An Adaptive Error Correction Scheme For Synchronous Digital Hierarchy-Based Wavelength Division Multiplexed Optical Network

In optical communications there are a variety of noise and distortion sources which can cause errors. These errors become essential and more intense in the high-capacity and long-haul wavelength-division multiplexing (WDM) systems. Therefore, the development of a forward error correction (FEC) technique to mitigate errors in WDM optical networks is very relevant and important. The existing FEC techniques for optical communications are based on fixed codes, which consume unnecessary overhead bandwidth even when there are no errors. This thesis proposes an adaptive forward error correction (AFEC) scheme for synchronous digital hierarchy (SDH)-based WDM optical networks, referred to as the SDH-AFEC. The scheme supports adaptive codes because it uses a dedicated WDM channel for transmission of different sizes of FEC redundancy for the payloads. Unlike most previous adaptive FEC techniques which change to a stronger code after an error has occurred, the SDH-AFEC is able to do so before an error occurs. This is achieved by using the combination of B2 error and corrected error count as the input parameters for the algorithm. Then the algorithm is designed in such a way that it adaptively assigns a suitable value of error correction capability, t for error correction, and the number of corrected errors is maintained not exceeding t/2. The SDH-AFEC adopts Bose–Chaudhuri–Hocquenghem (BCH) and Reed–Solomon (RS) codes for correcting random and burst errors respectively. A new technique is also proposed for estimation of the error pattern so that a suitable type of code can be assigned accordingly. This technique is based on the analysis of the corrected error locations, referred to as the error location analysis (ELA). Simulation results show that the SDH-AFEC is able to use different values of t adaptively for error correction. It assigns stronger t with increasing channel bit error rate (BER) or average burst length (ABL) to maintain the output BER below the target BER of 10-9, until the strongest value of t is assigned. The SDH-AFEC uses the maximum FEC overhead for high BER or long ABL. However, the FEC overhead requirement reduces with decreasing BER or ABL. Hence, in addition to the adaptive BER performance, the SDH-AFEC also provides a way to use the FEC overhead efficiently. Lastly, the results also show that by using ELA, the performance of the SDH-AFEC is further improved that it is able to correct about three times more random errors and three times longer burst length. Meanwhile, the average FEC overhead reduction after ELA is about 38% and 36% for random and burst errors respectively.

Preview PDF FK_2007_79.pdf Download (233kB)

Actions (login required)

View Item