Adaptive modulation and subcarrier allocation for multiuser OFDM/CDMA systems

Orthogonal frequency-division multiplexing (OFDM) is a promising technology to resist Intersymbol Interference (ISI) for transmitting data at extremely high rates, splits up the data stream and sends the data symbols simultaneously at a drastically reduced symbol rate over a set of parallel subcarriers. The focus of this thesis is on the adaptive modulation and subcarrier allocation methods for single- and multi-service transmissions with limited transmitted power in the Multi-Carrier DS-CDMA (called OFDM/CDMA in this thesis) system. Conventional FDMA and TDMA have hard capacity limitations, since they use a finite number of orthogonal resources. The comparisons of the OFDM/CDMA, OFDM Interleaved-FDMA and OFDM Group-FDMA have been analyzed for the data transmission. Because of the near-unity reuse factor, one of the important merits of CDMA, the combination of OFDM and CDMA, known as OFDM/CDMA, brings substantial increase in the transmission data rate per unit bandwidth when the frequency resources are limited. Even though the multiple access interference exists in OFDM/CDMA systems, better performance can still be achieved than that in other OFDM systems. Based on the OFDM/CDMA system and perfect channel estimation, a novel adaptive modulation method is developed for users to perform the single-service transmission with the given bit-error-rate (BER), data rate and transmitted power. It is significant that the quality of service (QoS) of this service transmission should be guaranteed and system capacity should be improved under the condition of the limited channel resource and power. Comparably, Some related research works emphasize particularly on the maximization of data rates, minimization of BER or minimization of power consumption. From a novel aspect to enhance the transmission efficiency in this thesis, a suboptimal solution to minimize theinterference from each user is put forward in the first step. Then the bit allocation adjustment scheme in the second step is presented to make the best use of the subchannels and reduce the interference further. By evaluating the performance of this algorithm and other related algorithms, we can know that the proposed algorithm for single-service provides better BER performance, higher data rates and higher system capacity, which is more vigorous to meet the demand of supporting more users in the future. For multi-service (Voice and Data) transmissions, voice and data services have different priorities, BER requirements, transmission rates and power consumption. Based on some related research works, these respective features of voice and data transmissions, such as delay tolerances and power consumptions, are slightly considered. In order to meet requirements of voice and data respectively, the subcarrier allocation and bit loading are adopted for the voice transmission at first. Then an adaptive modulation scheme for the data transmission is developed to maximize the data rate of each user according to the remaining subchannels and transmission power. Consequently the realtime transmission and QoS of voice are guaranteed and the data throughput is maximized simultaneously. Additionally, limited resources, i.e., the bandwidth and transmission power, are utilized efficiently. By comparing the performance between the proposed algorithm for multi-service and the other two adaptive schemes, we can know that higher data rates, lower BERs and outage probabilities can be achieved. The above two algorithms are performed based on the perfect channel estimation. However, the ideal estimation is not practical. Under the condition of the inaccurate channel information, the performance of the proposed algorithms are analyzed when the channel is overestimated or underestimated. The analysis and simulations show that the proposed algorithms can still keep tolerable performance based on some given QoS requirements when the channel estimation is not ideal. Since the proposed algorithms have the tolerance to the channel estimation errors, the channel estimation methods with lower complexity can be applied. Thus, the potential complexity problem of the adaptive system is reduced. Generally, it has been concluded that the proposed adaptive OFDM schemes are efficient and robust methods for transmitting data at high rates.DOCTOR OF PHILOSOPHY (EEE