Communications laboratory guide by kamran kiasaleh pdf free download

Hence, the results in [3] may not be extended readily to the case considered here since we are assuming a non-Gaussian due to turbulence field which does not lend itself easily to the KL expansion.

Manuscript received March 30, ; revised May 21, The pdf of is now given by under a fully developed speckle condition. In that event, the op-. As in [7], this integral may tical field amplitude is governed by a Rayleigh random variable, be approximated by the method of steepest descent. Letting which in turn leads to an NE-distributed field intensity.

That is , results in 9 2 where. For this case,5 where. Again, must be computed, 3 which is given by 6. Furthermore, is the product of Poisson distributed and random variables. Integration by part results in 4 10 with the point of stationarity is given by Note that this is an exact MMSE estimator. That is 6 11 Taking the derivative of , given by Considering only the numerator, we have , with respect of and set-.

Computation of the above ting the derivative to zero yields is similar to the computation of. That is, the point of stationarity remains the same.

Letting , results in and 12 This leads to the following equation: , and hence 7 13 By completing the square, the above assumes a Gaussian form, The second derivative is negative, confirming that consti- as shown in the equation at the bottom of the page. Dividing the tutes an ML estimator. As can be seen, the proposed estimator per- 8 forms better than its ML counterpart, especially for small values of and when the observation interval is rather small.

As where may be computed using 3 and 5. It is, nonetheless, in this letter, we are interested in the fully developed speckle result for which the assertion here that a channel model is available at the receiver.

Provided that for all , and , we have. Hence, which confirms the simulation re- sults [MSE and for and , respectively, and when photons Fig. This implies that the proposed estimator will have an asymptotic mean-square error which will be independent of the scintillation index. On the other extreme, i. This implies that the MSE for the proposed esti- mator will be zero under very small scintillation case which is confirmed in Fig.

Free Space Laser Communication Technologies, vol. San Jose, CA, Jan. Mean-square estimation error for the ML and the proposed estimator pp. Vilnrotter and M. Davidson and R. Theory, vol. IT, pp. Karp, R. Gagliardi, S. Moran, and L. Stotts, Optical Chan- nels. New York: Plenum, , pp. For a fully developed speckle result, the [6] H. Stark and J. Montazeri and K. Cao, A. Modiri, G. Sureka, K. Abaz and K. Modiri and K. Michael Cole and K. Kiasaleh, "Scintillation index of a multi-wavelength beam in turbulent atmosphere," Journal of Optical Society of America A, vol.

Burnham, C. Cantrell, A. Farago, A. Fumagalli, K. Kiasaleh, W. Osborne, and R. Kiasaleh, M. Srinivasan, T. Conference K.