Research On Resource Management Technology In OFDM System

Now the demand on wireless communication services is continually growing, but the available spectrum resources are more and more scarce. Only constantly improving the system spectrum efficiency can deal with it to a certain degree. And the scheduling algorithm with superior performance is an effective way to improve the spectrum efficiency. So it has attracted a lot of attention at present.But the scheduling algorithm designing faces a lot of difficulties and challenges. Firstly, the transmitter can’t obtain the perfect channel knowledge, which makes many channel-adaptive scheduling algorithms cannot be implemented. The compromise is to use the user feedback information. But too huge feedback will make performance degradation as a result of the system load increases. Second is the service quality of the cell edge users. In the co-located antenna system, it pays for the cell edge users’performance with the whole cell performance. But the distributed antenna system can improve the spectrum efficiency and the service quality of the cell edge users.In order to solve the limited feedback problem, in this paper, we consider the problem of the cross-layer scheduling and resource allocation, where utilizing the ACK/NAK in MAC layer and periodic feedback of users’SNR in physical layer simultaneously. This problem is modeled as a Partially Observable Markov Decision Process (POMDP) with cross-layer observations, and a simplified version is proposed to reduce the computational complexity. Numerical experiments show that the performance of our algorithm is relatively close to the upper bound and does not increase the system overhead of feedback.In the traditional DAS study, assuming the combined algorithm is perfect in the receiver. But in fact, with the increase of the combined way’s number, MRC and other combined algorithm will become more sensitive about the channel error. Especially when the number is larger than eight, the performance loss even can reach2dB.A greedy algorithm is proposed for the antenna selection in the distributed antenna system. Firstly, the distributed antenna (DA) selection is determined based on large scale fading. Secondly, relax the constraint and assign the power for the subcarrier by the water-filling algorithm. Thirdly, allocate the power of the subcarrier to the all selected DA. Finally, an integer program is formulated for the complete solution. Numerical results show that the DA selection is sparser and the performance loss is very little.The problem of resource allocation is formulated with the objective of maximizing the cell throughput. In order to solve the problem, this paper decomposes the original problem to several independent sub-problems and the sub-problems are solved from the view of analytic geometry. Numerical results show that the proposed algorithm has satisfactory convergence.