Spread spectrum communication over a fading multipath HF channel using transform domain signal processing and a transmitted reference signal
Joseph Michael Smallcomb · OhioLink ETD Center (Ohio Library and Information Network) · 1992
development of other types of transform algorithms, the field of Transform Domain (TD) signal processing has emerged from DSP. 1.2.History of Rake Correlator Classical digital communication over a High Frequency (HF) channel has been investigated since the 1930s [2].The ability to communicate over the horizon using an HF channel made it very attractive.However, the problems of fading and multipath that occurs naturally on an HF channel make it difficult to communicate reliably.Not until 1958, with the development of the Rake correlator by Green and Price [2], did it become theoretically feasible to attempt highly reliable digital communication over a wide band HF channel.Since 1958, research in the development of the analog Rake correlator has continued to the present.However, these analog versions are limited physically both in their size and sensitivity.1.3.Current Research in TD Signal Processing TransformDomain (TD) signal processing for digital communication signals has recently been employed on existing communication waveforms (e.g., Binary Phase Shift Keying) to preprocess received signals before demodulation [3]-[7].These techniques sample the received signal, transform it into the frequency domain using a DFT, petform the TD operation, and transform back into a time domain signal using an inverse DFf.After returning to the time domain, the received signal is then demodulated by conventional means.Other TD signal processing techniques have been suggested for use in demodulation and matched filtered operations performed by the receiver [7]-[9].In these papers, the communication waveform is primarily designed to facilitate the direct demodulation of the signal in the transform domain.1 It is assumed here that each cyclic shift of the base vector represents a valid symbol.It is possible, and in some cases may be desirable, to only use some binary fraction (i.e.Ml2, M/4, etc.) of the M possible symbols as the signal set. 4 1.5.Outline of this Dissertation Chapter 2. In this chapter we examine the fundamental concepts of HF communication in general.The physical properties and modelling of the HF skywave channel are discussed.The basic principles of various frequency diversity techniques and the use of Rake correlators to exploit these techniques are examined.Chapter 3. In this chapter we discuss the use of TD signal processing.The use of CCSK waveforms and Rake correlation in the TD are discussed.The basic discrete signals and channel model used in this dissertation are also presented.Finally, we introduce the use of a transmitted reference signal for estimating the channel impulse response.Chap ter 4. In this chapter we calculate the performance of the various communication systems described in chapter 3.These include traditional Rake systems for comparison with the TD systems.Also, the performance of the TD/CCSK Rake receiver under ideal conditions with and without reference signals is derived.The performance of the TD/CCSK Rake receiver using only the instantaneous reference signal to estimate the channel impulse response is also derived.Chapter 5.In this chapter we introduce variations of the basic TD/CCSK Rake receiver.This includes the concepts of frequency domain interleaving, tail clipping, center clipping and time-averaging to improve the estimate of the channel impulse response.The description as well as the theoretical performance improvements achieved over the basic TD/CCSK Rake receiver using these algorithms are presented.Chapter 6.Here we discuss the implementation factors of the TD/CCSK Rake receiver.The use of binary maximum length sequences, chirps, and filtered impulse trains as potential CCSK waveforms are examined.Also, the performance effects of both time and frequency synchronization errors are calculated and discussed.