© 2008 OPNET Technologies, Inc.
University: Julius-Maximilians Universität Würzburg
Performance-Energy Saving Trade-Off
R. Pries, D. Klein
While the focus of previous research was on increasing the performance of distributed systems, energy aspects play now a major role. The performance increase of a system often does not scale linearly with the power consumption. Research aspects are, to look at the trade-off between performance and energy savings. This requires that not only one side of a distributed system is looked at, but the overall system has to be evaluated because an energy-related optimization on one side can lead to more energy consumption on another part of the distributed system. The research in this area is carried out in the COST project on energy efficiency in large scale distributed systems.
Performance Evaluation of the IEEE 802.11 WLAN protocol
The Wireless LAN standard, officially known as IEEE 802.11, gained an enormous influence in wireless local area communication. Besides, compared to traditional mobile networks, the technology is cheap and the supported data rates with up to 54 Mbps are huge. Therefore, mobile network providers consider the integration of Wireless LAN technology into their legacy mobile networks. Users will be allowed to perform seamless handovers between the different technologies, such as WLAN, GPRS, or UMTS. However, the Wireless LAN technology has only rudimentary support for Quality of Service (QoS). Future Medium Access Control (MAC) protocol extensions will add such functionality, but it is still unknown how these mechanisms will perform in large-scale deployments. Our research focuses on these issues. This includes the following topics.
Wireless LAN systems, following the IEEE 802.11b and IEEE 802.11g standards, work in the 2.4 GHz frequency band. Data rates of up to 54 Mbps are supported. However, only three non-overlapping channels are available, which is not sufficient to deploy large-scale Wireless LAN networks, e.g. in office buildings. Therefore, overlaps of WLAN cells will be seen in practice. The additional interference within the overlapping area causes great quality degradation due to an increase in packet error probability. The simple Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) mechanism can not deal with this kind of problems, such that high levels of unfairness occur. The performance is greatly degraded and QoS can not be kept at an acceptable level. Future advancements in the MAC protocol, i.e. the Point Coordination Function (PCF) and Hybrid Coordination Function (HCF), might ease the problems. Both approaches are meant to add QoS functionality within single cells. Our research aims at using the different priority schemes to solve the fairness problems in overlapping cells. Nevertheless, service differentiation for various traffic classes such as voice, video, or best effort should be maintained.
Handover mechanisms are of great importance in all wireless systems. Depending on the QoS requirements, different handover mechanisms can be defined. The local Wireless LAN standard defines certain mechanisms to move a connection to a different Access Point when clients move to a new location. However, the mechanisms are restricted to the OSI layer 2, which means that only handovers within a single IP subnet can be supported. In order to seamlessly move within larger areas, handovers on OSI layer 3 have to be added. The well known Mobile IP (MIP) standard provides a simple solution. However, MIP cannot support QoS, such that a handover might take too long to keep the quality of e.g. a voice call at an acceptable level. More advanced handover techniques are necessary. Nevertheless, the problem becomes even worse when handovers between different wireless technologies, such as UMTS, GSM, or WLAN, should be possible. This is the goal for 4th generation mobile networks (4G). The mechanisms for such vertical handovers need to integrate the different architectures and must support handover policies that utilize the individual strengths of the different wireless technologies in various scenarios.
Simulative Performance Study of a Policy-based Vertical Handover Strategy in Heterogeneous Wireless Networks
R. Pries, A. Mäder, and D. Staehle
The integration of different wireless access technologies like UMTS, WLAN, WiMAX, and meshed networks into a heterogeneous network is one major topic for the near future. Since every access technology has been designed from a different perspective regarding user behavior, supported applications, and QoS requirements, such a heterogeneous network has to take into account the strengths and weaknesses from every technology.
The goal of this research project is to integrate WLAN and UMTS into one system and to design policies for the handover between these technologies. This work is based on two well-known architectures to couple these technologies: tight as well as loose coupling. Secondly we develop a combined handover and admission control for both technologies which can be either centralized or of distributed manner. This includes either one or several network entities which store the current situation of the network. In order to enable a comparison of both access technologies, an equivalent bandwidth metric has been designed for UMTS as well as WLAN.
This allows fine-grained handover policies which are far beyond simple best-connect strategies. Due to the evaluation of the current situation in every access technology, our framework enables policies which consider the key strengths and weaknesses of every technology in different situations.
1. Simulation techniques for performance analysis
This class introduces the basics of simulation as a technique to analyze the performance of distributed system, such as computer networks and telecommunication systems. The class follows the content describtion below:Content:
OPNET Technologies, Inc. is a leading provider of solutions for managing networks and applications. OPNET's best-in-class solutions address application troubleshooting, application monitoring, network monitoring, network configuration management, capacity management, and network simulation. OPNET’s solutions have been operationally proven in thousands of customer environments worldwide, including corporate and government enterprises, government and defense agencies, network service providers, and network equipment manufacturers. For more information about OPNET and its products, visit www.opnet.com.