Stochastic Methods For Data Rate Determination
No Thumbnail Available
Links to Files
Permanent Link
Collections
Author/Creator
Author/Creator ORCID
Date
2014
Type of Work
Department
Electrical and Computer Engineering
Program
Doctor of Engineering
Citation of Original Publication
Rights
This item is made available by Morgan State University for personal, educational, and research purposes in accordance with Title 17 of the U.S. Copyright Law. Other uses may require permission from the copyright owner.
Abstract
The flexibility and re-configurability of software-defined radios (SDRs) has led to the development of radios that are capable of resolving multiple waveforms, which use various modulation techniques- i.e., phase shift keying, frequency modulation, and amplitude modulation - and transmit at various symbol rates. One drawback of the SDR technology is that it still requires that the user set the necessary parameters that allow the radio to reconfigure itself and resolve a signal. A resolution to this problem is to make the SDR more aware of its environment, and enable it to make a decision about how it needs to reconfigure itself so that it can correctly resolve an impinging signal. These types of radios are referred to as cognitive radios. These cognitive radios are autonomous in their operation. Developing a cognitive radio requires the addition of a cognitive engine, which is comprised of detection and estimation algorithms that can detect various signal parameters, including, signal-to-noise ratio (SNR), modulation type, symbol rate, and symbol timing. This dissertation aims to improve the speed and accuracy of the cyclic autocorrelation (CAC) algorithm by combining it with the simulated annealing optimization algorithm. The symbol rate of the signal is a very important parameters as it used to assist in resolving other parameter such as the SNR and the modulation type. When estimating the symbol rate in a cognitive radio environment it is assumed that there is little or no a prior knowledge of the signal, and so the symbol rate estimation algorithm has to blindly make an estimate of the signals symbol rate. The cyclic autocorrelation algorithm is a process that estimates the symbol rate of a linearly modulated signal. It takes advantage of the fact that these signals are cyclostationary. The CAC algorithm correlated the signal with a frequency and time shifted version of itself. In doing this, it tries to identify the optimum frequency, known as the cyclic frequency, which maximizes the correlation. However, in order to find the optimum frequency an exhaustive enumerated search has to be performed, which can be slow and computationally intensive particularly if the resolution of the frequency shifts is fine, or if the number of lags is high. By using the simulated annealing to find the optimum cyclic frequency, it is possible to detect the optimum cyclic frequency without exhaustively evaluating all possible solutions.