Crosstalk-aware dynamic spectrum management algorithm for green DSL systems

Chee Keong Tan, Teong Chee Chuah, Mohd Saiful Bahri

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Dynamic spectrum management (DSM) techniques mitigate crosstalk in digital subscriber line (DSL) networks by adapting the transmit spectra to the actual noise and channel conditions. Conventional DSM schemes are designed based on single-objective optimization, either belonging to the rate-adaptive or margin-adaptive category. In this paper, an efficient crosstalk-aware DSM (CA-DSM) algorithm which jointly considers both the data rate and power is proposed to search for the best rate-power tradeoff solution based on the network conditions. The crosstalk-aware power strategy prevents transmitters which contribute excessive crosstalk from being allocated high power, thereby reducing the aggregate crosstalk noise in the system. A convex cost function is used to formulate the DSM optimization problem wherein two coefficients are introduced to make the CA-DSM algorithm adaptive to different network conditions. An iterative power update strategy is proposed for the CA-DSM algorithm to minimize the cost function. Convergence properties of the CA-DSM algorithm along with existence and uniqueness of optimal power solutions are examined analytically and illustrated graphically. Simulation results show that the proposed CA-DSM algorithm can provide a significantly better rate-power tradeoff performance compared to existing spectrum management schemes.

Original languageEnglish
Pages (from-to)717-732
Number of pages16
JournalTelecommunication Systems
Volume67
Issue number4
DOIs
Publication statusPublished - Apr 2018
Externally publishedYes

Keywords

  • Crosstalk
  • Digital subscriber line
  • Dynamic spectrum management
  • Energy efficiency
  • Spectral efficiency

Cite this