Luis C. Vieira

Predistortion of Radio Over Fiber Links: Algorithms, Implementation, and Measurements

Distributed antenna systems (DASs) have been proposed for use in high-data-rate wireless communications. Connecting the central unit to remote antenna units, radio over fiber (RoF) links become essential parts of the DAS. One of the major problems in the RoF links is the nonlinear distortion experienced in its optoelectronic devices. This paper considers the adaptive compensation of the nonlinear RoF link. In particular, we present an extension to the conventional least-mean-square algorithm for memory-polynomial-based predistortion. In addition, we study different combinations of the algorithms for the predistortion. Moreover, we implement the adaptive predistorter in hardware, build a real RoF link, and verify the performance by measurements. The results demonstrate that distortion of the RoF link can be considerably diminished with a low complexity design.

Analysis of and compensation for non-ideal RoF links in DAS [Coordinated and Distributed MIMO]

Distributed antenna systems have been found to be an elegant solution for the problems arising in high-data-rate wireless communication, particularly in large service areas. This article considers radio over fiber links as an essential part of the DAS, connecting the central unit with the remote antenna units. In particular, we analyze and discuss delays and nonlinearities stemming from the RoF links. In addition, we study the compensation for these impairments. Our studies indicate that the RoF links are a viable and cost-effective solution for implementing the DAS, although some of the RoF link non-idealities require compensation.

Behavioral modeling of radio-over-fiber links using memory polynomials

A behavioral, nonlinear radio-over-fiber (RoF) link model considering memory effects with low complexity is presented. Model validation against experimental results is demonstrated, with the effect of the memory length on the fitting accuracy being studied.

Architectures for Joint Compensation of RoF and PA with Nonideal Feedback

A high capacity wireless communication system requires careful design to minimize interference and distortion effects. This paper considers the adaptive predistortion of the nonlinear distortions induced by a Radio over Fiber (RoF) link and power amplifier (PA) connected in series. In particular, we study the architectures for the joint compensation of the nonlinearities in the presence of nonideal feedback. From the adaptive algorithm point of view, we study different combinations of algorithms for the predistortion. Our simulation results indicate that combined use of least mean squares (LMS) and recursive least squares (RLS) gives the best trade-off between complexity and performance. In addition, we show that the use of nonideal feedback causes a collapse in the performance of the predistortion. However, when using a compensated RoF link for feedback, the degradation of the adjacent channel power is very small compared to the case of the ideal feedback.

Fiber link design considerations for cloud-Radio Access Networks

Analog radio over fiber (RoF) links may offer advantages for cloud-Radio Access Networks in terms of component cost, but the behavior of the distortion with large numbers of subcarriers needs to be understood. In this paper, this is presented in terms of the variation between subcarriers. Memory polynomial predistortion is also shown to compensate for RoF and wireless path distortion. Whether for digitized or analog links, it is shown that appropriate framing structure parameters must be used to assure performance, especially of time-division duplex systems.

An experimental study on digital predistortion for radio-over-fiber links

Digital predistortion for directly modulated RoF links is experimentally investigated. A memory-polynomial-based predistorter model is directly identified from measurements of the baseband OFDM input-output signals and using the least-squares method in Matlab. The RoF link nonlinearity has been reduced implementing this technique.

Baseband behavioral modeling of OFDM-Radio over fiber link distortion

A memory polynomial (MP) model is compared against a modified envelope-memory polynomial (MEMP) model and a memoryless polynomial for characterization of OFDM-Radio over fiber link distortion, using a baseband behavioral modeling approach. The model accuracies are studied for 1.85-GHz RoF links of 2 and 20 km. An 18-MHz bandwidth 16-QAM/OFDM signal is used for the experiments, with 256-point IFFT size. For most cases, the three models can all reproduce very well the RoF link spectral regrowth. The best accuracies are found for the MP model, especially for the 20-km link, demonstrating the robustness of this model for representing RoF links with high distortion.

Study of complex-envelope behavioral models for radio-over-fiber link nonlinearities

Memory polynomials (MP) are demonstrated for modeling of RoF link nonlinearities, within the frequency range from 0.5 to 2 GHz, using a complex-envelope behavioral modeling approach. The model accuracies are studied for different RF frequency and input power cases, with the dependency of the MP model accuracy on the nonlinearity order and memory length parameters discussed. The experimental results show that the RoF link presents more memory effects at some RF frequencies and that the link phase nonlinearity tends to increase with the operating frequency. Excellent agreement between the measured and modeled results is reported for several validated MP models. In comparison with a memoryless polynomial model, the MP model performs significantly better for all experimental cases.

Clipping and predistortion for compensation of OFDM-Radio over Fiber link distortion

A combined clipping/predistortion scheme is proposed and demonstrated for the distortion compensation of OFDM-Radio over Fiber links at baseband. In this scheme, a memory-polynomial-based predistorter and a simple amplitude clipping method are used. No specific filter is used after the clipping function, which reduces the implementation complexity. Experimental results confirm that the proposed approach gives significantly better performance compared to the digital predistortion (without clipping) approach for highly distorted OFDM RoF links. Changes in the OFDM signals PAPR through the RoF link are analysed for the predistortion schemes with and without clipping. As a result of this, a way of optimizing the performance of the clipping/predistortion approach is proposed.