Over the past 15 years, our group has been developing passive wireless multi-sensor systems based on surface acoustic wave (SAW) encoded delay line sensors. The work has been principally developed through NASA SBIR/STTR programs, NASA developmental contracts, and NASA graduate fellowships. This period of effort has resulted in the development of orthogonal frequency coded RFID sensor devices, specialized transceiver (reader) technology, and adaptive filter correlator post processing techniques. Operational systems and devices were demonstrated from 250 MHz to 1 GHz.

This paper presents a multi-echo transit analysis for one-port reflective delay line SAW sensors. The closed form solution for the mixed-parameter analysis is presented. A discussion and results for SAW RFID and sensor delay lines is presented. The results are also applicable to delay lines having identical transducers.

Over the last 13 years, our group has been developing passive, wireless SAW device technology and the system has been continually evolving, providing gains in all aspects of the technology. Initial system proof-of-concept moved from 250 MHz to the current 915 MHz, which is a good choice from device and antenna size, and wireless RF component availability. Range has increased from less than a meter, to over 25 meters, and current testing is moving towards 100 meters for a NASA application. Several successful custom transceivers were developed and delivered to NASA and the US government. From custom-made systems, current efforts use a software defined radio (SDR) Universal Software Radio Peripheral (USRP) transceiver approach as the interrogator-receiver. This commercially off the shelf (COTS) hardware greatly reduces development time and cost and ensures rapid future improvements. The advantageous and SDR approach will be discussed, which include RF parameter adaptability in center frequency and bandwidth. The current status of the SAW wireless temperature sensors sensitivity and range will be used as the performance parameter measure, since it is the most ubiquitous of all sensor applications.

The universal software radio peripheral (USRP) is a versatile software defined radio (SDR) platform, developed by Ettus Research™, which is intended for a wide variety of applications ranging from communication links to RADAR. We have investigated another application of the USRP by implementing a transceiver capable of interrogating passive, wireless surface acoustic wave (SAW) sensors centered at 915MHz. Interrogation of wideband orthogonal frequency coded (OFC) SAW sensors imposes strict requirements on the timing and synchronization of the transceiver. In the standard mode of operation, samples are generated and streamed between the USRP and host computer, introducing latency and bandwidth limitations due to the sampling bus. To achieve the performance required for this application, the USRP FPGA has been modified to introduce new functionality. Extraction of the sensor temperature is accomplished with a custom matched filter correlator. The system is capable of interrogating multiple sensors and can quickly reconfigure the USRP. Demonstration of the USRP wireless sensor system is achieved by interrogating wireless SAW OFC sensors at 915MHz and extracting the sensor temperature.

Modern software defined radio (SDR) technologies have enabled surface acoustic wave (SAW) sensor interrogation systems that are small, inexpensive, and fully integrated. The universal software radio peripheral (USRP) is available as a commercial-off-the-shelf (COTS) SDR that can be programmed to fit a wide variety of applications and needs. Recent work at the University of Central Florida has demonstrated the USRP B200 as a SAW sensor interrogator. This paper expands upon previous efforts by implementing a standalone interrogation platform which utilizes the USRP B200 and embedded microcomputer. Recent advances in embedded processing platforms have enabled full Linux environments to be run, enabling high performance computation in a compact package. One such embedded platform, the MinnowBoard MAX, has been employed to perform the complex matched filter correlator post processing techniques as well as handle programming and communication with the USRP. The unit is self contained and can be operated without a traditional Desktop or Laptop. Performance of the interrogator has also been improved by adding and external RF switch and amplifier. The system is demonstrated by interrogating orthogonal frequency coded (OFC) SAW sensors at 915MHz and extracting the sensor temperature.

This paper will provide a fundamental approach to the design of SAW transducers and filters. Although SAW devices are often modeled with complex and large programs, the basic design principles can be analyzed with simple programs and commercially available analysis tools, such as Mathcad or Matlab. The beginning of the paper reviews the basis for the impulse response model. Design equations for fundamental window time functions as applied to apodized transducers will be presented. The development shows an analytical approach to the solution of apodized transducers. The solutions provide the SAW radiating beam profile as a function of frequency as well as the frequency dependent acoustic transducer parameters. The electrical network effects on the overall transducer response are presented which provides a complete first order analysis for fundamental SAW trandsucer design.

This paper presents a theoretical development of the effects of inter-sensor interference applicable to surface acoustic wave (SAW) multi-sensor systems. An approach is presented that statistically estimates desired wireless sensor signal degradation due to interference from other sensors in the system. A simple theoretical model is developed to estimate the inter-sensor interference. Simulations are performed that model the overlap of SAW signals, with consideration given to single frequency, single chip sensors as well as OFC SAWsensors. The simulations predict the degradation of the desired signal to noise due to other sensor responses moving into the desired time window. Results presented in this paper show that the SNR degrades rapidly as undesired sensors move into the desired sensor time window. The model is applicable to a broad range of possible sensor embodiments.

