Pekka Kyösti

Survey of channel and radio propagation models for wireless MIMO systems

This paper provides an overview of the state-of-the-art radio propagation and channel models for wireless multiple-input multiple-output (MIMO) systems. We distinguish between physical models and analytical models and discuss popular examples from both model types. Physical models focus on the double-directional propagation mechanisms between the location of transmitter and receiver without taking the antenna configuration into account. Analytical models capture physical wave propagation and antenna configuration simultaneously by describing the impulse response (equivalently, the transfer function) between the antenna arrays at both link ends. We also review some MIMO models that are included in current standardization activities for the purpose of reproducible and comparable MIMO system evaluations. Finally, we describe a couple of key features of channels and radio propagation which are not sufficiently included in current MIMO models.

Novel Radio Channel Models for Evaluation of DVB-H Broadcast Systems

A new area of digital television broadcasting has emerged in the from of hand held reception. This has created a need for new standard channel models, which describe more accurately the conditions in portable reception. Hence, the Finnish partners of the multinational CELTIC Wing-TV project performed a comprehensive measurement campaign during autumn 2005. The analysis of the measurement results show that in most scenarios a strong specular or line-of-sight component is present and clearly dominating. This is noted from the very large values of Rician K-factor. The SFN characteristics of the DVB-H reception is clearly visible while considering the RMS delay spreads, total excess delays and variations of the number of taps. Based on the analysis of the measured data this contribution presents novel tapped delay line (TDL) channel models suitable for DVB-H testing

Complexity Comparison of MIMO Channel Modelling Methods

Computational complexity of simulation of channel models has become an important issue due to the increased requirements for, e.g., bandwidth, number of antennas, quality of service (QoS), number of users, and advanced network topologies. This paper compares the computational complexity of generation of multiple-input multiple-output (MIMO) radio channel realizations with different modelling principles, namely correlation matrix based and sum of rays based methods. Three different aspects are compared. They are complexity of channel coefficient generation, number of required parameters, and the complexity of simulation. The comparison shows that the complexity of the channel coefficient generation is in about the same order of magnitude when the number of MIMO antenna pairs is up to 16 (e.g. 4times4 MIMO). When the number of antennas is higher, the correlation matrix method has higher complexity. The complexity of correlation method depends strongly on the number of antennas while the sum-of-rays method is quite insensitive to it. The simulation complexity dominates the overall complexity and is the same for both methods. The main conclusion of this paper is that the sum-of-rays based method is recommendable especially when the number of antenna pairs is high.

On a MIMO-OTA testing based on multi-probe technology

Standardization work of the OTA test method for MIMO terminals is currently on-going. This paper discusses, mainly from the electromagnetics viewpoint, field synthesis with multi-probe technology with the emphasis on 2-D field synthesis using a circular array. Furthermore, opportunities and practical challenges of multi-probe technology for MIMO-OTA testing are discussed briefly.

Cluster-level parameters at 5.25 GHz indoor-to-outdoor and outdoor-to-indoor MIMO radio channels

This paper describes cluster-level results of indoor-to-outdoor and outdoor-to-indoor double-directional MIMO radio channel measurements carried out at 5.25 GHz. The measurements were performed using a wideband multi-antenna radio channel sounder. Using a set of several thousands of snapshots at the two reciprocal environments, we provide a useful insight into the channel for further channel modelling and channel simulations. We discuss the cluster azimuth spread, cluster delay spread, inter-cluster shadowing, number of clusters per snapshot, number of paths within the cluster, cluster birth rate and cluster lifetime. These parameters are presented for different scenarios by their statistical distributions, and also by their cross-correlation. The results show that both outdoor-to-indoor and indoor-to-outdoor scenarios behave very similar, so we suggest to merge these two cases.

3D Channel Model Emulation in a MIMO OTA Setup

This paper presents a new channel reconstruction technique for 3D geometry-based channels in a multi- probe based MIMO OTA setup. The proposed method provides a general channel reconstruction framework for any spherical power spectrum. The channel reconstruction is formed as convex optimization problems, which give global optimal reconstruction accuracy and allow for a relatively low computational complexity.

The interdependence of cluster parameters in MIMO channel modeling

Novel geometry-based stochastic MIMO channel models like the COST 273 model are using multipath clusters to model the propagation paths of the channel. Recently, we introduced a novel framework to identify and track clusters automatically. MIMO measurements from various scenarios are currently evaluated in several laboratories to parametrize the models. From processing indoor measurements at 2.55 GHz and 5.25 GHz, we found that cluster parameters show significant correlations which have to be considered in the channel models in order to be physically relevant. In detail we observed strong positive correlation between cluster power and the number of paths within a cluster, and between all the cluster spread parameters, while we observe negative correlation between the number of clusters and the cluster spreads.

Evaluation of beam forming and multi antenna techniques in non-stationary propagation scenarios with HW emulator

This paper discusses a three-dimensional (3D) non-stationary channel modelling for the evaluation of multi antenna techniques with a hardware (HW) fading emulator or a software (SW) simulator. The proposed geometric stochastic channel modelling (GSCM) principle supports single or multi link emulations with both measured or simulated antenna patterns and dynamically evolving propagation environments. The modelling principle considers, e.g., 3GPP SCM(E), WINNER II, IMT-Advanced and WINNER+ channel models. The novelty of the paper is in the creation of smoothly evolving propagation and antenna orientation parameters with a linear interpolation of the specified location snapshots. We measured a throughput in four different non-stationary channels but also did a throughput comparison in 2D and 3D urban micro channel models.

Multiple-Screen Diffraction Measurement at 10–18 GHz

This letter presents analysis of diffraction over multiple shadowing screens at frequency range of 10–18 GHz. Up to ten pieces of thin metal sheets with dimensions of 130 × 100 cm at variable spacing were used as diffraction screens. The aim of this study is to investigate the total shadowing effect of multiple knife-edge diffractions at frequencies above the legacy cellular systems. The results show the necessity to adjust the Walfisch–Bertoni path loss model for the higher frequencies in the future fifth-generation systems.

Practical aspects of using IMT-advanced channel models

A geometry-based stochastic channel model was adopted for IMT-advanced evaluations. The complexity of the channel model has been under discussion within ITU-R working groups as well as in industry and academia performing system level simulations of future B3G/4G systems. The complexity reduction is a key task to improve the model usability. However, the complexity reduction should be made without reducing the model accuracy and generality too much. The large scale propagation parameters are random variables with certain distribution. One set of large scale and small scale parameters is called a ldquodroprdquo. To obtain statistical confidence, the number of drops has to be sufficient. Also in each drop, the length of simulation should be long enough. Therefore it is necessary to investigate the required number of drops and impulse responses per drop. Even more important is to study if we need multiple drops at all. This paper presents several approximation steps from full multi-drop IMT-advanced model to simple AWGN. The main message of this paper is to introduce the different approximation steps and their impact on simulation complexity. Simplification of accurate model is always possible, and the methods are known, but doing vice versa is impossible without additional information. Another result is that average channel model is not sufficient, but multiple drops are needed. Additionally, this paper discusses practical aspects on network level simulations such as network layout, user types and penetration losses.