对LR-WPAN中BPSK信号进行非相干检测是实现低功耗短距离物联网感知层可靠传输的有效方式之一。然而，现有检测方案都无法做到能效、成本和性能间的统筹兼顾。本课题拟以反三角函数泰勒级数展开，观测空间智慧划分和最大似然检测旋转不变性为出发点解决上述亟待解决难题。一方面，将在LR-WPAN现有和未来缩短前导符下，研究适用于传统检测机制的频偏影响估计技术；另一方面，将在无估计/补偿过程的限定下，研究全新检测机制；最后建立一套实验仿真验证系统。通过本课题研究，期望从全新视角实现传统检测机制中的强鲁棒性频偏影响估计量构造和全新检测机制下的强鲁棒性判决统计量设计，进而解决LR-WPAN当前和未来发展对绿色可靠信号检测技术的迫切需求问题。本课题将为低功耗短距离物联网感知层可靠数据传输提供极其新颖方法，将促进其与绿色通信的融合领域发展，对其在智能电网、精细农业和智能医疗等高可靠需求领域应用具有重要意义。

Non-coherent detection of BPSK signals in low rate wireless personal area network (LR-WPAN) is one of the efficient approaches to ensuring high-reliability communications in perceptual layer of low-power short-range IoT netwoks. However, the conventional detection scheme can't pay equal attention to energy efficiency, cost and performance. Based on Taylor series expansion of trigonometric function, intelligent division of observation space and rotation invariance of maximum likelihood (ML) detection, this project intends to solve the above urgent problems. The main topics are frequency-offset estimation schemes with existing and future shortened preamble for traditional detection, a detection scheme without estimation and compensation, and a validation system. Through this project’s study, it is desirable to develop strong-robustness estimator for the existing detection and strong-robustness detection statistic for a new one, and then to meet the urgent need of reliable signal detection technologies in current and future development of LR-WPAN. This project will provide novel ideas for reliable communication in perceptual layer of low-power short-range IoT netwoks, and will promote its integration with green communications. It is very important for the applications requiring very high reliability, such as smart grids, precision agriculture and medical monitoring systems.