Contribution on joint radio resource management in beyond 3g networks

• Autores: Lorenza Giupponi
• Directores de la Tesis: Ramón Agustí Comes (dir. tes.)
• Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2007
• Idioma: español
• Tribunal Calificador de la Tesis: Fernando José Casadevall Palacios (presid.), Oriol Sallent Roig (secret.), José María Hernando Rábanos (voc.), Mariano Fernández Navarro (voc.), Antonio Valdovinos Bardaji (voc.)
• Materias:
  • Ciencias tecnológicas
    • Tecnología de las telecomunicaciones
      • Radiocomunicaciones
• Texto completo no disponible (Saber más ...)
• Resumen

  • This work represents a contribution on Joint Radio Resource Management (JRRM) in a Beyond 3G and heterogeneous network.

    JRRM is the envisaged process to manage dynamically and co-ordinately the allocation and deallocation of radio resources in a heterogeneous radio access network.

    JRRM strategies may be useful to support a variety of objectives, such as avoiding disconnections due to lack of coverage in the current Radio Access Technology (RAT), blocking due to overload in the current RAT, possible improvement of QoS by changing the RAT, support of user's preferences in terms of RATs, support of operator's preferences for RATs usage or load balancing among RATs.

    To address these objectives and to define a comprehensive JRRM framework, the main problem we have to face is that the information on which the JRRM decision has to be based is in general dissimilar, qualitative and vague. For example, the different inputs coming from different RATs are in general not directly comparable, techno-economic and subjective inputs such as the operator's preferences and the user demand have to be taken into account, etc.

    In this context, we propose an innovative framework based on a fuzzy neural controller (FNC). The advantage of this choice is twofold. On the one hand, we can exploit the capability of fuzzy logic controllers of making effective decisions in situations where the available sources of information are interpreted qualitatively and inexactly. On the other hand, by improving the fuzzy logic controller with the learning capabilities of neural networks, we provide a framework capable of interacting with the surrounding environment, and accordingly self-tuning and acting, being that the basis of the so called cognitive networks. The thesis is organized into three parts.

    The first part of the thesis provides the reader with useful information to be able to fully understand the contents of the following two parts.

    In the second part of the thesis we deal wit