Carsten Jandura

Multisite field trial for LTE and advanced concepts

The 3GPP LTE standard is stable now in its first release (Release 8), and the question is how good its performance is in real-world scenarios. LTE is also a good base for further innovations, but it must be proven that they offer performance advantages for the price of their complexity. This article evaluates the performance of LTE Release 8 as a baseline and advanced concepts currently in discussion such as cooperative MIMO based on system-level simulations, and measurements in the laboratory and a multisite field testbed within the EASY-C project.

Intercell Interference Measured in Urban Areas

We report on measurements at 2.53 GHz with 20 MHz bandwidth in two triple-sectored urban macro-cell deployments in Berlin and Dresden, Germany. These measurements assist the design of multiuser and cooperative transmission and detection techniques promising to reduce the interference in future cellular systems. We study the geometry factor and the Top-N power statistics in the non-cooperative case and compare the results with a commonly used spatial channel model. Furthermore we observe that the median delay parameters in a coherent single frequency network deployment are roughly 1.5 times higher than in an isolated cell while the maximum excess delays are similar in both cases. Finally we study the correlation of the channel response for different antenna spacings at the base station. When the sector antennas are close together, the correlation is generally high which is also predicted by the channel model, but with spacings of about 20 m the correlation significantly decreases.

Verifying ray tracing based CoMP-MIMO predictions with channel sounding measurements

Multiple-Input Multiple-Output (MIMO) technology is ready for deployment in commercial cellular networks in the very near future. Thus, the need of incorporating this technology into radio network planning and optimization rises dramatically for network operators. The main question to answer is how accurate MIMO channel models reflect the real MIMO channel. In this contribution we verify a detailed ray tracing channel simulator with channel sounding measurements in the 2.53 GHz range by comparing simulated and measured eigenvalue characteristics for various Single-User (SU) downlink scenarios in a Coordinated Multi-Point (CoMP) environment. From our comparison we can conclude that carefully performed Geometrical Optics based ray tracing simulations are an adequate prediction model to reflect main characteristics of SU-MIMO channels even in a CoMP scenario.

Comparison of time- and angular dispersion between channel sounding measurements and ray tracing in CoMP-MIMO channels

The need of incorporating Multiple-Input Multiple-Output (MIMO) technology into radio network planning and optimization rises dramatically for network operators. One main question to answer is how accurate MIMO channel models reflect the real MIMO channel. To answer, we already compared eigenvalue statistics of various single-user (SU) and multi-user (MU) Cooperative Multi-Point (CoMP)-MIMO scenarios [1, 2, 3, 4] to verify a 3D ray tracing channel simulator with channel sounding measurements in the 2.53 GHz range, a major frequency band for LTE transmission in Europe. This contribution adds a comprehensive verification of time- and angular dispersion characteristics by carefully analyzing results of a dedicated high-resolution measurement campaign. These extra measurement data allow for an intermediate step of high-resolution post-processing of the measurement data [5, 6] to estimate radio channel parameters. These estimated parameters are then directly compared with ray tracing channel simulation results, especially in the rarely analyzed angular domain.

On timing constraints and OFDM parameter design for cooperating basestations

It is well known that symbol timing offsets larger than the cyclic prefix between transmitter and receiver stations destroy the orthogonality among OFDM subcarriers inducing interblock interference by coupling subsequent OFDM symbols. In conjunction with MIMO transmission over frequency selective channels, this effect on top of multi-user interference strongly degrades performance. In this paper we analyze the timing constraints in cellular MIMO OFDM networks that necessarily occur due to time-of-arrival differences of signals from different locations. We investigate a simple hexagonal model, where three users transmit their data on the same time-frequency resource to three cooperating base stations. Based on a detailed analytical derivation of the transmission model we present results for the post detection SINR in systems with symmetric and uniform user distribution within one cooperative cell.

Analysis of MIMO channel measurements in urban areas

Multiple-Input Multiple-Output (MIMO) technology is ready for deployment in commercial cellular networks in the very near future. Therefore, the need of incorporating this technology into radio network planning and optimization tools rises dramatically for network operators. In this contribution we show results of the validation of advanced ray tracing simulations with channel sounding measurements in the 2.53 GHz frequency band. These measurements were carried out inside the EASY-C project with the objective to learn more about coordinated multipoint (CoMP) [1] transmission in a real world testbed. We compare simulated and measured eigenvalue characteristics for various MIMO scenarios. As result we can conclude that carefully performed Geometrical Optics based ray-tracing simulations are a suitable prediction model to reflect main characteristics of these scenarios. A further objective of these experiments is to get insight into the physical structure of the radio channel [2].

Analysis of Measured and Simulated Delay and Angular Propagation Channel Parameters for MIMO Network Planning

A verification of delay and angular parameters of a deterministic ray tracing MIMO channel model requires dedicated high-resolution wide-band directional propagation channel measurements to estimate according channel parameters and is rarely done. This contribution presents such an analysis in the 2.53 GHz range through simulations and measurements in a real MIMO urban outdoor cellular environment. Our results indicate a good fit between measurement and simulation data for the analyzed delay and receiver angular channel parameters in macro-cellular deployments and especially in LOS environments. We notice, however, some deviations in case the transmit antennas are at rooftop level (micro-cellular deployment). This result suggests to use more reliable eigenvalue ratios for spatial multiplexing probability analysis instead of angular spreads in micro-cellular and heterogeneous MIMO network planning and optimization.