Resource allocation software algorithms for AMC-OFDM systems

Abstract

In recent years, adaptive modulation and coding (AMC) technologies, resource allocation strategies and user scheduling for single-cell downlink orthogonal frequency division multiplexing (OFDM) and orthogonal frequency division multiple access (OFDMA) systems have been widely researched in order to ensure that capacity and throughput are maximised. In terms of AMC technologies, the correlation between the channel coefficients corresponding to the transmitted sub-carriers has not been considered yet. In the literature of resource allocation and user scheduling, either channel coding is not considered or only a fixed code rate is specified. Consequently, with a fixed number of data sub-carriers for each user, all these criteria restrict the flexibility of exploiting the available channel capacity, which reflects negatively on system throughput. At the same time, the presented scheduling algorithms so far managed the data of each user regardless the fair services of all users. The philosophy of this thesis is to maximise the average system throughput by proposing novel AMC, resource allocation and user scheduling strategies for OFDM and OFDMA systems based on developed software engineering life cycle models. These models have been designed to guarantee the scalability, extendibility and portability of the proposed strategies. This thesis presents an AMC strategy that divides the transmitted frame into sub-channels with an equal number of sub-carriers and selects different modulation and coding schemes (MCSs) amongst them rather than considering the same MCS for the entire frame. This strategy has been combined with a pilot adjustment scheme that reduces the pilots used for channel estimation in each sub-channel depending on the measured coherence bandwidth, signal to noise ratio (SNR), and SNR fluctuation values. The reduced pilots are replaced with additional data sub-carriers in order to improve the throughput. Additionally, a novel resource allocation strategy has been introduced in order to maximise the system throughput by distributing the users, transmission power and information bit streams over the employed sub-channels. The introduced method utilises the proposed AMC strategy in combination with pilot adjustment scheme to tackle the problem of channel capacity exploiting efficiently. It presents the throughput as a new cost function in terms of spectral efficiency and bit-error rate (BER), in which both convolutional coding rates and modulation order can be varied. The investigated throughput maximisation problem has been solved by producing two approaches. Firstly, optimised approach that solves the adopted problem optimally using the well known Lagrange multipliers method. This approach requires a huge search processes to achieve the optimal allocation of the resources, which yields a high computational complexity. To overcome the complexity issue, the second approach decouples the considered maximisation problem into two sub-problems based on the decomposition method on the cost of performance particularly for low SNR values. The proposed resource allocation strategy has been developed to work with multi-input-multi-output (MIMO) based AMC-OFDMA systems. In this project, two MIMO transmission criteria are considered, i.e. traditional and eigen-mode. In contrast, a user scheduling algorithm combined with the proposed resource allocation and AMC strategies is presented. The user scheduling algorithm aims to maximize the average system throughput by arranging the users in distinct queues according to their priorities and selecting the best user of each queue individually in order to guarantee a fair user service amongst different priority levels. When the involved users are scheduled, the scheduled users have been passed to the resource allocation to implement the distribution of the available resources. The proposed strategies have been tested over different international telecommunication union (ITU) channel profiles. The obtained simulation results show the superior performance of the introduced approaches in comparison with the related conventional methods. Furthermore, the gradually improvement in the throughput performance of the AMC-OFDM/ODMA system throughout the combination of the proposed strategies is clearly explained.