| Channel characteristics must be comprehensively considered when designing a wireless communication system. Research on propagation properties of wideband channel is necessary with the increase of bandwidth and capacity of IMT-Advanced systems. Channel measurement is the most straightforward way to obtain channel properties. A variety of channel behaviors can be observed via channel measurement, and the key parameters depicting these behaviors are able to be extracted, which further set up the foundation for channel modeling and simulation work.Based on extensive field measurements, large scale characteristic analysis and channel capacity evaluation are presented in this thesis. The work can be categorized as three parts:First, path loss models for four typical scenarios, i.e. indoor hotspot, indoor multi-floor, outdoor-to-indoor and urban micro-cell, are given based on extensive field measurement. In indoor hotspot scenario, path loss exponent is 1.69 with the intercept of 39.2 for line-of-sight case. The propagation loss is less than that in free space due to the waveguide effect and thus forms a nice propagation condition. For non line-of-sight case, extra propagation loss is induced, which makes the path loss model 20 dB above the free space model. The path loss exponent is 4.33 and the intercept is 25.5. In indoor multi-floor scenario, measurement results show that the largest power attenuation occurs when penetrating the fist floor. Power attenuation per floor decreases as the penetrated floor increases. As for 2.35 GHz, penetration loss caused by penetrating the first floor is 26.3 dB, which far outweigh the loss caused by other floors (2nd floor:7.8 dB,3rd floor:10.8 dB,4th floor:10.0 dB). The same observation is obtain in the measurement at 4.9 GHz. Penetration loss caused by penetrating the first floor is 22.4 dB while the second floor counterparts is 10.2 dB. The attenuation factor model can well fit the path loss in this scenario. The prediction error can be limited to less than 4 dB when the number of floor is large. A classic conclusion is verified by the measurement in outdoor-to-indoor scenario:path loss decrease with the increase of the number of floors where the mobile station is located on. The path loss decreases with the average of 5.3 dB,9.7 dB and 7.0 dB when the mobile station moves from 1st to 2nd floor,2nd to 3rd floor and 3rd to 4th floor, respectively. In urban micro-cell scenario, the path loss exponent and intercept for line-of-sight and non line-of-sight cases are 2.21,45.6 and 3.67,12.4, respectively. Path loss models for indoor hotspot and urban micro-cell scenarios have been accepted by ITU as a part of the channel model for the evaluation of technologies in IMT-Advanced systems.Second, the cross-correlation properties between shadow fading and other channel parameters are investigated, which could be useful for the channel modeling and simulation. It is revealed by the measurement results that a strong negative correlation exists between shadow fading and Ricean K-factor for all cases while the correlation between shadow fading and angle spread of arrival/departure is hardly observed, especially in urban micro-cell scenario. The correlation between shadow fading and delay spread in urban micro-cell environment is weak. For non line-of-sight case in indoor hotspot scenario, they are negative-correlated.Third, path loss for relay-mobile station link, capacity improvement provided by relay transmission and the impact of Ricean K-factor on relay channel capacity are studied based on relay channel measurements. It is observed that the line-of-sight path loss model for base station to mobile station link under estimates the propagation loss at short distance when compared with measurement data. It is 3 dB below the estimated path loss model, as well as the free space model, which is unreasonable in outdoor condition. But the non line-of-sight model provides a better fitting. The path loss model for relay-mobile station link has been accepted by 3GPP and will facilitate the future evaluation of relay techniques. The relay channel capacities under different channel conditions are compared. It is shown that relay transmission leads to capacity loss when the direct link is in good condition while brings capacity improvement when the direct link is shadowed, especially for the outage capacity. Moreover, the decode-and-forward scheme always outperforms amplify-and-forward scheme in terms of capacity if the channel between base station and relay is in good condition. This advantage of decode-and-forward scheme vanishes as the Ricean K-factor of base station to relay link increases. It is proved in the thesis that the maximum mutual information for amplify-and-forward and decode-and-forward schemes will be identically distributed when the Ricean K-factor goes to infinity.From the research on the large scale characteristics presented in the thesis, it can be concluded that:in indoor hotspot scenario, line-of-sight propagation loss can be less than that of free space propagation while the non line-of-sight condition leads to a significant propagation loss; the path loss in indoor multi-floor scenario can be well predicted using the attenuation factor model; it is necessary to introduce the cross-correlation between shadow fading and Ricean K-factor. The cross-correlation between shadow fading and angle spread of arrival/departure can be neglected; the path loss model for base station to mobile station link can not be completely adopted to the relay to mobile station link, especially for the line-of-sight case; make sure the propagation between base station to relay is in good condition, which provides evident capacity improvement when the mobile station is in the shadowing area. The data and observation obtained from channel measurements will be useful in the future simulation and network planning work for the IMT-Advanced system. |
PDF Full Text Request