In this study, a software-defined radio based on orthogonal frequency division multiplexing (OFDM) is implemented to optimize universal software radio peripheral (USRP) for the optimal bit error rate. Two USRP B200s are used as the hardware platform, and an OFDM signal of 802.11a WiFi standard is used. The wireless system works with 64-QAM modulation and operates at the carrier frequency of 2.61 GHz to avoid interference. An OFDM signal generally has a large peak to average power ratio (PAPR) such that the transceiver is optimized to obtain the linearity requirement. To optimize the linearity of the USRP, a 14 dB back-off is set, and the base band signal of 20 MHz is up-sampled to 40 MHz. In addition, a digital filter is added to eliminate the unwanted signals. To analyze the performance of the implemented system, the bit error rate to signal-to-noise ratio is measured, and a maximum bit error rate of 1.75×10−6 is obtained without error correction. Furthermore, the standard channel codes are implemented to correct the bit errors occurring in a wireless channel. The image data are successfully transmitted and received without bit errors by using the channel codes.