Statistical Performance Comparisons of Optical OFDM Adaptive Loading Algorithms in Multimode Fiber-Based Transmission Systems

Abstract

Detailed statistical investigations of the effectiveness of three widely adopted optical orthogonal frequency division multiplexing (OOFDM) adaptive loading algorithms, including power loading (PL), bit loading, and bit-and-power loading (BPL), are undertaken, for the first time, over 1000 statistically constructed worst-case multimode fiber (MMF) links without incorporating inline optical amplification. It is shown that the BPL (PL) algorithm always offers the best (worst) transmission performance. The absolute transmission capacity differences between these algorithms are independent of signal bit rate and increase with both transmission distance and digital-to-analog converter/analog-to-digital converter (DAC/ADC) sampling rate. More importantly, numerical results also indicate that, for both worst-case and normal-case MMF links of less than 300 m, in comparison with the most sophisticated BPL algorithm, the simplest PL algorithm is sufficiently effective in escalating the OOFDM MMF system performance to its maximum potential. The effectiveness of the PL algorithm can be further improved when a large number of subcarriers are utilized. On the other hand, for relatively long MMF links (> 800 m) with their 3-dB bandwidths being much less than the transmitted OOFDM signal spectrum, the BPL algorithm has to be adopted. The aforementioned results have great potential for practical cost-effective OOFDM transceiver architecture design.

Keywords

• Optical fiber communication
• orthogonal frequency division multiplexing (OFDM)
• multimode fiber (MMF)
• adaptive loading algorithm