Fig. 1. Structure of (a) TOSA and (b) ROSA. (c) The photographs of the fabricated TOSA module. (d) Transmission of optical signals in the 42° oblique angle De-mux AWG-PLC.

Fig. 2. (a) Simulated results of the optical CE performance versus distance and oblique angle. (b) CE performance versus different distance when the oblique angle of the De-mux AWG-PLC is set at 40°, 42°, and 44°.

Fig. 3. Photographs of the fabricated optical transceiver module, PD array, TIA array, AWG-PLC, and DFB-LD.

Fig. 4. Scattering parameter performance versus the frequency for each lane of the fabricated optical transceiver module: (a) S11 and (b) S21. (c) Measured SFDR3 performance versus different frequency for each lane of the fabricated optical transceiver. (d) SFDR performance of lane 3 when the central frequency is 6 GHz, the fundamental item is the output two-tone RF signal, and IMD3 is the third-order intermodulation distortion.

Fig. 5. Schematic diagram of the modified single-loop LLL algorithm. H: matrix of the MIMO channel. Q, R are the matrix after QR decomposition of H, Q is a unitary matrix, and R is an uptriangular matrix. Im is an identity matrix. μ is the correction factor for R, H, and T. G is an orthogonal rotation matrix.

Fig. 6. Experimental setup of the proposed 4×4 MIMO 16QAM-OFDM RoF system for both (a) downlink and (b) uplink.

Fig. 7. Measured BER performance for 4×4 MIMO 16QAM-OFDM signals over 15.5 km SSMF and 1.2 m air transmission versus received optical power, (a) with ZF, (b) with MLSE, and (c) with LR-ZF in downlink, and (d) with ZF, (e) with MLSE, and (f) with LR-ZF in uplink.

Table 1. Real Multiplication for the Original LLL Algorithm, the Proposed LR Algorithm, and the Optimal MLSE in an $N\times N$ MIMO System