Showing papers by "Preben Mogensen published in 1997"

Power azimuth spectrum in outdoor environments

Abstract: From extensive 2D channel measurements it is found that the shape of the power azimuth spectrum at the base station in outdoor environments is accurately described by a truncated Laplacian function in both rural and urban environments.

Preliminary measurement results from an adaptive antenna array testbed for GSM/UMTS

Abstract: This paper presents preliminary measurement results for an 8 element adaptive antenna array in an urban environment. The results show that in most cases the received signal consists of a single narrowly scattered component impinging from the direction towards the mobile. The angular spread (AS) is found to be no larger than 5 degrees for the test measurements. The correlation coefficient between slow fading and AS is -0.56, indicating that the AS increases when the local signal strength fades. The paper also shows the results for a direction of arrival (DoA) algorithm, which is designed for downlink beamsteering in a GSM related system. The results indicate that an averaging window size of one SACCH frame (0.48 s) is a wise choice for DoA estimation in a non-FH network.

High resolution of electromagnetic waves in time-varying radio channels

Abstract: The applicability of the SAGE (space-alternating generalized expectation-maximization) algorithm to the estimation of time variant radio channels is demonstrated. This algorithm allows one to separate the complex multi-dimensional optimization problem required to compute the estimate of the parameters characterizing the impinging waves, i.e. their delay, incidence azimuth, Doppler frequency, and complex amplitude, into separate one dimensional optimization processes that can be performed sequentially. The performance of the scheme, i.e. its convergence behavior and its accuracy, is investigated in synthetic and real channels.

On the capacity of a GSM frequency hopping network with intelligent underlay-overlay

Abstract: This paper describes a method to increase the capacity of a digital cellular network, like GSM, without having to increase the number of available frequencies or base stations. The idea is to split the frequency spectrum into two bands. One consisting of frequencies that can only be used when a high C/I ratio is ensured, the super frequencies, and the other consisting of frequencies that can be used throughout the whole cell, the regular frequencies. The minimum frequency reuse distance of the super frequencies can therefore be smaller than the reuse distance of the regular layer frequencies. By combining this reuse partitioning with frequency hopping, an increase in the network capacity in terms of carried traffic per cell is achieved. Simulations have indicated that for slow moving mobiles a gain of approximately 35% is achieved by this new feature when compared with a frequency hopping network. For faster moving mobiles the simulated gain is lower, approximately 26%. Modifications are proposed to optimise the simulated algorithms, to improve the capacity in terms of carried traffic per cell even further.

Improved intelligent underlay-overlay combined with frequency hopping in GSM

Abstract: IUO (intelligent underlay-overlay) in a combination with random frequency hopping in GSM is analysed. Several improvements to the original IUO concept analysed in Nielsen et al. (1997) are introduced. With the improved IUO concept it is possible to load a network configuration consisting of 4 regular frequency channels and 3 super layer channels almost to the hard blocking limit. A capacity gain of more than 40 percent compared to a one layer frequency hopping 3/9 reuse network has been found.

Algorithm test using measurements from an adaptive antenna array testbed for GSM/UMTS

Abstract: This paper shows the results for a direction of arrival (DOA) algorithm, which is designed for downlink beamsteering in a GSM related system. Measurement campaigns have been conducted for an 8 element adaptive antenna array in an urban environment. The results using measurement data indicate that an averaging window size of one SACCH frame (0.48s) is a wise choice for DOA estimation in a non-FH network or a FH network using DTX. Downlink beamsteering performance is analysed for different antenna array configurations and different number of fixed beam directions. The performance evaluations are made for both the measured data and data obtained from simulations.

A preamble based diversity scheme for DECT using a new signal quality measure

Abstract: PCS and RLL (radio in the local loop) are applications which set new demands on the DECT receiver, due to the fact that the RMS delay spread often exceeds 200 ns for outdoor environments. For comparison a typical low-cost DECT receiver based on a limiter-discriminator detector is not able to cope with more than 100 ns delay spread. A well-known solution which increases the receiver resistance to time dispersion is antenna diversity. The optimum realizable selection diversity scheme is based on the status of the R-CRC (cyclic redundancy check) check and it requires two parallel receiver chains. The simpler switching diversity scheme, normally used in commercial equipment, requires only one receiver chain, but it suffers from a strong velocity dependency. In this paper a diversity scheme using a new signal quality measure is presented. The signal quality is extracted successively in the two antenna branches during the DECT preamble which makes the scheme velocity independent as no significant delay is introduced. With the new quality measure it is possible to closely emulate genuine R-CRC selection diversity by use of only one receiver chain. For an E/sub b//N/sub 0/ of 25 dB, the maximum allowable RMS delay spread is increased from 180 ns for RSSI driven diversity to 220 ns with the new scheme. This is only 10 ns less than R-CRC selection diversity. For realistic hardware conditions, the performance is only degraded 10 ns which makes the new diversity scheme applicable for a practical implementation.

magnetic o Chan~~ls

Abstract: Absfracr-The applicability of the SAGE (Space-Alternating Generalized Expectation-Maximization) algorithm to the estimation of timevariant radio channels is demonstrated. This algorithm allows to separate the complex multi-dimensional optimiition problem required to compute the estimate of the parameters characterizing the impinging waves, Le. their delay, incidence azimuth, Doppler frequency, and complex amplitude, into separate one dimensional optimization processes that can be performed sequentially. The performance of the scheme, i.e. its convergence behavior and its accuracy, is investigated in synthetic and real channels. Design and analysis of advanced communications systems equipped with an antenna array require realistic models of the radio channel. Indeed, system performance evaluations by means of either theoretical investigations or Monte-Carlo simulations yield representative results only provided that the underlying channel models accurately reflect the relevant features of the physical channel. Extensive measurements are a prerequisite for the development of accurate channel models. Moreover, appropriate sophisticated processing tools need to be employed in order to extract the parameters of interest from the measurement data. Different methods have been proposed nowadays to estimate the incidence direction of impinging waves. The Beam-Forming and Fourier algorithms [ 11 are presently the most well-known and probably the widest used among these techniques. Recently, high resolution schemes have been successfully applied to resolve the waves according to their azimuth andor delay. They rely either on nonparametric or spectral estimation approaches like the MUSIC [2], [3] and ESPRIT [4], [5] algorithms or on parametric estimation approaches, such as maximum-likelihood-()-based methods [6], [7]. The scheme presented in the last reference jointly resolves the impinging waves in a time-invariant propagation environment according to their delay and incidence azimuth. It relies on the space-altemating generalized expectation-maximization (SAGE) algorithm [SI, which is an extension of the expectation-maximization (EM) algorithm [9]. The former method allows to replace the highdimensional (D) non-linear optimization problem necessary to compute the ML, estimate ofthe waves’ parameters by several l-D maximization procedures that can be performed sequenIn this paper the scheme proposed in [7] is extended to include the estimation of the Doppler frequency of the waves in time-varying environments. This additional dimension allows to resolve waves having close incidence azimuths and relative delays. Such a situation occurs quiet frequently at the base station (BS) in macro-cells. In this case the power received * B. FYeq is with Communication Technology Laboratory of the Swiss