| OFDM (Orthogonal Frequency Division Multiplexing) and MIMO (Multi-Input and Multi-Output) are believed to be two necessary techniques for the intending mobile wireless communication since they can provide high specturm efficiency and high data rate in mutli-path envieronment.The integrated utility of radio resource is an important support for the future mobile communication systems, which play a crucial role in resaining the frequency efficiency and QoS(quality of service) guarenttee. In the mobile communicationsystem based on MIMO and OFDM, the radio resource manegment is extended from one dimension to multiple dimensions, and from variable rate, variable modulation mode, variable power and bandwidth under single input single output, to the variable number and the different mode of the antennas as well as their working cooperatively, moreover, involving the cross-layer resource management in MAC and PHY. All in all, the issue of radio resource allocation becomes more and more complicated.On the basis of the models for MIMO OFDM systems and the MAC protocol of WLAN (Wireless Local Area Networks), this dissertation researches on the resource management of combining PHY and MAC layer for MIMO OFDM systems. The author's main research contents and contributions are summarized as follows:1. After investigation of the MAC protocol for IEEE 802.11 WLAN, a delay based back-off scheme is proposed, through which the CW (Contention Windows) can be dynamically adjusted according to the requirements of users and the current delay, so that the starvation of the users with low priorities can be avoided under the condition of guaranteeing QoS for the users of high priorities. Its performance has been validated by theoretical analysis and computer simulation.2. Through researching the power allocation, bit loading and sub-carrier allocation techniques, as well as the proportional fairness algorithm, the dissertation proposes a MAC and PHY resource allocation algorithm for multi-user OFDM systems. The algorithm allocates the resource by combining the QoS requirements and the characteristics of OFDM. Furthermore, the number of the acceptable users is estimated by the limitation of the capacity region. This characteristic of the proposed algorithm has been verified by theoretical analysis and the results of simulation to demonstrate the performance in bandwidth and delay for every user. 3. The resource allocation of multi-user MIMO systems is analyzed. Aiming to the limitation of the traditional block diagonal (BD) algorithm that the number of the users must be less than the number of the antennas of the base station, we propose a space-time domain resource allocation algorithm. The resource management is extended to the space-time 2-dimensional domain by combining the selection of the antennas for users in the space dimension with the time slot allocation in the time domain, which makes the users' QoS requirement in average bandwidth and delay can be guaranteed in a given size of time period.4. Based on the study of the resource allocation algorithm of OFDM and MIMO systems, the research is extended to the space-time-frequency 3-dimensional domain for multi-user MIMO OFDM systems, where the interference between some users caused by their high correlation in space domain is avoided by separating them in frequency domain, i.e. allocating them with different sub-carrier, and where the antenna selection and sub-carrier allocation are performed according to the users' bandwidth request and the correlation between different users. Besides, the rate compensation mechanism is adopted in the time domain so as to satisfy the fairness. The principle of the compensation mechanism is that if some user channels appear to be too poor in some duration, its bandwidth is temporarily allocated to other users so as to obtain higher spectrum efficiency; when any one of the channels is getting better, some rate compensation is given to satisfy its band width requirement.
|
PDF Full Text Request