Low complexity indoor wireless data links using chirp spread spectrum

This dissertation proposes the use of Chirp Spread Spectrum (CSS) modulation for the transfer of high-speed (>10 megabits per second) data in the indoor wireless environment. CSS utilizes the properties of chirp signals to combat the effects of multipath fading. The major advantage CSS has over other spectrum spreading techniques is its analog-only implementation. Two main methods for deploying CSS are described: Binary Orthogonal Keying (BOK), in which chirp signals with different characteristics represent different data symbols, and Direct Modulation (DM), where data is modulated using a conventional non-coherent modulation scheme, and then has its bandwidth spread by the chirp signal. Performance equations for the Gaussian channel are derived for both systems. Several other key system design parameters are identified and explored, including chirp windowing, occupied bandwidth, and quasi-RAKE filter configuration. Computer simulations and prototype measurements show close agreement with the developed theory. Raw bit error rates less than 10−5 at 20 megabits per second are achieved, with a bandwidth of 200 MHz in non line of sight wireless channels over 10 m with less than 2 milli-Watts of transmit power. Techniques for designing SAW chirp filters with both linear and nonlinear time-frequency responses are described, and measurements of a prototype SAW device are presented. Novel methods for enhancing the throughput of a binary CSS system are presented which increase the number of bits per symbol of the modulation CSS signal. Computer simulations show that data rates of 50 megabits per second are achievable for a single channel system using this method.