| 1. IntroductionOrthogonal Frequency Division Multiplexing (OFDM) is a widely used multi-carriermodulation scheme which divides the multipath fading channel into a number of parallelfrequency dependent flat fading channels. It has been suggested that multiuser OFDMsystems employ adaptive subcarrier allocation as well as adaptive bit loading. Byadaptively assigning subcarriers depending on channel characteristics, multiuser OFDMcan take advantage of channel diversity among users in different locations.In multirser environment, the so-called 'multiuser diversity' by dynamically allocatingsubcarriers among the users can be used. Since the subcarriers that appear to be in deepfade for one user may not be in deep fades for others, spectral efficiency can be improved,and equivalently, transmit power can be reduced.Two classes of resource allocation schemes exist: fixed resource allocation anddynamic resource allocation. Fixed resource allocation schemes, such as time divisionmultiple access (TDMA) and frequency division multiple access (FDMA), assign anindependent dimension, e.g. time slot or subchannel, to each user. A fixed resourceallocation scheme is rigid regardless of the current channel condition. On the other hand,dynamic resource allocation allocates a dimension adaptively to the users based on theirchannel gains. Due to the time-varying nature of the wireless channel, dynamic resourceallocation makes full use of the multiuser diversity to achieve higher performance. Twoclasses of optimization techniques have been proposed in the dynamic multiuser OFDMliteratures: margin adaptive (MA) and rate adaptive (RA). The objective of MA is toachieve the minimum overall transmit power under the constraints on the users' data rate orbit error rate. While the objective of RA is to maximize each user's error-free capacity witha total transmit power constraint.2. Margin AdaptiveA dynamic multiuser subcarrier and bit allocation algorithm with low computationalcomplexity for wideband OFDM downlink transmission is proposed to exploit themultiuser diversity in this paper.The problem of margin adaptive has been discussed first. We propose a new algorithmthat defines and solves the problem as a circular of two allocation tasks, which can reducethe complexity of calculation effectively. Dynamic channel allocation for subcarrier isconducted, after the initial allocation of subcarrier completed, further allocation to reducetotal power is made. Then the associated subcarrier swapping is done between users if thepower reduction is maximum between the users. Or a subcarrier from one user to anotheris relocated and the related bit loading is adjusted as described before. After the subcarrierswapping or the relocation operation is completed, updating of all the power reductionfactors is necessary, and continues this process until all the power reduction factors arenegative. This means will reduce the total transmit power further by either two subcarrierswapping or one subcarrier relocation. In order to reduce the complexity, the iterationtimes of the algorithm can be controlled to make a satisfied condition. Once the subcarriershave been allocated, the single user allocation methods, for example greedy method, canbe used to solve the problem. Then after a fixed interval, the subcarriers can relocated byusing dynamic channel allocation algorithm with greedy method too. This proposedalgorithm can greatly reduce computational cost that is the most critical obstaclepreventing the applications of the allocation algorithms. At the same time, a goodperformance is guaranteed. Furthermore, the algorithm can be used in a variousenvironments by changing the time interval between two subcarrier allocations.3. Rate AdaptiveIn the paper, a subcarrier adaptation method that maximizes the total data rate ofmultiuser orthogonal frequency division multiplexing systems in a down link transmissionis developed. The data rate maximization problem that multiusers could share a subcarrieris formulated. The assumption, that says only one user can use a special subcarrier, isproved mathematically. Precisely, the subcarrier assignment strategy for multiusers tomaximize the data rate of a specific subcarrier in a downlink multiuser OFDM system isthat the subcarrier should be assigned to only one user who has the best channel gain forthat subcarrier. Base on this, the model could be simplified and the dynamic channelallocation methods to solve the subcarrier allocation problem can be adopted. Sincemaximum modulation level supported by each subcarrier is limited, the maximum systemthroughput can be obtained by searching all possible bit rate and taking the maximum onein satisfaction of the power constraint. In other words, the optimal problem can beequivalently decoupled as a fast searching problem.4. Simulation and ConclusionThe simulations of the improved resource allocation algorithm on computer withMATLAB are performed. The simulation results indicate that the proposed adaptivedynamic subcarrier and bit allocation algorithm offers about 4-5dB performance gain overOFDM system using modulation with fixed subcarrier allocation. |