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 Modelling and forecasting metal prices: evidence from developing African economies(2017) Handura, Vetjevera MercyDeveloping Africa has been heavily dependent on primary commodities for decades as these countries are rich in mineral resources and often tend to export that and little else. However, commodities are highly susceptible to volatility and their effects on these economies are enormous. This paper investigates the extent to which the GARCH and EGARCH models can accurately be employed to model and forecast metal prices. Also, a p-dimensional VECM is formulated in establishing the extent to which the metals are co-integrated. Seven metals - Aluminium, Copper, Gold, Lead, Nickel, Platinum and Zinc have been employed for the purpose of this study. The models yielded satisfactory prediction results, albeit mixed findings in terms of the superiority of the models. Nonetheless, we conclude that the results are sufficient in aiding African economies in deriving appropriate policies and trading strategies so as to capitalise on export revenues, resulting in increased GDP and overall economic growth and development of their countries. Key Words: volatility, GARCH, EGARCH, VECM forecast, metal pricesItem An investigation into jamming GSM systems through exploiting weaknesses in the control channel forward error correction scheme(2017) Timm, GarethThe ability to communicate effectively is of key importance in military scenarios. The ability to interfere with these communications is a useful tool in gaining competitive advantages by disrupting enemy communications and protecting allied troops against threats such as remotely detonated explosives. By reducing the number of corrupt bits required by using customised error patterns, the transmission time required by a jammer can be reduced without sacrificing effectiveness. To this end a MATLAB simulation of the GSM control channel forward error correction scheme is tested against four jamming methodologies and three bit corruption techniques. These methodologies are aimed at minimising the number of transmitted jamming bits required from a jammer to prevent communications on the channel. By using custom error patterns it is possible to target individual components of the forward error correction scheme and bypass others. A ran dom error approach is implemented to test the system against random errors on the channel, a burst error approach is implemented to test the convolutional code against burst errors, and two proposed custom error patterns are implemented aimed at exploiting the Fire code’s error detection method. The burst error pattern approach required the least number of transmitted jamming bits. The system also shows improvements over current control channel jamming techniques in literature.