射频能量收集为下一代能量受限无线中继网络提供了新的供电方案，它大大延长了网络节点的寿命，引起国内外学者的广泛关注。目前相关研究主要集中在理论领域，且都假设理想瞬时信道信息（CSI）。然而，由于非理想信道估计和能量限制，这种理想假设难以实现，非理想条件下实用能量收集中继系统成为本领域亟待研究的课题。. 本项目旨在开发实用型非相干能量收集中继系统，它无需瞬时CSI，彻底避免了信道估计并大幅度提高了节点能量利用率。具体而言：本项目将为放大-转发和解码-转发网络设计非相干能量收集中继协议，提出标准化的非相干无线信号和能量同时传输（SWIPT）架构。对任意非相干调制信号，本项目将设计最大似然检测器并分析其误码率，然后通过设计最优能量收集参数和最佳中继位置来优化系统性能。最后，我们将对新提出的能量收集协议做深入比较，确定各种现实衰落环境下的最优协议，为无线能量收集中继系统提供实用非相干解决方案。

Radio-frequency energy harvesting (EH) is an emerging solution to power the next generation energy-constrained wireless relay networks. Currently, EH has attracted intensive attention from all over the world due to its capability of prolonging the lifetime of wireless terminals. To our knowledge, current research only dealt with the theoretical aspects of EH relay networks, based on the ideal assumption of perfect instantaneous channel state information (CSI). However, this ideal assumption cannot be satisfied in practical EH relay networks due to imperfect channel estimation and energy constraints, thus motivating an urgent need to study practical EH relay systems without the ideal assumption. . This project aims to develop realistic noncoherent EH relay systems, which completely bypass the need for sophisticated channel estimation and hence boost the energy efficiency significantly. Specifically, we will first propose noncoherent EH relaying protocols for amplify-and-forward and decode-and-forward networks, and then develop a unified noncoherent framework of simultaneous wireless information and power transfer (SWIPT). For arbitrary noncoherent modulated signals, we will propose the noncoherent maximum-likelihood detectors (MLD) and analyze its error performance. Then, we will optimize the system performance through the derivation of the optimum EH parameters and optimum relay positions. Finally, a comprehensive comparison of the proposed noncoherent EH relaying protocols will be conducted to determine the superior protocol under various realistic fading environments. We believe that this project will provide a useful solution for practical noncoherent SWIPT.