Systolic Viterbi decoding for VLSI implementations

Abstract

Convolutional coding with Viterbi decoding is a Powerful forward error correction method for channel coding in communication systems. The most straightforward implementation of the Viterbi algorithm is a completely sequential one where every state is evaluated, in sequence, in one processor. The required area of the fully sequential Viterbi decoder is small but throughput is disappointing. For small state Viterbi decoder (small interconnection area), the fully parallel implementation is used where one state is assigned per processor. We propose the systolic architecture for both trace-back and register-exchange methods. The throughput is very high and the amount of data storage is reduced in our designs. The register-exchange method has even better performance than trace-back method. For large state Viterbi decoder (large interconnection area), we design an area-efficient architecture that offers a favorable area- speed tradeoff. Our architecture has the smallest overhead among the existing designs. The time-efficient architectures are suitable for small state and high speed Viterbi decoder, while the area-efficient architecture is suitable for large- state and moderate high speed Viterbi decoder. Waterflood infill drilling has been used successfully in increasing oil recovery from many heterogeneous West Texas carbonate reservoirs. The majority of the carbonate reservoirs in West Texas have a 20-acre well spacing and some are lower than 10-acre well spacing. The potential of future waterflood infill drilling in these reservoirs will hinge on its economic return. As an alternative to waterflood infill drilling, miscible C02 flooding has been used for increasing the oil recovery from the carbonate reservoirs since the early 1980's. The success Of C02 flooding depends largely on reservoir conditions, process implementation, and economic environments. This thesis presents a technical and economical comparison of waterflood infill drilling and C02 flooding on Monahans Clearfork Unit. It provides a case study to show how the basic field data and related information an be used to evaluate the potential of waterflood infill drilling and C02 flooding for future field development. Results show that the impact of well spacing reduction on oil recovery is substantial. Estimated ultimate recovery efficiencies from waterflood infill drilling at the average well spacing of 46 acre, 36-acre and 19-acre well spacings are 12.0%, 15.4% and 23.7% OOIP, respectively. The projected ultimate recovery efficiencies for the 15-acre and 10-acre well spacings based on well spacing and recovery efficiency correlation are 26.2% and 29.8% of OOIP, respectively. The economic evaluation of waterflood infill drilling under current economic environment indicates that waterflood infill drilling above 19-acre is profitable. However, pattern waterflood infill drilling down to 10-acre is not economically feasible. The economic sensitivity analysis based on a selected section study shows that the economic return of 10-acre infill drilling is more sensitive to the capital investment than the operating cost. The large investment is mainly related to drilling and completion cost. An oil price of $27 per barrel is necessary to make the 10-acre infill drilling profitable. An analogous study Of C02 flooding indicates that an incremental oil recovery of 12% OOIP by C02 flooding is attainable. At this incremental recovery, C02 flooding at current 19-acre well spacing is profitable. The results based on economic sensitivity study show that C02 flooding with incremental recovery efficiency below 11% of OOIP is not economically feasible and the main factor affecting the economic outcome of C02 flooding is the cost Of C02. Based on the comparative study of waterflood infill drilling and C02 flooding, this study is in favor Of C02 flooding for the future development of Monahans Clearfork Unit.