3. Electronic Theses and Dissertations (ETDs) - All submissions
Permanent URI for this communityhttps://wiredspace.wits.ac.za/handle/10539/45
Browse
2 results
Search Results
Item Channel estimation techniques for filter bank multicarrier based transceivers for next generation of wireless networks(2017) Ijiga, Owoicho EmmanuelThe fourth generation (4G) of wireless communication system is designed based on the principles of cyclic prefix orthogonal frequency division multiplexing (CP-OFDM) where the cyclic prefix (CP) is used to combat inter-symbol interference (ISI) and inter-carrier interference (ICI) in order to achieve higher data rates in comparison to the previous generations of wireless networks. Various filter bank multicarrier systems have been considered as potential waveforms for the fast emerging next generation (xG) of wireless networks (especially the fifth generation (5G) networks). Some examples of the considered waveforms are orthogonal frequency division multiplexing with offset quadrature amplitude modulation based filter bank, universal filtered multicarrier (UFMC), bi-orthogonal frequency division multiplexing (BFDM) and generalized frequency division multiplexing (GFDM). In perfect reconstruction (PR) or near perfect reconstruction (NPR) filter bank designs, these aforementioned FBMC waveforms adopt the use of well-designed prototype filters (which are used for designing the synthesis and analysis filter banks) so as to either replace or minimize the CP usage of the 4G networks in order to provide higher spectral efficiencies for the overall increment in data rates. The accurate designing of the FIR low-pass prototype filter in NPR filter banks results in minimal signal distortions thus, making the analysis filter bank a time-reversed version of the corresponding synthesis filter bank. However, in non-perfect reconstruction (Non-PR) the analysis filter bank is not directly a time-reversed version of the corresponding synthesis filter bank as the prototype filter impulse response for this system is formulated (in this dissertation) by the introduction of randomly generated errors. Hence, aliasing and amplitude distortions are more prominent for Non-PR. Channel estimation (CE) is used to predict the behaviour of the frequency selective channel and is usually adopted to ensure excellent reconstruction of the transmitted symbols. These techniques can be broadly classified as pilot based, semi-blind and blind channel estimation schemes. In this dissertation, two linear pilot based CE techniques namely the least square (LS) and linear minimum mean square error (LMMSE), and three adaptive channel estimation schemes namely least mean square (LMS), normalized least mean square (NLMS) and recursive least square (RLS) are presented, analyzed and documented. These are implemented while exploiting the near orthogonality properties of offset quadrature amplitude modulation (OQAM) to mitigate the effects of interference for two filter bank waveforms (i.e. OFDM/OQAM and GFDM/OQAM) for the next generation of wireless networks assuming conditions of both NPR and Non-PR in slow and fast frequency selective Rayleigh fading channel. Results obtained from the computer simulations carried out showed that the channel estimation schemes performed better in an NPR filter bank system as compared with Non-PR filter banks. The low performance of Non-PR system is due to the amplitude distortion and aliasing introduced from the random errors generated in the system that is used to design its prototype filters. It can be concluded that RLS, NLMS, LMS, LMMSE and LS channel estimation schemes offered the best normalized mean square error (NMSE) and bit error rate (BER) performances (in decreasing order) for both waveforms assuming both NPR and Non-PR filter banks. Keywords: Channel estimation, Filter bank, OFDM/OQAM, GFDM/OQAM, NPR, Non-PR, 5G, Frequency selective channel.Item A spatio-temporal modelling and analysis of digital sensor data for underground mine health and safety(2017) Opiti, Calvin OduorHealth and safety of employees within their work environment is critical. In the mining industry and especially in underground mines, monitoring and management of health and safety of employees is particularly important Most underground mines today are not fully mechanized, except for coal mines. The industry thus still relies on and employs human personnel. Monitoring and managing these mines and hence personnel health and safety as they undertake their trade is therefore a necessity. Implementation of technology, especially in digital sensor systems and real-time spatial analysis systems, provides a means by which health and safety risk factors can be monitored and information gathered to facilitate determination of prevailing risks or prediction of such risks. Technology therefore can be used to make better decisions and implement specialized emergency response to avert or reduce the extent of injuries, casualties and damages in an underground mine. This research project looks into determination of prominent risk factors in an underground mine, determination of parameters for modeling of such risk factors and the implementation of ESRI’s ArcGIS platform for the retrieval and analysis of streaming sensor data about this parameter from an underground mine. A proof of concept (POC) system is developed that analyses streaming digital sensor data and determines the status of the underground mine environment. The results from this analysis are displayed in a dashboard application for a control room environment. The results and achievements of this research project, especially from a dashboard system perspective, show the possibilities of an integrated GIS-based solution for real-time data processing and determination of the prevailing conditions in an underground mine. This solution also opens up a wide pool of possibilities through which systems integration and its benefits can be achieved, especially in underground mines and focusing on health and safety, as previously silo systems can be integrated at data levels, enabling data sharing, analysis, predictions and making of informed decisions.