The aim of this article is to propose and compare linear and nonlinear precoding schemes for multicell multiuser MIMO-OFDM based systems. The considered linear precoder is designed in two phases: first the intercell interference is removed by applying a linear zero-forcing algorithm. Then the system is further optimized by proposing three power allocation algorithms with per base station power constraint and different complexity tradeoffs: one optimal to minimize the average bit-error-rate and two suboptimal. The proposed nonlinear precoding is designed to minimize average bit-error-rate, over the users, conditioned to a channel realization and the transmitted data. In the high SNR regime, this problem simplifies from a constrained quadratic nonlinear optimization to a single quadratic problem with a scaling, allowing to reduce the complexity. The performance of the proposed schemes is evaluated, considering typical pedestrian scenarios based on LTE specifications. Numerical results show that the performance of the nonlinear scheme outperforms the linear ones. The nonlinear algorithm selects and inverts part of the correlation matrix unlike the linear zero-forcing where full inversion is required. This leads to a better performance as the selection allows to get a better conditioned matrix. Also, it is shown that the complexity of the nonlinear scheme is similar to the linear suboptimal closed-form one.