Dahal, Saurav ORCID: 0000-0002-8918-2036 (2020) Millimetre Wave for Fifth Generation of Wireless Communications. PhD thesis, Victoria University.

Abstract

Mobile communication technology is continuously evolving. Early Fifth generation (5G) products, due 2020, should support high capacity, higher data rates, lower la- tency, lower energy consumption and should be cost effective. High data rates will require wider bandwidths, which are available in the higher frequency millimetre wave (mmWave) bands. Millimetre waves have much shorter wavelengths compared to to- day’s microwave mobile systems. Understanding mmWave propagation characteristics is important in the physical layer design of future wireless systems. Statistical Channel models are required in the standardization process to evaluate implementation proposals without the expense of building costly hardware test-beds. Statistical channel models are described by a few key parameters and a variance. For example, transmission path loss is a function of environment, range, path loss exponent (PLE) and the standard deviation (σ) of a statistical (shadowing) component. These parameters are generated by fitting an equation, in a minimum mean square error (MMSE) fashion, to a collated measurement database from a number of research organizations in different countries. Considering this, a number of channel sounding empirical measurements at mmWave frequencies were conducted to characterize, model and evaluate the propagation chan- nel properties in different 3rd Generation Partnership Projects (3GPP) scenarios (envi- ronments) under Australian conditions. Particular outdoor areas of interest were Rural Macro (RMa), Urban Macro (UMa) and Urban Micro (UMi) while indoor areas (InH) of interest included office and shopping malls. An in-house channel sounder was modified to obtain signal strengths with respect to (wrt) location and angle of arrival (AoA). Angular resolution was nominally 10°, 20° and 55° depending upon the size of the horn antenna. The equipment covered frequencies between 24 GHz to 40 GHz and was capable of measuring path losses up to 160 dB with the appropriate antenna. The experimental work involved performing measurements in different scenarios and then comparing to existing channel models (if such exist) and suggesting improvements where appropriate. The thesis focuses on outdoor and out- door to indoor mmWave wireless propagation channels and contains contributions in each of the key scenarios as follows: Rural Macro (RMa) channel measurements were performed at 24 GHz. Average path loss and Angle of Arrival (AOA) spread were close to 3GPP predictions limited to sub 7 GHz frequencies. The results were submitted to the Third Generation Partnership Project (3GPP) enabling the extension of the existing RMa model to 30 GHz. Urban Macro (UMa) have base station (BS) antennas above roof-top height. Measure- ments at 27.1 GHz were performed in two environments; dense urban and light urban, both classified as NLOS UMa. The results were compared against two competing 3GPP path loss models; the alpha beta (AB) model and the close-in free space reference dis- tance (CI) model. The AB model had the lower RMS error. The AB model required input parameters such as average building height and street width etc. which made it sensitive to the choice of input parameters, whereas the CI model required no input pa- rameters and was preferred if the cell consists of wide variety of dense and light urban regions. Urban Micro (UMi) have BS antennas below roof-top height. Measurements were con- ducted at 39.5 GHz in Open Square (OS) and Street Canyon (SC) scenarios. Results showed that the extracted parameters are in close agreement with 3GPP specifications, particularly for the CI model. Additionally, we reported 31 dB for 100 m down the SC side streets out of which 18 dB loss occurs just around the corner. We further analysed the base station (BS) height gain effect in OS and found a marginal benefit of 0.5 dB for a 4 m height change. Cross polar discrimination were also reported in SC which reduced by 2.5 dB per 10 dB increase in path loss using directional antennas while correlation disappeared on a omnidirectional basis. We presented a double-directional measurements for an UMi OS environment. Results showed that multiple Angles of Departure from a given user equipment (UE) position often result in few (often one) Angles of Arrival at the BS. Similarly different UE lo- cations can often share a common angle of arrival at the BS. This could cause rank reduction in some MIMO system. An outdoor to indoor measurement campaign at 24 GHz emulating the satellite to in- door propagation channel was presented. The results are applicable to the satellite/mo- bile co-existence problem as well as in-building coverage from high altitude platform. The mean building entry loss increased by 0.43 dB per degree of (satellite) slant eleva- tion angle, almost twice the ITU recommendation. Further we showed that the signal linearly decays with distance with a slope that increases with slant angle. Additionally, we showed that high gain narrow beam antenna outperformed low gain wide beam an- tenna both in terms of signal maximization to high altitude platform as well as signal minimization to a co-existing satellite uplink channel. Further, there is an antenna gain reduction in this type of environment due to internally generated multipath.

Search Google Scholar

Repository staff login