Magne Pettersen

Automatic antenna tilt control for capacity enhancement in UMTS FDD

Simulations were performed to evaluate the effect of adaptively adjusting the base station antenna vertical tilt angle to adapt to the current geographical traffic load distribution and thereby achieve an overall capacity increase in UMTS FDD. The capacity increase is very dependant on the traffic conditions. Using a generic 21-cell scenario and positioning three traffic hotspots within the area with varying position relative to the base stations, the capacity increase was in the order of 30% on average at a 99% grade-of-service. The capacity improvement by using tilt adjustment increases with increasingly unbalanced traffic, as expected. Automatic tilt control represents a viable alternative for interference reduction purposes and thereby provisioning of better overall network capacity during dynamically changing traffic situations.

Characterisation of the directional wideband radio channel in urban and suburban areas

Directional wideband channel measurements in urban and suburban areas have been performed using a channel sounder and an eight element linear array. A statistical analysis of the results has been performed. It was shown that in urban area a base station antenna at lamppost level lead to more severe azimuth spread (AS) than a base station at rooftop level. Using a phased array smart antenna reduced the azimuth spread by more than 70% in both urban and suburban areas. The smart antenna reduced the delay spread (DS) by 45% in suburban area, but only by 10% in urban area. The correlation between AS and DS was low in most of the measurement routes.

An evaluation of adaptive antennas for UMTS FDD by system simulations

A system simulator for performance evaluation of adaptive antennas in UMTS FDD has been developed. The simulator is flexible, it uses realistic channel and system assumptions, and, for a given scenario, provides the maximum number of users that can be supported by the network. The antenna types included in the simulator are switched lobe, phased array and adaptive array, as well as conventional fixed sector antennas. Results point to a capacity increase of more than a factor of two by using adaptive antennas compared to conventional sector antennas. Most of the increase is obtained by using the relatively simple switched lobe antenna implementation. Using adaptive antennas can alternatively increase range or reduce transmit power. The reduction in transmit power points to MS battery lifetime being increased by at least 4 times compared to conventional sector antennas. Sector and switched lobe antennas are approximately equally sensitive to downlink orthogonality loss. The total network capacity was roughly 20 % higher when increasing the downlink code orthogonality from 0.4 to 0.8.

DECT offering mobility on the local level. Experiences from a field trial in a multipath environment

The digital enhanced cordless telecommunications (DECT) is a candidate for radio in the local loop applications. Telenor has established a DECT field trial. DECT most often gives satisfactory quality of service (QoS), but in some areas the system shows poor performance despite adequate signal strengths. To investigate the influence of multipath propagation on DECT performance, we have conducted combined wideband propagation measurements and DECT performance measurements in areas where poor DECT QoS has been experienced. Some simple means to cope with multipath propagation would considerably improve DECT performance in an outdoor environment.

Space diversity measurements for DECT in indoor and outdoor microcells

Diversity measurements have been performed in microcellular environments where DECT typically might be implemented. The intention has been to investigate how well different implementations of antenna space diversity can combat the problems DECT experiences in multipath environments. Diversity with 2 antennas and selection/switching based on signal level and on delay spread has been investigated, and the effect on the delay spread statistics has been studied. Measurements have been performed in three different types of scenarios; both radio base station (RBS) and portable part (PP) indoors (indoor case), RBS outdoors and PP indoors (outdoor to indoor case) and both RBS and PP outdoors (outdoor case). For the indoor case the delay spread values were small. The most severe time dispersion was observed in the outdoor to indoor case, where delay spread values exceeded 100 ns in almost 50% of the cases. In the outdoor scenario delay spread values above 100 ns were observed for approximately 10% of the cases. It seemed that for both the outdoor, and outdoor to indoor cases, where the time dispersion may cause problems for DECT performance, the diversity algorithm should be based on some other selection/switching criterion than signal level, preferably BER or some other quality criterion. In the indoor environment, signal level as a selection/switching criterion should be sufficient. The work presented in this paper is part of an investigation of DECT performance in environments which DECT has not been originally designed for, to possibly extend the usage of DECT for RLL applications and as a public access system.

An evaluation of adaptive antennas for a UMTS FDD network deployed in a suburban area of Norway

A system simulator for performance evaluation of adaptive antennas in UMTS FDD networks has been developed. The simulator is flexible, it uses realistic channel and system assumptions, and for a given scenario provides the maximum network capacity. The antenna types included in the simulator are switched lobe, phased array and adaptive array as well as conventional fixed sector antenna. Analyses of a UMTS network deployment in a suburban area of Oslo, Norway, reveal that adaptive antenna techniques potentially can more than double the radio capacity compared to conventional sector antennas. The majority of the gained capacity is obtained by using a relatively simple switched lobe antenna implementation, providing approximately 75% of the capacity compared to the adaptive arrays. The phased array alternative offered a slight improvement over switched lobe, with average throughputs per base station of 2.4 Mbit/s and 2.3 Mbit/s respectively. These results demonstrate that beam forming techniques can provide significant radio capacity gains for UMTS FDD.

Air interface capacity in future heterogeneous mobile networks

This paper gives an overview over expected applications for the Universal Mobile Telecommunications System (UMTS). Based on current traffic statistics and the expected penetration of new applications, a traffic prediction is performed. This traffic prediction, together with a capacity and performance analysis of the future air interface, forms the input for network planning. The majority of existing telecommunication operators have GSM networks, which are expected to be upgraded with the General Packet Radio Service (GPRS) in 1999/2000. With the introduction of UMTS in 2002, most operators will operate a heterogeneous network, where traffic will determine the load of the GSM/GPRS and UMTS network. The paper performs a capacity analysis for 144 and 384 kbit/s services, and provides guidelines for network planning.

Applicability of rough surface scattering to 3D radio planning

A demonstration 3D channel prediction model was used in example scenarios in hilly terrain and undulating rural terrain. The model is comprised of a 2D vertical Tx-Rx-plane component as well as 3D components due to off-axis scattering. The latter components are found using rough surface scattering estimation. In the implementation a novel high performance scattering model called the amplitude/phase model was used. It was shown than when the vertical plane 2D prediction was attenuated more than typically 15 dB compared to free space a 3D channel model is necessary. This value was exceeded in approximately 40% of the coverage area in the hilly terrain, and in about 15% of the area in undulating rural. This leads to the conclusion that in a hilly terrain traditional 2D models are not sufficient. When the 3D components were dominating the distribution was approximately Rayleigh, with a typical difference between the 5 and 95 percentiles in the distribution of more than 15 dB. Weather and seasonal variation may change the mean received signal level by up to 4-5 dB. An erroneous assumption about the permittivity or surface roughness by a factor of two could change the mean of the received radio signal by up to 10 dB.

An experimental performance comparison between antenna diversity schemes at 1800 MHz

A measurement campaign to compare four different base station two-branch antenna diversity schemes in the 1800 MHz band has been carried out. The measurements were performed in such a way that the performance of the diversity schemes could be compared directly for different environments. The main conclusion to be drawn from this measurement campaign is that polarisation diversity schemes performed better than space diversity schemes.