Practical Considerations for Radar Embedded Communication

Abstract

This thesis expands on the previous work done in the area of intra-pulse radar embedded communication by examining some of the practical aspects of the waveform design. Communication waveform mismatches between the tag and receiver due to multipath distortion, sampling rate differences and using different lengths for the radar waveform are explored for each of the three previously developed design methods. The Dominant-Projection approach is shown to be robust to most mismatches while the other two approaches significantly degrade or fail with any mismatch. Lack of synchronization between the receiver and tag is shown to increase the occurrence of symbol errors, since the receiver is required to search over multiple samples for the communication waveform sent by the tag. Attempts to reduce the number of errors caused by the lack of synchronization are also made, first by taking a three sample average of the filter output and second by generating waveforms with lower local cross-correlation, with both attempts shown to be unsuccessful. Other attempts are also made to improve the waveform design. It is shown that temporal expansion can be used to either improve symbol error rate or reduce the amount of bandwidth expansion required. A rule-of-thumb is developed for the bandwidth expansion versus temporal expansion trade-off. It is also shown that more of the dominant space can be projected out with Dominant-Projection to reduce the probability of symbol error, but this comes at the cost of being more susceptible to being detected by an intercept receiver.