oa Baseband DSP for dirty RF front-ends: Theory, algorithms, and test bed

Orthogonal Frequency Division Multiplexing (OFDM) is widely adopted as the transmission scheme of choice for almost all broadband wireless standards (including WLAN, WiMAX, LTE, DVB, etc.) due to its multipath resilience and practical implementation complexity. The direct-conversion radio frequency (RF) front-end architecture , where the down-conversion to baseband is accomplished in a single stage, requires fewer analog components compared to the super-heterodyne architecture where the down-conversion is accomplished with one or more intermediate frequency (IF) stages. Fewer analog components result in reduced power consumption and cost. However, direct-conversion OFDM based broadband wireless transceivers suffer from several performance-limiting RF/analog impairments including I/Q imbalance and phase noise (PHN) [1]. I/Q imbalance refers to the amplitude and phase mismatches between the in-phase (I) and quadrature (Q) branches at the transmit and receive sides. In an OFDM system with I/Q imbalance, the transmitted signal at a particular subcarrier is corrupted by interference from the image subcarrier. To compensate for the effect of I/Q imbalance, the received signals from each subcarrier and its image subcarrier are processed jointly. PHN refers to the random unknown phase difference between the carrier signal and the local oscillator. In an OFDM transceiver, PHN rotates the signal constellation on each sub-carrier and causes inter-carrier interference between the sub-carriers resulting in significant performance degradation. In this project, we designed novel algorithms for the efficient estimation and compensation of RF impairments including I/Q imbalance and PHN for beamforming OFDM systems such as 4G LTE cellular systems and 802.11 wireless local area networks (WLAN) [2]. Conventional OFDM transceivers ignore I/Q imbalance effects and each OFDM subcarrier is processed separately which causes an irreducible error floor in the bit error rate performance. However, in our proposed method, each OFDM subcarrier and its image subcarrier are processed jointly to mitigate the effects of I/Q imbalance. This novel method is capable of eliminating the error floor and obtaining performance close to the ideal case where no I/Q imbalance exists. Furthermore, we have developed an experimental OFDM testbed to implement the proposed algorithms. Our testbed uses the Universal Software Radio Peripheral (USRP) N210 RF frontends and it is based on packet-based OFDM similar to the IEEE 802.11.a WLAN standard. The baseband processing is done in MATLAB where the MATLAB driver for USRP is used for stream processing of the transmitted and received signals. The measured experimental results demonstrate that the proposed algorithms improve the performance significantly at low implementation complexity. [1] B. Razavi, RF Microelectronics. Englewood Cliff, NJ: Prentice-Hall, 1998. [2] O. Ozdemir, R. Hamila, and N. Al-Dhahir, “I/Q imbalance in multiple beamforming OFDM transceivers: SINR analysis and digital baseband compensation,” IEEE Trans. on Comunications, vol. 61, no. 5, pp. 1914-1925, May 2013. Acknowledgement: This work is supported by Qatar National Research Fund (QNRF), Grant NPRP 09-062-2-035.