The flexible architecture of the open radio access network (O-RAN) provides effective support for the deployment of federated learning (FL). However, existing FL schemes in wireless networks often suffer from packet errors, which lead to reduced model test accuracy and increased training delay. In this paper, we propose a robust FL scheme in O-RAN, named adaptive retransmission-based hierarchical FL (HFedAR), which improves model test accuracy and reduces training delay. Specifically, HFedAR first analyzes the similarity of local gradients using a clustering algorithm to associate users with edge servers. Subsequently, we employ an adaptive retransmission mechanism for both edge and global aggregation, thereby facilitating rapid convergence of FL on non-independent and identically distributed data. Considering the limitations of spectrum resources and user energy, we formulate a multi-objective optimization problem to minimize FL training loss and delay. Due to the implicit nature of the objective function, the coupling of decision variables, and network dynamics, it is difficult to solve the problem through traditional convex optimization and machine learning algorithms. Therefore, we derive a convergence upper bound for HFedAR and present a two-stage resource allocation algorithm. The algorithm can jointly make optimal user power and computing resource allocation decisions in the first stage, and user scheduling, retransmission selection, and radio resource block allocation decisions in the second stage. Extensive simulation results on the MNIST and CIFAR-10 datasets demonstrate that the HFedAR scheme significantly improves model test accuracy by 3.5% and 6.5% and reduces training delay by 40.1% and 34.2% compared with existing FL benchmarks.
Kai Qiao, Hongchao Wang, Zi-Hao Zhang et al.· IEEE Transactions on Mobile...· 0 citations
Emerging technologies including wireless power transfer (WPT), integrated sensing and communication (ISAC), and fluid antennas (FAs), have significantly advanced the capabilities and performance of modern satellite communication systems. This paper investigates an FA-assisted integrated sensing, communication, and power transfer (ISCPT) framework for low Earth orbit (LEO) satellite networks, which operates in two phases: 1) an energy-transfer and target-sensing phase (Phase I), and 2) an information-transmission phase (Phase II). Specifically, in Phase I, a space solar power satellite (SSPS) transmits a dual-functional waveform to simultaneously charge multiple LEO satellites and illuminate a sensing target, while in Phase II, these LEO satellites coordinately serve multiple ground user equipments (UEs) leveraging the harvested energy. We formulate a sum-rate maximization problem subject to the SSPS’s transmit power constraint, LEO satellites’ energy harvesting and sensing requirements, UEs’ information rate demands, and the FAs’ movable regions. To tackle the highly-coupled and non-convex optimization problem, a three-stage alternating optimization (AO) algorithm is proposed, which decomposes it into resource allocation, SSPS-side FA placement, and LEO-side FA placement subproblems. In particular, the resource allocation subproblem is reformulated by adopting the Cauchy–Schwarz inequality and semidefinite relaxation (SDR), and is efficiently tackled via the successive convex approximation method. The two FA placement subproblems are addressed leveraging trust-region-based optimization. Simulation results validate the superior performance gains of the proposed algorithm over seven benchmarks and demonstrate that FAs can enhance multi-functional wireless services by adjusting inter-channel diversity according to service types. Notably, a non-trivial trade-off arises among FAs-enabled multi-functional services requiring distinct channel characteristics, as FAs cannot simultaneously provide optimal channel conditions for all services.
Weihao Mao, Yang Lu, Dong Yang et al.· IEEE Journal on Selected Are...· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