The software defined radio (SDR) provides a unique platform to interrogate passive, wireless sensors. The SDR platform defines many radio functions in software, rather than hardware. This allows for a very versatile interrogation platform that can be quickly reconfigured for diverse scenarios. This paper investigates the Universal Software Radio Peripheral (USRP), a commercial SDR developed by Ettus ResearchTM, to determine its viability to interrogate passive, wireless sensors. In particular, the N200 USRP and the WBX daughterboard are considered because they allow for the greatest possible bandwidth for this platform (40MHz) and frequency tuning that is ideal for surface acoustic wave (SAW) sensor interrogation (50MHz-2.2GHz). In the default operation mode, the USRP continuously streams data to and from the host computer. The host computer generates samples for transmission and processes any received samples. Synchronization of the transmit and receive chains becomes difficult due to latency in the communication medium (USB or Ethernet) between the host and USRP. While this mode is sufficient for most narrowband applications, wideband applications are more difficult to achieve because of the high sampling rates required (even in a baseband system such as the USRP). A prototype transceiver (passive tag reader) is developed by modifying the N200 FPGA to introduce new functionality to the USRP. These modifications include a custom interrogation signal generator (linear chirp) and triggering of the receiver based on the transmission state and desired listening time. These modifications are discussed and the modified transmit and receive characteristics are analyzed. Finally, a passive, wireless SAW OFC sensor is interrogated to demonstrate performance as a passive sensor transceiver.

This paper presents a wafer-level integrated SAW sensor that uses conventional thin-film fabrication for the SAW, and the direct write of a thicker dissimilar metal for the antenna. An automated thermal spray process deposits the antenna conductor onto the SAW substrate, providing ease of fabrication, optimal film thickness, superior adhesion, and application specific materials. Results will highlight the direct printing of thick copper traces onto whole LiNbO 3 wafers. The design of a 915 MHz meandered dipole antenna with low mismatch loss and maximized radiation efficiency over a 7% SAW fractional bandwidth is demonstrated. Experimental performance results of antennas fabricated on standard FR4 and on-wafer by the direct write process are contrasted. Temperature sensors are fabricated on YZ-LiNbO 3 and their wireless performance is evaluated.

This paper discusses a novel coherence multiplexed system for passive surface acoustic wave (SAW) RF-identification (RFID) tags. The method is adapted from optical coherence-domain reflectometry (OCDR) which is widely used today in optical coherence tomography (OCT) for medical applications. The paper discusses the coherence multiplexing technique as applied to SAW RFID. A broadband white Gaussian noise source is used as the interrogation signal and one of the major advantages is the large signal dynamic range obtained. A prototype transceiver system was built at 915MHz and used for wireless interrogation of SAW tags. The signal processing technique is discussed and the results of wireless multiplexing of four RFID SAW tags are shown.

Novel wireless RFID surface acoustic wave (SAW) sensor devices have been developed to address needs that often cannot be met by other technologies. These sensors are fully passive, and use only the interrogation energy to encode the device RFID and sensor information.This work will present results from an operational 915 MHz wireless SAW sensor system that demonstrates 8 SAW temperature sensors, received simultaneously from a single interrogation signal, having over 3 meters range. In addition, examples of room temperature reversible hydrogen SAW gas sensors with over 20 dB dynamic range and real-time response time (<1 sec) will be shown. An adaptive matched filter and synchronous correlator techniques are used at the receiver to extract the sensor information. The sensor examples and transceiver hardware and software approach will be discussed.

The thin-film acoustoelectric effect in surface acoustic wave (SAW) devices describes the interaction of electrical energy between a SAW in a piezoelectric medium and a thin-film placed in the wave's propagation path. The real-time observation of the thin-film acoustoelectric interaction is useful in the design and characterization of SAW-based thin-film chemical and physical sensors (i.e. temperature, humidity, viscosity, voltage, current, hall effects, etc.). An in-situ test fixture was designed to be mechanically, thermally and electrically stable. Data has been taken for many SAW devices and over a wide range of frequencies. The results show that the use of the in-situ procedure yielded: good agreement between theoretical predictions and the measured data, characterization a SAW-H-2 gas sensor in real-time and various methods to calibrate the film deposition system and procedure. This paper presents the approach taken in configuring an electron beam evaporation system for ultra-thin-film characterization and the design of test fixtures, data acquisition configuration, and experimental procedures to extract and analyze SAW parameters in real time, and to extract the thin-film properties under test.

The focus of this paper is to discuss theoretical and practical applications of a coherent correlator transceiver (CCT) system approach for SAW RFID sensors, with the vision toward low cost sensing. For high volume applications, it appears achievable to meet the ultimate system goals and costs with RF device integration. As with most technologies, the sensor device and system volumes will drive the cost down, and it seems reasonable to expect a similar product cycle as with SAW filters. Theoretical predictions of range, SNR, and correlation properties will be shown for differing system parameters, such as bandwidth, power, coherent integration, and ADC parameters. A comparison of SAW device resonant, CDMA and OFC type signal formats will be given and an approach for signal detection and extraction discussed. Recent measurement results for a 915 MHZ, 64 MHz bandwidth, CCT system using OFC SAW temperature sensors will be given and compared to predictions, which demonstrates the practicality and implementation of the system architecture.

