Back to list of Group's MSc Theses
Abstract:
The Square Kilometre Array (SKA) is a multi billion dollar international project to create a receiving surface of a million square metres, one hundred times larger than the biggest receiving surface now in existence. The SKA core array will have to be located in a remote area. Therefore countries interested in hosting the SKA core array were requested to perform Radio Frequency Interference measurements at their site of choice. The systems that are to be used in the measurements must conform to a document called the "RFI Measurement Protocol for Candidates SKA Sites", the SKA Memo 37.
The RFI protocol divides measurements into two parts, Mode 1 and Mode 2. Mode 1 is defined for the observation of strong RFI and is relevant for SKA receiver linearity analysis. Mode 2 is defined for the observation of weak interferences, which potentially threatens to obscure weak signals of interest.
In Mode 1, the RFI protocol specifies a dwell time of 2 µs duration over a large 1 MHz bandwidth in the 960 -1400 MHz band (L-band). The reason for this short dwell time is to capture and characterize pulsed interference from radars and Distance Measuring Equipment (DME) in this band. This kind of interference is expected, potentially with very high peak power and short dwell time. Executing these measurements with the spectrum analyzer is impractical because of the very long measurement times. It was therefore proposed to build a dedicated FFT spectrometer from standard components, state of the art FPGA board with a high speed 14 bit ADC.
The proposed system was designed by Dr. Adrian Tiplady and is called the RFI measurement system 4. This thesis investigates whether system 4 measures impulsive RFI as required by the protocol. This was done by simulating the receiver of system 4 and feeding the receiver with simulated DME signals. The simulated receiver findings were then compared to the real receiver findings when similar DME-like signals were injected into the real receiver.
The practical system was found to be able to withstand high power signals from DME systems without suffering from compression and inter-modulation distortion. The maximum signal that the receiver can handle under automatic gain control is 9 dBm, equivalent to 8 mW into 50 Ohms. However, the automatic gain mode was found not to be desirable for RFI measurements because of the unknown gain in the AGC. The simulated receiver was found to measure RFI as expected by the protocol without any compression and intermodulation distortion. Manual mode is recommended for making level measurements in absolute terms.
Back to list of Group's MSc Theses
Abstract:
Traditionally, direct GPS signals are used for navigation and positioning, while the indirect reflected signals are considered a nuisance. However, recent studies show that indirect reflected signals contain some useful scientific data. GPS reflected signals introduce a new and exciting way of doing ocean and land remote sensing, and have even more advantages over traditional remote sensing tools. This project discusses the basic principles and theory of this new technology, and concentrates on reflection points and Fresnel zones. The GPS receivers are placed at different coastal regions within South Africa, and the simulation of the reflection points and Fresnel zones are observed as the GPS satellites pass over South Africa. The East London area was chosen as the location to place the receiver throughout my analysis. Areas of the Fresnel zones reaching a maximum of about 6500 km2 were observed at different receiver heights and the software to make these simulations was written using the IDL language. Results shows that this new tool of remote sensing is feasible and has potential to be used in South Africa. The uses for this new tool include ocean altimetry, ocean, land, ice sheet remote sensing etc.
Back to list of Group's MSc Theses
Abstract:
This thesis describes the design and implementation of a hardware platform for Software Defined Radio (SDR) applications. The Software Defined Radio Forum describes Software Defined Radio as:
a collection of hardware and software technologies that enable reconfigurable system architectures for wireless networks and user terminals. SDR provides an efficient and comparatively inexpensive solution to the problem of building multi-mode, multi-band, multi-functional wireless devices that can be adapted, upgraded or enhanced by using software upgrades. [1]SDR applications perform demodulation, modulation and other signal-processing of a digitised source signal using software modules running on a PC or other capable device. SDR systems typically consist of a transceiver, with a wide-band ADC for signal reception and a DAC for signal generation implemented in hardware, with signal-processing software running on a host PC. In many cases pre- or post-processing of the data occurs in FPGA’s or other programmable devices.
This project was aimed at creating an ADC-based system that could capture samples for processing in software on a host PC, while providing a framework for functionality enhancements through system extensions. The system was designed to integrate with the existing GnuRadio open-source SDR toolkit, with the hardware design was based on the existing USRP system [4]. In the context of the Radar and Remote Sensing Research Group, this system can be implemented in sample capture applications, as well as software RADAR systems through system extension.
The USRP system was developed by the Free Software Foundation GnuRadio project to complement their open-source SDR software. It is a single board housing a large FPGA which accepts input signals from add-on ADC daughter-boards [5], and feeds output signals to DAC boards. The USB data transfer to and from the host is carried by a Cypress FX2 microcontroller.
This system is split into two modules:
The ADC module implements a 12-bit 65MSPS dual-channel ADC [8] to allow for wide-band sampling of analogue inputs. The sample stream is transported to the USB module via LVDS, where it is buffered before transfer to the host PC. FPGAs are housed on each of the modules, which provide the logic for buffering of data and flow control. Firmware developed for the FPGAs handles the data flow and debug signals, while the FX2 firmware handles the USB transfers to the host PC.
The FX2 firmware used in this project is a modified version of the open-source C code, as are the low-level libraries which interface with the application-level GnuRadio software. As the GnuRadio toolkit is Linux-based, this system was developed on an Ubuntu 6.06 Linux platform. The open nature of the Linux development platform meant that existing code (both firmware and software) could be readliy re-used and modified, which sped up the development cycle considerably. Several proprietry applications and sources are available for Microsoft Windows development using the Cypress FX2 microcontroller, but these were found to have lower levels of support and documentation, which would have led to a longer development cycle (alo considering that a large number of freely-available SDR modules have already been developed by GnuRadio).
The system was assembled and tested. Due to time constraints however, several limitations were imposed on the effecitve system performance, namely; the sample resolution was reduced to 8 bits to ensure data frame integrity, and the sample rate was reduced considerably to ensure that no discontinuities were introduced in the captured sample streams. These two limitations introduced compromises in the overall system performance, but recommendations were presented to recover the losses.
Back to list of Group's MSc Theses
This page was last updated in July 2007 (RL)