A Suboptimal Algorithm for PTS-Based PAPR Reduction

In Orthogonal Frequency Division Multiplexing (OFDM) systems, high Peak-to-Average Power Ratio (PAPR) is a common issue that can lead to signal distortion and reduced power amplifier efficiency. The Partial Transmit Sequence (PTS) algorithm is an effective method for PAPR reduction, but it suffers from high computational complexity, especially when dealing with large numbers of subcarriers. To balance performance and computational requirements, researchers have developed suboptimal variants of the PTS algorithm. The core concept of suboptimal PTS algorithms is to reduce computational overhead by either restricting the search space or employing heuristic approaches instead of exhaustively searching all possible phase combinations. Common implementation strategies include: Group Optimization: Subcarriers are partitioned into groups where phase factors are optimized independently within each group, significantly reducing the global search space. In code implementation, this can be achieved using grouping functions and applying phase rotation optimization per group. Iterative Optimization: Phase factors are adjusted progressively through iterations to approach near-optimal solutions, rather than computing all combinations simultaneously. Algorithmically, this can be implemented using gradient descent or other iterative optimization techniques with early termination conditions. Threshold-Based Optimization: A target PAPR value is predefined, and the search terminates immediately when this threshold is reached, avoiding unnecessary computations. This approach typically requires real-time PAPR calculation functions and conditional break statements in the optimization loop. Although these methods cannot guarantee global optimum solutions, they achieve significant PAPR reduction with substantially lower computational costs in practical applications. Suboptimal PTS algorithms are particularly suitable for communication systems with high real-time requirements, such as 5G and Wi-Fi systems, where they can be implemented using optimized matrix operations and parallel processing techniques for improved efficiency.