Abstract
Both Device-to-device (D2D) and full-duplex (FD) have been widely recognized as spectrum efficient techniques in the fifth-generation (5G) networks. By combining them, the FD-D2D aided underlaying networks (FN) has exhibited considerable technical advantages in terms of both spectral efficiency (SE) and energy efficiency (EE). Considering the fact that the performance of FN may be severely affected by users’ workload, the workload-driven FN (WFN) must be investigated. In this paper, a deep learning based transmit power allocation (TPA) method is proposed for automatically determining the optimal transmit powers of co-spectrum cellular users (CUs) and D2D users (DUs) relying on a deep neural network. Unlike the conventional transmit-power-control schemes, in which complex optimization problems must be addressed in an iterative manner (it usually requires a relative longer computational time), the proposed scheme enables each DU to determine its transmit power with a relatively shorter time. Furthermore, an improved iterative subspace-pursuit algorithm, as the performance benchmark, is formulated for WFN. In addition, to reflect the influence imposed by the workload, the penalty-based statistical sum-date-rate (PSS) can be employed as the performance metric of WFN. Numerical results show that the proposed scheme is capable of achieving a PSS comparable with that of the traditional iterative-based algorithms even under heavy-workload scenarios, but the computational complexity of the former can be significantly reduced.
Highlights
W ITH the rapid development as well as the increasing commercialization progress of the fifth-generation (5G) communication technology, new services
To express the link gain or distortion caused by Successfulor-Unsuccessful-Transmit (SUT), a penalty coefficient (PC) is introduced in the received SINR expression to represent the performance loss caused by the unsuccessful transmission
We can use the indicator penalty-based statistical sum-date-rate (PSS) to reflect the impact of workload on the performance of workloaddriven FN (WFN): on the one hand, a user with a heavier workload can contribute more to the sum data rate; on the other hand, a user with a heavier workload will definitely impose a stronger interference on its neighbors, eroding the performance of the WFN
Summary
W ITH the rapid development as well as the increasing commercialization progress of the fifth-generation (5G) communication technology, new services Since the FD devices are allowed to concurrently transmit and receive signal over a single spectrum, the spectral efficiency (SE) of the cellular networks (CN) can be doubled as compared to the traditional half-duplex (HD) technology [16]–[18]. On this basis, if we can combine both FD and D2D (that is, to form a new technology called FD-D2D) and give full play to their advantages, the SE of CN will inevitably be further improved. Once an improper TPA is implemented, it is very likely to cause an interference strong enough to run out of control, resulting in a fall back of the performance of CN [19], [20]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
