Threshold Detection Scheme Based on Parametric Distribution Fitting for Optical Fiber Channels

Background: When data is transmitted by encoding it in the amplitude of the transmitted signal, such as in Amplitude Shift Keying, the receiver makes a decision on the transmitted data by observing the amplitude of the received signal. Objective: Depending on the level of the received signal, the receiver has to decode the received data, with minimum probability of error. For binary transmission, this is achieved by choosing a decision threshold. Methods: A threshold detection mechanism is evaluated in this paper, which uses parametric probability distribution fitting to the received signal. Results: Based on our earlier work on visible light communication, the idea is extended to a fiber optic channel and it is found that in a fiber optic channel, the threshold value obtained by fitting a Rayleigh distribution to the received data results in error probability approaching zero. Conclusion: It is found that the proposed threshold estimation method works well for both equal as well as unequal apriori probabilities of the binary symbols. The proposed method of threshold detection is adaptive when the apriori probabilities for 1’s and 0’s change.