As the size and power consumption of microelectronic circuits continues to decrease, passively-powered sensors promise to come to the forefront of commercial electronics. One of the most promising technologies that could realize this goal is backscatter sensing. Backscatter sensors could harvest power from and modulate data onto an impinging carrier waveform. Currently radio frequency identification (RFID) technology passively powers itself and transmits statically stored data. However, this technology has two major weaknesses: lack of resiliency against narrowband interference and slow data rates. Both of these issues could be detrimental in sensing applications. This thesis will lay out a method for addressing both of these weaknesses through a unique application of spread spectrum encoding. Instead of spread spectrum being viewed as the multiplication of an already encoded data sequence with a periodic pseudorandom sequence, each sequence could be viewed in an aperiodic manner, where a single period of a pseudorandom sequence represents a data symbol. In this manner, backscatter sensors not only benefit from the increased resiliency that spread spectrum provides, but also can have higher data rates, since multiple bits can be encoded on a single symbol and multiple nodes can be read simultaneously, using spread spectrum multiple access techniques. In this thesis, 63-chip and 255-chip Kasami sequences, as well as 127-chip Gold sequences, will be analyzed for their use in various aperiodic direct sequence spread spectrum/multiple access system configurations (systems that have up to three nodes and use up to four different aperiodic sequences per node to represent different symbols). For each different configuration, near-"ideal" code configurations/rotations will be determined for use in the system.

Backscatter radio - the broad class of systems that communicate using scattered electromagnetic waves - is the driving technology behind many compelling applications such as radio frequency identification (RFID) tags and passive sensors. These systems can be used in many ways including article tracking, position location, passive temperature sensors, passive data storage, and in many other systems which require information exchange between an interrogator and a small, low-cost transponder with little-to-no transponder power consumption. Although backscatter radio is maturing, such systems have limited communication range and reliability caused, in part, by multipath fading. The research presented in this dissertation investigates how multipath fading can be reduced using multiple antennas at the interrogator transmitter, interrogator receiver, and on the transponder, or RF tag. First, two link budgets for backscatter radio are presented and fading effects demonstrated through a realistic, 915 MHz, RFID-portal example. Each term in the link budget is explained and used to illuminate the propagation and high-frequency effects that influence RF tag operation. Second, analytic envelope distributions for the M x L x N, dyadic backscatter channel - the general channel in which a backscatter system with M transmitter, L RF tag, and N receiver antennas operates - are derived. The distributions show that multipath fading can be reduced using multiple-antenna RF tags and by using separate transmitter and receiver antenna arrays at the interrogator. These results are verified by fading measurements of the M x L x N, dyadic backscatter channel at 5.8 GHz - the center of the 5725-5850 MHz unlicensed industrial, scientific, and medical (ISM) frequency band that offers reduced antenna size, increased antenna gain, and, in some cases, reduced object attachment losses compared to the commonly used 902-928 MHz ISM band. Measurements were taken with a custom backscatter testbed and details of its design are provided. In the end, this dissertation presents both theory and measurements that demonstrate multipath fading reductions for backscatter-radio systems that use multiple antennas.