Abstract
In this paper, the challenges of parameter estimation for the Gamma‐Gamma turbulence channels with generalized pointing errors are addressed. The Kolmogorov‐Smirnov goodness‐of‐fit statistical test results indicate that the approximate probability density function obtained by the saddlepoint approximation (SAP) method provides a better approximation for a larger value wz, and this means that the proposed method is more efficient for the FSO links over long distances when the transmit divergence angle at the transmitter side is fixed. Also, an additional parameter k needs to be estimated in addition to the shaping parameters α and β under the SAP method. An estimation scheme for the shaping parameters is proposed based on the SAP method. The performance of the proposed estimation is investigated by using the mean square error (MSE) and normalized mean square error (NMSE). The results indicate the proposed estimator exhibits satisfactory performance in both noiseless and noisy environments. The effects of the receiver aperture on the estimation performance are also discussed.
Highlights
Free-space optical (FSO) communication systems offer strong alternatives to radio-frequency (RF) communication systems because of their merits including an unregulated spectrum, ease of deployment, enormous available bandwidth, and high level of security
We study the problem of parameter estimation over Gamma-Gamma turbulence channels with generalized pointing errors
The normalized mean square error (NMSE) is an important performance metric to measure the quality of an estimator and is defined as NMSE = MSE1⁄2bθ/θ2 [38]
Summary
Free-space optical (FSO) communication systems offer strong alternatives to radio-frequency (RF) communication systems because of their merits including an unregulated spectrum, ease of deployment, enormous available bandwidth, and high level of security. The lognormal-Rician distribution [9] is shown to have excellent fit with both simulation and experimental data over a wide range of weak-to-strong turbulence conditions. The maximum likelihood estimation (MLE) with expectation-maximization (EM) or the saddlepoint approximation algorithm is applied to characterize the lognormal-Rician turbulence model parameters [15, 16]. To avoid calculating the complicated integral, a saddlepoint approximation (SAP) method is introduced With this approach, a highly accurate approximation can be achieved over a wide range of channel conditions according to the results of Kolmogorov-Smirnov (KS) goodness-of-fit statistical tests. A highly accurate approximation can be achieved over a wide range of channel conditions according to the results of Kolmogorov-Smirnov (KS) goodness-of-fit statistical tests Based on this method, an approximate log-likelihood function is developed. The simulation results demonstrate that the proposed estimator provides satisfactory performance in both noiseless and noisy scenarios
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call