Airborne Laser Communications with Impulse Response Shortening and Viterbi Decoding
Lee,Sangwoo
The Pennsylvania State University, Department of Electrical Engineering
Kavehrad,Mohsen
The Pennsylvania State University, Department of Electrical Engineering
Free Space Optics (FSO) through cloud has lately attracted considerable attention for a variety of applications. FSO is a promising candidate for emerging broadband applications, considering that RF spectrum is already congested, rendering assignment of additional RF channels difficult and costly.
In FSO, as in Figure 1, the delay spread can be prohibitively large for high optical thickness values and high data rates. One way to reduce ISI is through frequency domain filtering in the form of ultrashort wavelet pulsed transmission as in [2] and [5]. An alternative approach is through employment of an impulse response shortening filter (SIRF) [7][9][10][11][12] followed by maximum likelihood sequential detection algorithm (MLSD) [8]. SIRF has been by now studied mostly in the context of OFDM, but it was also studied in DSL applications with single carrier modulation.
SIRF basically reduces the length of the discrete channel impulse response. The justification for use of SIRF in FSO is the severely distorting nature of the atmospheric channel itself. The ensuing MLSD scheme is known to be the best equalization algorithm, and its complexity constraint indeed requires a short channel length, thus further justifying our employment of SIRF.
A preliminary result shown in Figure 1 shows an FSO channel model for optical thickness value of 10 and cloud length of 1000m. Figure 2 and Figure 3 show its shortened versions when using a 5-tap SIRF to focus most of its energy onto 4 taps of the resulting shortened impulse response, using either MSSNR or MMSE criterion. A preliminary error performance with Viterbi decoding is shown in Figure 6.
In this paper, the authors aim at the following goals. First, they will investigate a channel shortening filter combined with maximum likelihood sequential detection algorithm and compare its performance with suboptimal equalization schemes studied in [5]. Second, based on these results, the authors will discuss the effectiveness of the proposed scheme of MLSD combined with SIRF in extremely time-dispersive FSO channel model to advance beyond the performance reported in [5]. Third, adaptive LMS and RLS algorithms will be applied to SIRF algorithm to evaluate the learning characteristics, as a way to test for the SIRF capabilities in slowly time-varying FSO models.
Sangwoo Lee received his B.S. degree and his M.S. degree in electrical engineering from Yonsei Univ., Seoul, Korea in 1998, and from Pennsylvania State Univ., University Park, US in 2003, respectively. He is currently a PhD candidate in electrical engineering at Pennsylvania State Univ., University Park, US. His main research interests are in resource-efficient multicarrier modulation systems, MIMO signal processing schemes, high-speed equalization schemes, and 3G and beyond mobile communications.