This paper explores the effect of time overlapping multiple Bragg reflectors on the waveform of a coded, passive SAW sensor. Typical SAW sensors utilize serial reflectors, or chips, with a constant waveform shape; however, by using a multi-track SAW configuration the overlap of chips can be optimized for a desired waveform independent of bandwidth. Theoretical results for common signal envelopes are discussed and fabricated results on YZ-LiNbO 3 are presented. By properly designing the chip overlap an arbitrary envelope and weight can be achieved. These techniques are used to design a new type of orthogonal frequency coded (OFC) SAW sensor. The updated coding technique decreases device response length while preserving code diversity and bandwidth as compared to a traditional, serial OFC layout. Results of these devices in a multi-sensor system show the ability to simultaneously track temperature and range, with 4 concurrent devices tracked at distances between 5-14 meters. Single sensor operation at a range of 14 meters is highlighted with a range determination accuracy of 0.25 meters (at +/3°C).

Nonlinearity distortion measurements and accurate modeling techniques for surface acoustic wave (SAW) devices are desired for third generation (3G) and fourth generation (4G) telecommunication systems. In this paper, a novel nonlinear Mason equivalent circuit model including a 3rd order nonlinear coefficient in the wave propagation is presented. The model is able to accurately simulate both small signal (S parameters) and large signal (3rd harmonic and 3rd order intermodulation) responses of a SAW duplexer. Harmonic and intermodulation measurement techniques are also discussed.

This paper presents the development of sampled Gaussian noise transducers for surface acoustic wave (SAW) correlation filters for use in ultra-wideband (UWB) communications and tagging. The orthogonal frequency coded (OFC) SAW correlator concept was previously demonstrated using a uniformly weighted OFC coded dispersive transducer in conjunction with a wideband apodized transducer. The interdigital SAW transducer can accurately represent the noise-signal by electrode apodization. For an ideal-Gaussian white noise signal, there are no correlation sidelobes and cross-correlation of other codes is extremely small. For a finite truncated white-noise Gaussian signal, that is also frequency band-limited, the auto-correlation sidelobes are produced, a function of the signal time bandwidth product. The implementation of noise-like generators and coding can be very useful in UWB systems. The UWB noise transducer used as a generator, can provide greater resistance to jamming and allows an UWB communication system to coexist with other systems. Code generation is performed in the transmitter using the noise weighted SAW filter and is correlated at the receiver using a matching filter device; eliminating the need for complex signal processing. This can lead to a very robust and simple short-range UWB communication system. The theoretical foundation for the signal analysis and transducer implementation will be presented. Coupling of mode (COM) theory will be used to show the important transducer design parameters for transducer evaluation. SAW correlators with fractional bandwidth of greater than 25% are fabricated on lithium niobate (LiNbO3) having a center frequency of 250 MHz. Discussion of the transducer design, analysis and measurements are presented. Results are shown for operation in a matched filter correlator for use in an UWB communication system and compared to predictions, showing good results.

Surface acoustic wave (SAW) technology, when paired with orthogonal frequency coding (OFC), has proven to be a versatile platform for the design of passive, wireless sensors. This paper presents the design and demonstration of a passive, wireless SAW strain sensor that is uniquely identifiable in a multi-sensor system. A cantilever fabricated out of the SAW substrate acts as the strain sensing mechanism. Force applied at the end of the cantilever causes a strain distribution at the root of the beam. In turn, a measurable change in the SAW propagation delay occurs between the SAW transducer and OFC reflector bank. Relating SAW propagation delay changes to strain is accomplished by developing a simple 1-D model. This model assumes strain in only the Z direction (YZ-LiNbO 3 ) which allows the stiffness matrix to be ignored and an effective stiffness constant to be used instead. Derivations for strain coefficient as well as the strain and force equations needed to measure strain are given. The sensor test setup and experimental results are detailed and discussed. Additionally, a magnetic field sensor is demonstrated as an application for this design.

There has been little published data on spread spectrum SAW RFID correlator receiver performance, since most approaches published have used an FMCW system. The purpose of this paper is to discuss issues related specifically to correlator receiver performance parameters with respect to range, detection, and noise. The minimum detectable signal (MDS) at the ADC is used as the measure for prediction of the maximum range. The loop gain of the system, MDS, noise and processing gain bound the predicted achievable range for a correlator receiver. It will be shown that in a correlator receiver nano- to micro-joules of energy can obtain ranges greater than 50 meters, dependent on the key system parameters. A model is developed for prediction of the maximum range as a function of center frequency, output signal and power, MDS, synchronous interrogation, and loop gain. From the predictions, it appears feasible to have a sensor range of over 100 meters with modest interrogation energy in a pulsed correlator system.