University of Patras
Συγχρονισμός σε συσκευές δορυφορικών επικοινωνιών: η περίπτωση των πολλαπλών δακτυλίων
Abstract
dc:descriptionThe objective of this thesis is the analysis and study of the synchronization mechanisms performed by digital satellite terminal receivers when multiple ring constellation diagrams are used. The aim of this thesis is to develop new synchronization techniques that exhibit improved performance and to also propose new methods and ways for evaluating the effectiveness of such receiver submodules. The new trends in the field of digital communications systems and, especially, in modulation and error coding techniques, along with the increasing demand for more effective and interactive applications and services through limited satellite links, have initiated the development of new satellite communications standards. The newest standard is DVB-S2, by the European Space Agency (ESA), in which modern and up-to-date techniques for the design of satellite terminal receiver are required. Meanwhile, the Software Defined Radio (SDR) technology comprises a promising implementation approach as it incorporates the necessary flexibility and versatility for supporting various functionalities and rates into modern receiver structures. Synchronization functions of satellite receivers are complicated and demanding procedures that are related to the estimation of transmission parameters, which correspond to the nominal symbol rate, carrier frequency, phase and to the boundaries of the physical layer frames. These functions determine the complexity and performance of receiver realizations. Thus developing more efficient and simple, in terms of implementation complexity, algorithms and mechanisms is a key objective in such processing platforms. A significant problem that was encountered during the research for the present thesis, was the design and implementation of an efficient digital IF down-converter architecture that is able to manipulate the input IF signal of an SDR receiver through proper processing of its digital input sample stream. The objective of this unit is the shifting of the IF input signal to baseband and the generation of the corresponding baseband I, Q signals. A usual limitation in such realizations is the maximum sampling frequency of front-end ADC circuits. The presented solution addresses this constraint with an architecture that is based on two cascaded units of frequency down-conversion, one with fixed and one with programmable frequency. The programmable unit is driven by the frequency offset estimations of a following baseband processing stage. The advantage of this architecture is the double precision that is achieved compared to the typical approach and for a given sampling frequency. It is worth mentioning that the frequency converter is intended for use in SDR receivers utilizing ADC circuits of moderate sampling frequency. Additionally, in the framework of this thesis, the Symbol Timing Recovery (STR) mechanism based on a second order feedback loop driven by the signal of a timing error detector (TED), was studied and analyzed. The fundamental characteristics of such a control loop, mainly the duration and quality of the initial acquisition are defined through the gain value of the two paths included into the first order loop filter (Proportional-Integral, P-I). This structure comprises a usual approach for communications applications. Conforming to this general analysis for the configuration and the design of the feedback loop, the thesis focuses on the feedback loop incorporating the Non-Data-Aided (NDA) Gardner TED. Using the above analysis and due to the fact that multiple ring constellation diagrams exhibit insufficient performance in such closed loops in comparison to the single ring counterparts of the same mean energy, this thesis proposes a modification of the typical loop deploying the Gardner TED that improves its performance. The increased variance of the input signal of the TED that stems from the changes of symbols with variable magnitude comprises the main reason for the increased variance during the tracking of the timing error in such loop structures. The proposed approach is based on the insertion of a subunit inside the loop structure that adjusts the symbol magnitudes of all rings to a reference magnitude before they are fed into the Gardner TED logic. The above subunit makes use of the internally generated clock of the loop in order to control the magnitude of specific signal samples and does not affect the sample stream at the matched filter input. The application of the specific subunit has the advantage of minimizing the noise during the tracking operation of the loop, which leads to the decrease of the standard deviation of the estimation error when compared to the typical loop structure. The performance of synchronization mechanisms is usually evaluated based on internal parameters that are strongly related to the utilized architecture, the estimated parameter and the operational status of the specific mechanism. The present thesis exploits the use of multiple ring constellation diagrams in modulation process and proposes a generic and new `figure of merit' that is able to determine the performance of various synchronization mechanisms that are incorporated into multiple ring constellation (M-APSK) receivers. A significant advantage of this metric is that it solely based on the processing of the signal at the mechanism's output (extracted symbol values) which enables the utilization of this metric in the presence of frequency, phase and symbol rate offset errors. Another advantage of the proposed metric is that it does not require any knowledge on the transmitted symbols, in contrast to other widely used performance metrics that are applied at the demodulator output, such as the Error Vector Magnitude (EVM) e.t.c. The mathematical analysis of the mean value of the metric under additive white Gaussian noise (AWGN) that is exhibited in this document, includes the theoretical and practical cases. In the first, the receiver is aware of the ring derivation of received symbols, whereas in the second case this information is absent and the receiver determines the metric according to the nearest ring for each symbol. The second case corresponds to realistic receiver realizations. As is shown, in both cases there is a fixed relation between the proposed metric and the commonly used performance metric ratio Es/No for AWGN channels. According to the characteristics described above, the proposed metric can be utilized for the estimation of channel condition under additive white Gaussian noise. This is accomplished through the processing of the STR output signal (symbol values) which usually comprises the first synchronization mechanism in digital satellite terminal receivers. It is worth mentioning that the channel estimation is feasible even under significant carrier frequency offset errors. The significance of the above process is related to the fact that this estimation can be exploited by the following synchronization subunits (of carrier frequency and phase) of the receiver in order to properly adjust and make their operations faster. A second application of the proposed metric is the recovery of large frequency offset errors by processing the signal at the input and the output of the matched filter of the previously mentioned STR structure. The control of frequency offset errors at such point of the receiver processing chain under specific limits, is critical as it strongly affects the performance and efficiency of the following synchronization mechanisms. Finally, this thesis analyzes and presents the implementation of a complete SDR IF demodulator that is compliant to DVB-S2 technology and is based on a reconfigurable hardware platform. This platform incorporates hardware (FPGA) and software (DSP) circuits in a unified environment. The IF demodulator supports single (QPSK/8PSK), two (16APSK) and three (32APSK) ring constellations and comprises a significant part of a full receiver implementation that includes all the necessary functions ranging from the manipulation of the input IF signal to the forwarding of the recovered user information to a Gigabit Ethernet (GbE) LAN. In addition, the IF demodulator implementation includes the hardware realization of the IF digital down-converter into an FPGA device and the software realization of the remaining synchronization procedures starting from the STR into the available DSP processors of the reconfigurable platform. The other necessary procedures for the proper demodulation of the input signal, are: frame synchronization, carrier frequency/phase recovery and amplitude normalization. Furthermore, information is also given on the implementation of the corresponding demapping, error correction and LAN interfacing procedures that are performed in the following processing stages of the DVB-S2 receiver.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2010
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Savvopoulos, Panayiotis
- Σαββόπουλος, Παναγιώτης
Subjects
dc:subject × 16- Ψηφιακές δορυφορικές επικοινωνίες
- Ψηφιακοί δορυφορικοί τερματικοί δέκτες
- Ψηφιακές τεχνικές συγχρονισμού
- Αρχιτεκτονική δορυφορικών τερματικών
- Τεχνικές διαμόρφωσης πολλαπλών δακτυλίων
- Τεχνολογια software defined radio (SDR)
- Digital satellite communications systems
- Digital satellite terminal receivers
- Digital synchronizations techniques
- Satellite terminal receiver architecture
- Multiple ring modulation techniques
- Software defined radio technology (SDR)
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/26276