In full-duplex radio, simultaneous transmission and reception of a signal in the same frequency band is possible only if the transmitting signal interfering with the own receiver is cancelled properly. The full-duplexing technique can double the channel capacity making it a potential candidate for next generation wireless standard. The aim of this research is to investigate effects of self-interference on full-duplex transceivers and propose a method to suppress it in the digital domain for OFDM system. In this method, A copy of the transmitted signal generated by the transmitter is presented to the receiver segment of the transceiver. The receiver subtracts an aligned and scaled version of this signal from its received composite signal to cancel the interference and extract the desired signal from the composite.Moreover, an adaptive algorithm implementation of the recursive least square (RLS) is used for fast and correct estimation of self-interference. A model is designed with the full-duplex transceiver in OFDM system to illustrate research method and is also simulated in MATLAB. The results show that self-interference can be eliminated from desired signal using digital canceller after analog or antenna canceller in series. According to simulation results, RLS is a suitable algorithm for estimation of self-interference based on the rate of convergence and amount of suppression in interference. The main drawback for RLS algorithm is high computational complexity, and longer adaptation time for time-varying channel compared to time non-varying channel. However, adaptation time is short compared to least mean square (LMS) algorithm.