A management scheme for distributed cross-layer reconfigurations in the context of cognitive B3G infrastructures

  • Authors:

  • G. Dimitrakopoulos;K. Tsagkaris;K. Demestichas;E. Adamopoulou;P. Demestichas

  • Affiliations:

  • University of Piraeus, Department of Digital Systems, 80 Karaoli Dimitriou Street, 18534 Piraeus, Greece;University of Piraeus, Department of Digital Systems, 80 Karaoli Dimitriou Street, 18534 Piraeus, Greece;National Technical University of Athens, School of Electrical and Computer Engineering, Athens, Greece;National Technical University of Athens, School of Electrical and Computer Engineering, Athens, Greece;University of Piraeus, Department of Digital Systems, 80 Karaoli Dimitriou Street, 18534 Piraeus, Greece

  • Venue:
  • Computer Communications
  • Year:
  • 2007

Quantified Score

Hi-index 0.24

Visualization

Abstract

Current research efforts in wireless communications are targeted at the evolution of B3G (Beyond the 3rd Generation) wireless infrastructures. The operation of B3G infrastructures envisions dynamic adaptations to external stimuli, which can be facilitated through the exploitation of cognitive networking potentials. Cognitive networks dispose mechanisms for dynamically selecting their configuration (algorithms and parameter values, at different layers of the protocol stack), through appropriate management functionality that takes into account the context of operation (environment characteristics and requirements), profiles, goals, policies and knowledge that derives from previous experience. This paper focuses on such management functionality and it addresses a problem, dealing with ''Distributed, Cross-Layer Reconfigurations'' (DCLR), which aims at assessing and selecting the most appropriate configuration per network element in a cognitive network. In essence, this work contributes in four main areas. First, a fully distributed formulation and solution to the DCLR problem is provided, which is important for the management of a particular reconfigurable element in a cognitive context. Second, robust learning and adaptation, strategies are proposed, for estimating and gaining knowledge of the performance potentials of alternate reconfigurations. Third, a computationally efficient solution to the problem of exploiting the performance potentials of reconfigurations is provided, in order to rate reconfigurations and finally select the best ones. Finally, results that expose the behaviour and efficiency of the proposed schemes, are presented.