Adaptive Multi-Homing Resource Allocation for Time-Varying Heterogeneous Wireless Networks Without Timescale Separation

Weihua Wu, Qinghai Yang, Peng Gong, Kyung Sup Kwak

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)

Abstract

In this paper, we design an adaptive multi-homing resource allocation algorithm for time-varying heterogeneous wireless networks (HetNet), where the algorithm iteration timescale is the same to the network state acquisition timescale. First, the network utility maximization is characterized by a stochastic optimization model. Second, the multi-homing resource allocation (MHRA) algorithm is developed to accommodate the dynamic wireless network states, i.e., time-varying wireless channels between the access points (AP) and mobile terminals and as well the queuing dynamics at the APs. Then, we investigate the tracking error between the MHRA algorithm output and the target optimal resource allocation solution. Based on these results, an adaptive-compensation multi-homing resource allocation (AMRA) algorithm is proposed to offset the tracking error so as to enhance the network utility. Specifically, we give a sufficient condition that the AMRA algorithm asymptotically tracks the moving equilibrium point with no tracking errors. Finally, we derive a tradeoff between network utility and media transmission delay, where the increase of average delay is approximately linear in V and the increase of network utility is at the speed of 1/V with the control parameter V. Simulation results validate the theoretical analysis of our proposed scheme.

Original languageEnglish
Article number7523272
Pages (from-to)3794-3807
Number of pages14
JournalIEEE Transactions on Communications
Volume64
Issue number9
DOIs
Publication statusPublished - Sept 2016

Keywords

  • Adaptive resource allocation
  • Continuous-time Lyapunov drift
  • multi-homing

Fingerprint

Dive into the research topics of 'Adaptive Multi-Homing Resource Allocation for Time-Varying Heterogeneous Wireless Networks Without Timescale Separation'. Together they form a unique fingerprint.

Cite this