Hierarchical power allocation algorithm for D2D-based cellular networks with heterogeneous statistical quality-of-service constraints

Device-to-device (D2D) communication can increase network coverage, spectrum efficiency and energy efficiency (EE) within the existing cellular infrastructure, which makes it a promising architecture for the future networks. Due to the diversification of services, heterogeneous statistics quality-of...

Full description

Saved in:
Bibliographic Details
Published inIET communications Vol. 12; no. 5; pp. 518 - 526
Main Authors Liu, Yuanfei, Wang, Ying, Sun, Ruijin, Miao, Zhongyu
Format Journal Article
LanguageEnglish
Published The Institution of Engineering and Technology 20.03.2018
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:Device-to-device (D2D) communication can increase network coverage, spectrum efficiency and energy efficiency (EE) within the existing cellular infrastructure, which makes it a promising architecture for the future networks. Due to the diversification of services, heterogeneous statistics quality-of-service constraints are considered in this study, where cellular users are concerned about the delay constraint and D2D user groups pay more attention to the outage probability of data transmission. The power allocation problem of cellular users can be solved by optimising the capacity-payoff power-loss game model. Upon exploiting Lagrange dual decomposition and the Newton iteration method, the power optimisation problem of cellular users is transformed into a parameter optimisation problem. Due to the limited energy resource of D2D users, EE is the focus of D2D user groups. Using fractional programming and convex optimisation techniques, energy-efficient optimal power allocation algorithms of D2D users are proposed subject to the outage probability constraint. As a result, a power allocation algorithm based on hierarchical game is conceived for efficiently solving the power optimisation problem. The simulation results show that the proposed algorithm can obtain a performance improvement compared with other algorithm and converge within a certain number of iterations.
ISSN:1751-8628
1751-8636
DOI:10.1049/iet-com.2017.0488