This Small Business Technology Transfer (STTR) Phase I project will model, design, simulate, develop and transition to industrial partners the key elements of Frequency Agile Reconfigurable Radios that incorporate both medium-data-rate wireless wide area network (WWAN) radios and ultra-high-data-rate short-range wireless personal area network (WPAN) radios. Demand for wireless data will soon outstrip available RF spectrum. While limited new spectrum is available, the complexity of wireless device RF front ends is already daunting and adding more RF bands will just add more switches and more surface- acoustic-wave (SAW) filters. In the near term, the Frequency agile aspect of the proposed work will enable RF front ends that can cover the entire spectrum from 0.8GHz - 5GHz with only 3 RF front ends. This research focuses on the design of tunable LC filters coupled with adaptive cancelation of transmit tone leakage and the design of high efficiency watt-level power amplifiers. Finally, the potential for peer-to-peer ad-hoc wireless networking to provide auxiliary access to large data files of common interest will be assessed; and, a short range 60GHz transceiver front end will be designed such that the overall radio can be rapidly reconfigured to operate between 0.8-5GHz or at 60GHz.
The broader impact/commercial potential of this project is that a frequency agile RF front end architecture can dramatically improve the cost and performance of wireless front ends by allowing a single transceiver and tunable RF Filter to cover any frequency band within an entire octave of frequency. An early payoff for the project will be the reduction in the number of RF switches and RF filters in the front end of wireless devices, especially cellular phones. As the number of distinct frequency bands that must be addressed by modern cell phones increases, their cost increases and their RF performance decreases because the on/off impedance ratio of available RF switches is small. The frequency agile RF front end will improve the RF performance of cellular phones due to decreased RF switch losses. This would simplify the government?s process to issue new frequency bands without considering radio complexity. The long term commercial potential for the project is its ability to dramatically reduce the data that travels over the WWAN by facilitating high data rate ad-hoc networks between mobile devices using a WPAN radio.
Under Phase 1 of an STTR award, Carley Technologies, Inc. (CarleyTech) has developed and optimized the design of tunable Microelectromechanical Systems (MEMS) based Inductor (L) Capacitor (C) resonators (tunable LC resonators) for use in tunable RF front end filters. These tunable high quality factor (Q) LCs enable Frequency Agile Reconfigurable Radios that can be tuned in real time across a wide range of frequnecies. Such tunable RF front ends would enable an entirely new class of mobile radios that could efficiently operate on any frequency band across a wide frequency range. In order to achieve the performance required for compatibility with existing wireless standards, our research concluded that tunable LCs with a minimum unloaded Q of over 100 would be required. An example of one of the tunable RF front ends is shown in Image 1. Q’s of over 100 are significantly higher than can be achieved using normal on-chip or in-package inductors. Note, wirebonds have unloaded Qs on the order of 50 in the 1-2GHz range; however, there is a strong trend in RF front end module design to move from wirebonded assembly to flip-chip assembly, which removes wirebonds as an option for creating modest-Q inductors. Finally, the tunable RF front ends require a large number of inductors making traditional board level inductors too bulky and too expensive. The major impact of this research is the development of a design for integrated tunable LC based RF front ends that are fabricated using traditional VLSI processing equipment. More specifically, under the Phase 1 STTR award we developed designs for micromachined high-Q inductors which, when combined with high-Q MEMS tunable capacitors, allows the construction of RF front end filters that can be tuned over an octave in frequency range and that can provide over 40dB of transmit to receive isolation at a 4% bandwidth separation and an insertion loss of less than 2.5dB on the receive path. The properties of these high-Q inductors were verified using COMSOL field simulations to explore variations in geometry and we also demonstrated the potential improvement in Q that could be achieved through the use of magnetic fill compound using additional COMSOL field simulations. Finally, we carried out initial fabrication experiments in order to validate the viability of fabricating the out-of-plane MEMS inductor concept that was developed above. Through these cleanroom experiments, we were able to demonstrate the self assembly of an out-of-plane MEMS inductor with a coil diameter of approximately 200um (see Image 2). In addition, we developed, on paper, a number of manufacturing process flows that supported the idea that these inductors could be moved into high volume high yield production.
