Dual-band and multi-band power amplifiers (PAs) becomes popular as they can effectively reduce the equipment cost and operation cost. On the other hand, the nonlinear interference among signals in different bands needs special attention. Previous modeling results [1] suffer performance degradation when the intermodulation terms among bands fall into bands of the signals. In this paper, we start from the passband Volterra model and derive a general baseband Volterra model considering all possible intermodulation terms. Existing model in [1] is a special case of the proposed baseband Volterra model. Simulation results show that the proposed dual-band Volterra DPD model can provide additional 2 dB performance improvement comparing to the existing dual-band Volterra DPD model.

Recently, visible light communication (VLC) has attracted much attention as a possible candidate technology to meet the ever growing demand in wireless data. However, current low-cost white LED has limited modulation bandwidth, which limits the throughput of the VLC. Optical MIMO can provide spatial diversity and thus achieve high data rate. Traditional multiple-input multiple-output (MIMO) techniques used in wireless communications cannot be directly applied to VLC. This paper studies the precoder and equalizer design of optical wireless MIMO system for VLC. First, we propose a MIMO VLC system, which can effectively support the flickering/dimming control and other VLC-specific requirements. Second, besides the transceiver design with perfect channel state information, we also take into account channel uncertainties for joint optimization in the MIMO VLC system. Numerical results show that the proposed MIMO solution for VLC is robust to combat the influence caused by the channel estimation imperfection. By taking into account the channel estimation errors, the proposed joint optimization method demonstrates the bit error rate (BER) improvements in the scenario of imperfect channel estimation.

In digital predistortion (DPD) implementation, the computational complexity of coefficients estimation of the DPDmodel is a key performance metric. Conventional coefficients estimation algorithms, such as least squares (LS), recursive least squares (RLS), and least mean squares (LMS) cannot achieve a fast convergence with little computation. In this paper, we propose an RLS algorithm with reduced complexity by introducing orthonormal polynomial basis functions. The proposed algorithmis as simple as LMS algorithmyet as efficient as RLS algorithm. Simulation results validate our analysis.

In many predistorter designs, polynomial model was used to describe the nonlinearity of power amplifiers (PAs). However, numerical instability can occur during model coefficients estimation. Orthogonal polynomial models presented in [1] [2] can help to alleviate the numerical instability problem in fixed-point implementation. In this paper, we propose a new orthonormal polynomial model to further improve the numerical stability of the polynomial model. Both simulation and experiment results show that the orthonormal polynomial model can achieve more spectrum regrowth suppression than the conventional polynomial model as well as the orthogonal polynomial model with the same finite word length.