Novel Low Duty Cycle Schemes: From Ultra Wide Band to Ultra Low Power

To enable future dense heterogeneous wireless networks, a wireless technology with high scalability potential is required supporting devices with ultra low power consumption as well as high data rates. Ultra wideband (UWB) is a promising technology inherently supporting such a system scalability. However, ultra low power communication is still a major challenge.

A rigorous low duty cycle operation of UWB impulse radio (UWB-IR) transceivers seems a potential key to ultra low power communication. Thereby, two different approaches are possible. Low pulse rate (LPR) systems realize the low duty cycle by a large pulse repetition period. High pulse rate (HPR) systems achieve the low duty cycle by burst-wise transmission at high peak data rate. Analysis and optimization of these two approaches is the main contribution of this project.

National authorities impose hard constraints on peak and average transmit power of UWB devices, which makes transmit power an important optimization parameter. Therefore, the impact of peak and average power constraints on UWB-IR signals is analyzed with focus on the Federal Communications CommissionÕs (FCC) regulations. A joint maximization of peak and average transmit power leads then to modified LPR and HPR schemes with important gains in link margin, coverage and performance.

The modified LPRs correspond to a novel type of UWB-IR schemes directly derived from the FCC peak power constraint. They are especially suited for rich scattering environments and drastically increase the TX/RX-scalability in the sense of a tunable trade-off between transmitter and receiver complexity. Moreover, for energy detectors (ED) with a fixed, sufficiently large integration window, the modified LPRs minimize the bit error rate (BER) under the FCC power constraints even in case of full channel state information at the transmitter (CSI-T).

While the burst-wise transmission of HPR systems allows simple hardware realizations, due to the high peak data rate, even a moderate delay spread can lead to inter-symbol interference (ISI) and major performance degradation. In particular, this is true for EDs, which seem the best suited receivers for ultra low power communication. Therefore, a low complexity postdetection based on maximum likelihood sequence estimation (MLSE) is proposed for non-linear ED frontends and binary pulse position modulation (BPPM). The MLSE is realized by a Viterbi algorithm which operates at symbol rate and works with very limited CSI and number of states. In parallel to ISI cancelation, the proposed MLSE can also be used for efficient symbol synchronization in practical ultra low power receivers. Optimal MLSE metrics based on different levels of CSI are evaluated as benchmarks and highlight the potential of the proposed low complexity MLSE. As a further benchmark, the slightly more complex transmitted reference (TR) scheme is investigated, too. While TR and BPPM with energy detection (BPPM-ED) show the same BER performance in absence of ISI, it is shown that TR is much more robust to ISI than BPPM-ED and for a certain setup even benefits from moderate ISI.

To prove the presented concepts in practice, the more attractive HPR scheme is integrated into an ultra low power modem design. With a MLSE post-detection for ISI cancelation and symbol synchronization in parallel, an ISI robust ultra low power modem design is achieved. Based on a meta study of currently available devices, the power consumption of the modem supporting an average data rate of 500 kbps is estimated to about 1 mW, while its promising real time BER is demonstrated using an over-the-air testbed in a strongly ISI-limited environment with many scattering objects.