Abstract : The channel reciprocity is an advantageous hypothesis in cognitive radio (CR) technologies, however it is not practically exploitable due to miscellaneous discrepancies including the radio frequency (RF) impairments. Consequently, previous studies proposed calibration solutions to solve these RF impairments. This paper is a step ahead in the implementation of calibration algorithms through the EURE-COM's experimental OpenAirInterface (OAI) platform. It proposes to implement and evaluate a less computationally intensive multiple-input multiple-output (MIMO) calibration algorithm, according to a realistic scenario based on long term evolution-time division duplex (LTE-TDD) specification. The result from OAI simulator shows that it is feasible to restore the reciprocity after calibration in a non reciprocal channel, and provides an overview of the challenges in the channel estimation in a real time case.