm, represented by HughesHartogs algorithm, can be used to reduce the transmit power of orthogonal frequency division multiplexing (OFDM) system. However, two major disadvantages exist for this kind of algorithm: One is the long loading time delay caused by the bittobit loading, i.e., only one bit is loaded in every iteration; the other is that there is no flexibility in freely predefining the candidate modulation set before the loading is finished. In order to solve these problems, we propose the lowloadingdelay parallel adaptive loading algorithms aiming at reducing the transmit power under the condition that the data throughput and error rate are maintained to target values. Two improvements are achieved by the new algorithm. One is that it divides the successive adaptive loading into several independent smallscale loading (SSLoading) procedures. "SSLoading" can be performed in parallel mode. To support this, we propose two subband division methods (successive and sorted subband divisions). The simulation results show that for a large range of subband number (1128), the loading time delay can be remarkably decreased (especially for the parallel adaptive loading based on sorted subband division, i.e., SRT parallel adaptive loading algorithm) with neglectable power efficiency loss, compared with HughesHartogs algorithm. The second improvement is that the new algorithm allows us to predefine the candidate modulation set, which provide flexibility for the system design, e.g. we can exclude those rarely used modulation modes. We also reveal that HughesHartogs algorithm is actually a special case of the newly proposed algorithm.

