Full-Duplex Cooperative Rate-Splitting for Multigroup Multicast With SWIPT
Full-Duplex Cooperative Rate-Splitting for Multigroup Multicast With SWIPT
复制标题
DOI：
10.1109/twc.2021.3129881
复制
发表时间：
2022-06-01
影响因子：
10.4
通讯作者：
Yuan, Dongfeng
中科院分区：
文献类型：
--
作者：
Li, Tiantian;Zhang, Haixia;Yuan, Dongfeng
提出了一种下行两组多播系统中的全双工协作速率分割（FD-CRS）方案。在发送端，根据速率分割多址（RSMA）的原理，将两组用户分别请求的两个不同的消息进行分割，然后编码成一个公共流和两个私有流。一组小区中心用户（CCU）依次解码公共流和自己的私有流，然后协作形成分布式波束形成器，辅助小区边缘用户（CEU）进行公共流传输。为了充分利用协作过程中的时间资源，所有CCU都工作在FD模式，使信息接收和转发同时进行。此外，由于牺牲合作者的能量来转发是不公平的，因此通过采用功率分割协议，每个CCU能够从接收信号中收获能量。以最小化系统发射功率同时保证所有组的目标速率为目标，建立了波束形成器、消息分割和功率分割比联合设计的优化问题。利用凸规划的差分方法将非凸问题转化为一种新的形式，并提出一种基于逐次凸逼近的迭代算法来求解该问题，以获得局部最小值.在此基础上，针对信道状态信息不完全的情况，提出了一种结合半正定松弛（SDR）技术和罚函数法的鲁棒算法。虽然我们提出的FD-CRS方案采用了看似浪费能量的无线功率传输技术，但仿真结果仍然证实了所提出方案的优越性，即，在各种用户部署、网络负载和目标速率下的功耗方面，它优于其他基准方案。这是由于FD合作增益、空间复用增益以及功率复用增益的综合利用。
We propose a full-duplex cooperative rate-splitting (FD-CRS) scheme in a downlink two-group multicast system. At the transmitter, two distinct messages requested by the two groups respectively are split and then encoded into one common stream and two private streams, based on the principles of rate splitting multiple access (RSMA). The cell-center-users (CCUs) in one group decode the common stream and their own private stream successively, then cooperatively form a distributed beamformer to assist the cell-edge-users (CEUs) in common stream transmission. To make full utilization of the time resources during cooperation, all the CCUs operate in FD mode to enable information receiving and forwarding simultaneously. Moreover, since it is unfair to sacrifice the cooperator' energy to forward, each CCU is enabled to harvest energy from the received signal by adopting power-splitting protocol. With the objective of minimizing the system transmission power while guaranteeing all the groups' target rates, an optimization problem is formulated to jointly design the beamformers, message splitting and power-splitting ratio. We reformulate the non-convex problem by using the difference of convex (DC) programming, and then propose an iterative algorithm based on successive convex approximation to solve it to obtain a local minimum. Further, a robust algorithm combining the semi-positive definite relaxation (SDR) technique and penalty function method is developed for the case with imperfect channel state information. Although our proposed FD-CRS scheme adopts the seemingly energy-wasting wireless power transfer technique, the simulation results still confirm the superiority of the proposed scheme, i.e., it outperforms the other baseline schemes in terms of power consumption under various user deployment, network loads and target rates. That is attributed to the comprehensive utilization of FD cooperation gain, spatial multiplexing gain as well as power multiplexing gain.