You are here: Project > Technical Highlights > Mobility and QoS Management > 
27.3.2017 : 12:35

Mobility and QoS management

A framework to support Quality of Service (QoS) and mobility in the context of QoSMOS has been specified in WP5.

QoS and mobility framework

The analysis of this framework contributed to the identification of the challenges raised by the management of the QoS in a dynamically changing radio environment, where, contrary to legacy wireless networks, QoS cannot be guaranteed by an exclusive access to radio spectrum. In particular, two types of mobility are faced in QoSMOS: spectrum mobility: when the whitespaces used for opportunistic communications are pre-empted upon appearance of incumbent users, the opportunistic user must vacate the resources and the service needs to be promptly handed over to another available spectrum portion; physical mobility: the above procedures may occur in addition to mobility management procedures dealing with the physical mobility of the served user.

From these challenges, the concept of a Cognitive Manager for Resource Management (CM-RM) has been defined (see further below), and requirements to manage the QoS and to ensure the protection of the incumbents have been specified. These requirements have led to the elaboration of a set of QoS and mobility functions to be implemented in the CM-RM, as depicted in the figure below.


QoS and mobility functions [D5.2]


Cognitive Manager for Resource Management

The concept of a Cognitive Manager for Resource Management has been introduced to manage the problem of efficiently enforcing QoS for wireless networks with intermittently available spectrum resources. This QoSMOS functional entity is complementing the Cognitive Manager for Spectrum Management (CM-SM) introduced in WP6.

A detailed CM-RM reference model has been defined, based on the description of its functional blocks and interfaces (internal and external) and on topological views for centralized and distributed architectures. The description of the CM-RM is completed by the study of a set of operations, defined through Message Sequence Charts, in which the CM-RM is involved.

The reference model was first presented in [1] and then slightly revised in [2]. The functional entities of the CM-RM are:

  • Networking domain Cognition (NC);
  • Terminating domain Cognition (TC);
  • Resource Allocation (RA);
  • Resource control Support (RS);
  • Resource Exploitation (RE);
  • Access Control (AC);
  • Mobility Control (MC).

They are organized in two groups, the ones dedicated to the Resource Control (RC) and those reflecting the Resource Usage (RU) by the opportunistic users, accordingly to the relevant topological domains, here terminating and networking domain (see the highlights on system architecture and WP2 paper [3] for details):



Reference model of the Cognitive Manager for Resource Management [2]


The full description of the framework and the CM-RM reference model, as well as the algorithms for transmit power control, resource allocation and interference management presented in [1] and [2] for the target use-case scenarios for cognitive ad hoc network, cognitive femtocell and cellular extension in whitespace, is reported in QoSMOS deliverables D5.1, D5.2 and D5.3. D5.3 also presents various tools for performance evaluation.

Adaptation Layer

The Adaptation Layer (AL) developed in QoSMOS (see also D2.3) allows interworking of remote blocks, managed dispatching of messages to functional units and monitoring the status and triggering alarms if needed. The kernel of the functionalities resides in the AL_CORE, and the interface to the QoSMOS entities is placed in the AL_END, see the following figure.


Internal architecture of the Adaptation Layer [D5.3]

CM-RM solutions in target scenarios

The QoSMOS systems architecture (see the highlights on architecture) is designed to be applicable to a number of scenarios (see the highlights on scenarios). Example solutions for resource management in the three main target scenarios of QoSMOS are illustrated in the following. Further details are available from the cited references.

A cognitive access control applicable to the cellular extension in TV whitespace was studied by simulations for various QoS metrics under different system configurations. For example, here below are shown the connection dropping rate (CDR) and connection blocking rate (CBR) against time before and after a cognitive access control (CAC) algorithm is applied.


CDR and CBR vs time: with CAC inactive, in reactive mode and in preventive mode [D5.3]

For the femtocell scenario the downlink power control algorithm detailed in the following figure, was shown by simulations to be capable of sensibly reducing the interference to nearby macrocell user equipments (UE) caused by femto-user equipments (FUE).


Message sequence chart for the downlink power control [D5.3]

For the cognitive ad hoc network scenario, it was developed the algorithms for the selection of the operating channels illustrated below. Simulation results showed that it controlled well the interference to incumbent users.


Example channel selection [D5.3]



 [1] G. Mange, C. Rosik, S. Leveil, U. Celentano, D. Olasunkanmi and A. Kamran, "Cognitive resource management for QoS support in mobile opportunistic communications", Proc. Future Network & Mobile Summit (FuNeMS 2011), Warszaw, Poland, 15-17 June 2011.

[2] S. Leveil, C. J. Le Martret, H, Anouar; K. Arshad, T. Zahir; J. Bito; U. Celentano; G. Mange; J. Rico, A. Medela, "Resource management of centrally controlled cognitive radio networks", Proc. Future Network & Mobile Summit (FuNeMS 2012), Berlin, Germany, 4-6 June 2012.

[3] U. Celentano, B. Bochow, C. Lange, F. Noack, J. Herrero, B. Cendón, O. Grøndalen, V. Mérat, C. Rosik, "Flexible architecture for spectrum and resource management in the whitespace", Proc. Int. Symp. Wireless Personal Multimedia Commun. (WPMC 2011), Brest, France, 3-7 October 2011.