A Simulation Setup Validation Framework for Modeling Architectures and Algorithms in Designing Large Scale Wireless Communication Systems P.M.PAPAZOGLOU 1,3 , D.A.KARRAS 2 , R.C.PAPADEMETRIOU 3 1 Department of Informatics & Computer Technology Lamia Institute of Technology, Greece p.m.papazoglou@hotmail.com, papaz@teilam.gr 2 Department of Automation Engineering Chalkis Institute of Technology, Greece dakarras@ieee.org, dimitrios.karras@gmail.com, dakarras@teihal.gr 3 Department of Electronic & Computer Engineering University of Portsmouth, United Kingdom Abstract – Channel allocation is one of the fundamental issues in designing wireless communications systems due to the fact that it determines how the available bandwidth will be managed over the changeable user demands. The limited channel capacity and the increasing requirements for advanced services such as real time video grant channel allocation strategies a special role. Plethora of channel allocation strategies have been proposed in the literature for supporting GSM and 3G communication as well as multimedia data services and tested through generic or specific simulation architectures for medium size networks. The problem, however, still remains, that is how reliable are the simulation results attained, compared to the performance of real world mobile communication systems. In large scale networks, the problem is getting even worse, since generic simulation systems are well adapted to medium scale problems but not to higher complexity systems, where the traffic conditions cannot be predicted. Therefore, some strategies should be involved to investigate the validity of the simulation models implemented in the proposed simulators. This paper, presents a validation framework for such simulation systems based on an architecture involving a hierarchy of several key components in order to manage this important issue. Some initial experimental results are compared favourably to the theoretical ones. Key-words: Simulation Systems, Validation Framework, Bandwidth management, voice services, multimedia services, Channel allocation, 1 Introduction Models and real experiments, to some degree, can only be approximations as having complete control over all of the factors is simply not fully achievable in any real system. This is a basic notion affecting work in simulation verification in general but, also, in mobile communication systems specifically, starting from the seminal research ideas of D.B. Johnson [1]. The basic directions introduced then, are: 1) The “trace emulation” approach, where a trace of the desired network’s behaviour is generated using simulation, and then uses this trace to drive the standard trace modulation system in the operating system kernel of the real machines on the real network 2) The Initial approach to the validation of simulation work is to check the operation of the system according to a number of logical consistency checks. Although initial validation checks give considerable confidence in results, they do not actually fully validate the simulator results matching to the real world. 3) The approach involving a comparison based on the progression of some performance metric as a function of time is considered effective by comparing the results from simulations and real measurements (or emulations). 4) An alternative approach to validation could be to record a trace of all significant events (e.g., all packets sent, received, or forwarded) during the experiment in the real network, and to create a similar trace during a corresponding simulation run. Recent Advances in Telecommunications and Circuits ISBN: 978-960-474-308-7 201