| With the rapid development of wireless communication and internet of things, all kinds of intelligent devices need ubiquitous wireless connection. 5G utilizes high-frequency wireless bands. Meanwhile, diverse wireless communication standards allocated in different frequency bands are anticipated to co-exist for a long time. 80% wireless data accur at indoor scenario, the distributed antenna system is needed. The traditional distributed antenna systems use RF cables and power amplifiers, which is not suitable for future trend. The cable loss and free-space loss is extrem high for high-frequency wireless signal, traditional distributed antenna systems may substantially increase the cost.Therefore, multi-standard multi-band integrated technologies are under intensive research for potential cost and energy saving. Meanwhile, in the context of distributed antenna systems, multi-band analog radio-over-fiber(RoF) systems based on subcarrier multiplexing (SCM) also attract much attention due to their well-known benefits such as low-loss transmission, radio-technology transparency, shared front-haul fiber infrastructure, and good scalability. However, ROF links suffer from inherent nonlinearities. For wireless location applications, interferences from negative spectrum and other frequencies deteriorate the location accuracy. In order to address thess issues,we did the following innovative work:A multi-dimensional digital predistortion (MD-DPD) technique is proposed to improve the performance of multi-band radio-over-fiber links. We investigated the multi-band nonlinear behavior of ROF links based on multi-dimensional memory polynomial model. Based on this model, multi-(?)dimensional digital predistortion is experimentally studied for multi-band ROF links. The experimental results show that the cross-band distortions is dominant with respect to the in-band distortions. Output EVMs are reduced from 3.78%and 4.16% to 1.22% and 1.15% for two bands respectively. Moreover, the results also suggest us to implement a memoryless multi-dimensional digital predistortor with involving nonlinearity orders up to 5 for a practical multi-band RoF system. With the suggested parameters, the proposed multi-dimensional DPD technique is able to improve the signal transmission performance.A multi-dimensional crest factor reduction (MD-CFR) is proposed to improve the performance and efficiency of multi-band radio-over-fiber links.Cooperating with multi-dimensional digital predistortion (MD-DPD), MD-CFR increases the performance of RoF links, in terms of error vector magnitude(EVM) and adjacent channel power ratio (ACPR). More than 5 dB output ACPR reduction is obtained, output EVMs are reduced from 11.83% and 12.47%to 7.51% and 7.26% for two bands respectively, while only a slight improvement to 11.58% and 10.78% is obtained solely using MD-DPD. Given a threshold in EVM or ACPR, the RF power transmit efficiency is also further enhanced.A frequency estimator for real sinusoidal signals using only three DFT samples is proposed to improve the accuracy of wireless location in radio-over-fiber links. Simulations and experiments confirm that our proposed method outperforms traditional alternatives based on three DFT samples. Compared with estimators for complex sinusoidal signals, our method increases the estimation accuracy thanks to its consideration of negative spectrum components. Compared with the estimators with windowing technique, our method also has a better performance in spite of a lower complexity. For multi-frequency scenario, a new estimator based on iteration is also proposed and achieves superior performance. |
PDF Full Text Request