Hierarchical modulation with signal space and transmit diversity in Nakagami-m fading channel.

Abstract

Hierarchical modulation (HM) is a promising scheme for wireless image and video transmission, exploiting the benefits of unequal error protection to ensure enhanced system performance. However, there is a limiting factor to the benefits of using only hierarchy to improve bit error rate (BER) performance of a transmission system. Diversity, namely signal space diversity (SSD) and Alamouti transmit diversity (ATD), can be introduced to improve BER performance results for HM systems. This dissertation presents the BER analysis of hierarchically modulated QAM with SSD and using maximal ratio combining (MRC) to retrieve the transmitted symbol from 𝑁 receiver antennas. In addition, the study includes the BER analysis of an identical system in an ATD scheme employing two transmit antennas and 𝑁 receiver antennas with MRC. SSD comprises of two fundamental stages: constellation rotation and component interleaving. The angle at which the constellation is rotated can affect the performance of the system. In the past, the rotation angle is determined based on a design criterion which maximizes the diversity order by minimizing the Euclidean square product or, alternatively, minimizes an SER expression. In this dissertation, a simple method for determining a rotation angle at which system performance is optimal for hierarchical constellations is presented. Previously, the BER analysis for HM involves an intricate approach where the probability of an error occurring is determined by considering the probability of a transmitted symbol exceeding past a set decision boundary. This dissertation presents the Nearest Neighbor (NN) union bound approach for determining an accurate approximation of the BER of an HM system with SSD. This method of analysis is later extended for an ATD scheme employing HM with SSD. Although introducing diversity elevates the system performance constraints on HM, it does so at the cost of detection complexity. To address this issue, a reduced complexity maximum-likelihood (ML) based detector is also proposed. While the conventional ML detector performs an exhaustive search to find the minimum Euclidean distance between the received symbol and all possible modulated symbols, the proposed detector only considers the nearest neighbors of the received symbol. By reducing the number of comparisons, a complexity reduction of 51.43% between the proposed detector and the optimal detector for 16-QAM is found.