Evaluation of a Mobility/QoS-aware Architecture for Mobile Internet

Nuno Lopes, Maria J. Nicolau, Alexandre Santos

Universidade do Minho
Escola de Engenharia
4710-057 Braga, Portugal

Tel.: +351 253 604430
Fax.: +351 253 604471
E-mail: vascolopes (at) dsi.uminho.pt, joao (at) dsi.uminho.pt, alex (at) di.uminho.pt


Abstract

The massive expansion of wireless networks represents a new world when it comes to application's services which now require new quality demands. This uncontestable reality results in the necessity to find new QoS models or adjust existing ones to the specific characteristics of mobile environments. In the future, Internet must provide acceptable Quality of Service (QoS) to mobile users that are running real-time applications and are moving across different access points at high speeds. The user mobility presents a great challenge to the network layer in order to maintain users on going connections. Currently, the Internet protocol that manages the user mobility at the network level is the Mobile Internet Protocol (MIP). This protocol, when a mobile user changes its point of attachment, maintains the same IP address for mobile node, so that user mobility became invisible to the application level and thus avoiding a connection interruption. Although, MIP standard allows the user mobility while maintaining an uninterrupted connection to an application, it does not have any concerns with the QoS support provided to applications with more strict performance requirements such as real-time applications.

This paper addresses the issue of mobility and QoS management principles as well as the mobility and QoS management integration in the sense of build a QoS-aware architecture for mobile Internet. After covering the mobility and QoS management principles and integration, this paper also proposes a new QoS-aware architecture for mobile Internet. This new architecture takes into account the specific characteristics of mobile networks in order to design an integrated Mobility/QoS-aware management architecture suitable for real-time applications requirements.

The solution avoids network congestion as well as the starvation of less priority DiffServ classes while increasing resource utilization for priority classes while maintaining the QoS offer to MN's applications by making each DiffServ class unchangeable with MNs mobility. The proposed mobility management model is simple, easy to implement, and takes into consideration mobile internet requirements. It proved to be capable of providing Internet applications running in mobile devices with acceptable levels of quality of service. The simulation results indicate that the suggested architecture is able to provide acceptable QoS levels to real-time applications that are running in mobiles devices.


JCOMSS - Journal of Communications Software and Systems (ISSN 1845-6421), Volume: 8 Number: 1, 2012