Enhanced Uplink (funded by T-Mobile)
Analytic Models for UMTS Radio Network Planning
|Research project financed by|| |
|T-Mobile Germany, Bonn||Heads||Prof. Dr.-Ing. Phuoc Tran-Gia,|
|Dr. Dirk Staehle|
|Partners||Dr. Hans Barth|
|Funded period||2000 - 2006|
The scientific cooperation with T-Mobile started in the year 2000 and is still ongoing. The general focus of the project is the development of algorithms for the UMTS radio network planning software PegaPlan.
Current Project: Interference Prediction for the Enhanced Uplink (2005-2006)
The purpose of the current project is the inclusion of the Enhanced Uplink (or E-DCH, HSUPA) in the UMTS planning process. This requires the calculation of the interference densities in the cell coverage area under consideration of the new features of the E-DCH. These features comprise amongst others a reduced transport time interval of 2ms, hybrid ARQ and fast NodeB controlled rate scheduling which is performed by scheduling grants for the allowed uplink transmit power. Further projects will consider the effects of scheduling grants of non-serving NodeBs and the NodeB hardware dimensioning with new radio bearer types like the HS-DSCH for the HSDPA service.
Modelling of aggragated traffic streams at the NodeB (2004-2005)
A generic model framework for aggregated mobile user traffic was developed, which uses an recursive approach to capture the various layers of data traffic like web or video. The recursive model approach is flexible enough for a wide spectrum of traffic types but at the same time provides an unified interface for the simple calculation of statistics like mean channel utilisation or mean holding times.
Dimensioning of hardware components at the NodeB(2003-2004)
The objective of this project was the develoment of an algorithm for the dimensioning of hardware elements in the base station (NodeB). The algorithm can be used with an arbitrary number of service classes and considers the impact of the WCDMA soft cell capacity, which describes the phenomenon that the air interface capacity in UMTS is a stochastic factor and depends on the interference situation in the system. The algorithm delivers the optimal number of required hardware elements for which the soft (due to the air interface) and hard (due to hardware capacity) blocking probabilities fullfill given target values.
Impact of Packet-Switched Services on UMTS Radio Network Planning (2001-2002)
A key difference of UMTS networks in contrast to conventional GSM networks is the possibility to offer a large variety of services to the UMTS users. Many of these - in the context of wireless networks - novel services are packet-switched, circuit-switched services like the telephony in GSM networks will become less important. These new technologies and services have to be considered in the planning and optimization process of UMTS networks. The project comprises the derivation of a traffic model for various kinds of applications like audio and video streaming services or web and WAP traffic. These models are used to generate traffic in a simulation of a UMTS network. This simulation includes the technical specifications which are applied in UMTS to perform admission and rate control.
Analytic Models for Predicting the Coverage Area of UMTS Networks (2000-2001)
The focus of this project was on the development of algorithms and analytical models for the uplink cell coverage area prediction in UMTS mobile radio networks. The first project outcome was a snaphot based approach to compute the interfence densities in a UMTS network consisting of a set of base stations and mobile stations which my operate with various services. While this approach was very flexible and accurate, it turned out to be quite time consuming for the planning of very large networks. The next step was therefore the development of an analytic method which includes the interference interdependencies between the base stations. The model is based on a traffic theoretic approach which requires to calculate the steady state distribution of the number of mobiles in the system. Although this analytic method is inherently less computational complex in most network scenarios, the demand for computing power was further decreased with the use of a state space reducing approximation method based on the Kaufman-Roberts algorithm.