APCompCode: an Optimized Solution for Over-the-Air Computing in Wireless Communications
Weijun Zheng,Jinjiang Tang,Ying Wang,Zheng Wang,Bin Du,Qingqing Wu
DOI: https://doi.org/10.1109/icicn62625.2024.10761666
2024-01-01
Abstract:With the rapid development of the Internet of Things (IoT) and big data technologies, modern communication systems face significant demands for data transmission and computational capabilities. Traditional communication protocols and methods exhibit notable limitations when dealing with large-scale datasets, particularly spectrum efficiency and computational accuracy. Over-the-air computation (AirComp), an emerging wireless communication technology, significantly enhances data processing efficiency and spectrum utilization through signal aggregation during transmission. However, AirComp faces challenges in practical applications, especially the destructive overlap of constellation points at the receiver, which complicates the decoding process and severely impacts system reliability and performance. To address these challenges, this paper proposes a novel modulation and coding scheme named APCompCode. This scheme first assigns unique weights to channel signals, followed by amplitude and phase aggregation of signals within the transmission channel. This method effectively resolves the destructive overlap of constellation points at the receiver, thereby improving decoding accuracy and overall system robustness. Through a series of simulation experiments, this paper verifies the bit error rate of the APCompCode scheme under different channel numbers and signal-to-noise ratio conditions. Compared with traditional modulation schemes, the results show that APCompCode exhibits excellent performance in various communication environments.