Increasing concern about the rapid escalation of environmental pollution has led to strong legislation to ensure, for example, that the emission of pollutants from vehicles and industries is controlled to an acceptable level. As a consequence, there has been a rapid expansion of research into developing more efficient and low-cost gas monitoring systems. Currently, commercial solid-state atmospheric gas detection systems are based on one sensor for each gas, while research systems are an array of sensors for the detection of multiple gases. In this research, techniques are developed whereby more than one gas is detected using a single resistive gas sensor. A novel modulated temperature technique was used to enhance the selectivity of the resistive SnO2 gas sensor. Fast Fourier transforms was used to extract the Fourier coefficients. These in turn were used as input to neural networks for training and subsequently for prediction purposes. The result has shown that a single doped SnO2 resistive microsensor can be used to classify binary gas mixture in air. The research objectives have been fulfilled in that a novel way in detecting the components and the concentration level of a binary gas mixture was developed. Additionally, a low-cost low-power intelligent gas monitoring system was designed. This included the design of a novel temperature/thermometer circuit.