Measurement, evaluation, and monitoring of subsurface objects often require wireless data transmission between an embedded sensor and an exterior host system. Such technologies find applications in many areas-medical imaging, space exploration, earth formation evaluation in oilfield industries, for example. This paper describes a complete wireless data acquisition system that includes design of a transceiver unit, as well as communication protocols for data encoding and decoding. Although the application considered in this paper pertains to the oilfield industry, the method is applicable to other areas as well. The transceiver design is highly constrained due to requirements on physical size, mechanical stability, and low-power consumption. The host antenna generally consists of ferrite-backed coils wound on a metallic cylinder. A critical requirement in the design of these coils is to produce a desired spatial variation of the magnetic field in a specified region. A genetic algorithm is used to optimize the location and excitation of each coil. A combination of finite-element method and superposition principle is used to speed up the solution of forward problem. The data measured by an embedded sensor are encoded with seven-bit cyclic redundancy code concatenated with Manchester code for error detection/correction and bit synchronization. Coded data are finally transmitted by binary frequency-shift keying modulation scheme. Numerical and experimental results for magnetic field, signal-to-noise ratio, and data demodulation are presented.