NSF CAREER (4/15-3/20)


Project Description

IDEA: Integrated Data and Energy Access for Wireless Sensor Networks (April 2015 - March 2020)

  • Client:

    Project Sponsor- NSF
The proposed Integrated Data and Energy Access (IDEA) framework will help realize the promise of jointly powering and communicating with wireless sensors with the help of electromagnetic radio frequency (RF) waves. This research will result in a cross-layer protocol suite, modeling tools, and experimental systems that will be instrumental in building future RF energy harvesting networks for healthcare, wearable electronics, smart homes, and embedded implants with extended lifetimes. The controllable and predictable nature of RF energy and the lack of dependence on environmental conditions distinguish this form of energy harvesting over other renewable sources. IDEA will directly support national and institutional level energy-efficiency related program goals by reducing dependence on batteries and enhancing sustainability in networking. The project involves an integrated education plan involving demonstrations at public museums, pedagogical tools for K12 students, a summer school focused on RF energy harvesting, continuing education for industry personnel, along with dedicated outreach efforts to attract students from under-represented groups.

Current Personnel: 
Yousof Naderi
Ufuk Muncuk
Kai Li
Kubra Alemdar

Previous Personnel:
Raul Cid-Fuentes
Luiza Pelucio
Zachary Deschaux
Kunal Sankhe
Subhramoy Mohanti

Project Objectives:
  1. We will research methods for sharing the channel for the charging and data communication functions in the network, while addressing the challenges posed by phase variance and the constructive and destructive interference of energy signals in space.
  2. We will propose new routing metrics and protocols for creating energy paths, which will charge only selected sensors to ensure data delivery to the sink occurs timely over these pathways. It will yield concrete insights on the relationship between the number of ETs, their placement, the network lifetime, and the data communication capacity given the energy transfer requests within the network.
  3. We will answer fundamental questions on how much energy can be obtained through indoor ETs, as well as outdoors from the ambient RF environment in the TV band, leading to the creation of spatial-temporal energy maps.
  4. We will transform the concept of RF harvesting sensors to a platform for network deployment through a testbed of software­-defined radio ETs and circuit designs for RF harvesting boards.

Key Intellectual Merits:


1. Software-Defined Wireless Energy Beamforming:
We developed and demonstrated a software-defined solution for wirelessly charging the sensors using RF energy. Here, the actions of more than one energy transmitter (ET) were synchronized in phase and frequency in realtime using periodic feedback from the target sensor, but without any common clock reference. The controller selected the optimal subset of ETs to satisfy the energy request from a given sensor, which cooperatively beamform RF energy towards that sensor.

Fig. 1. Prototyping distributed energy beamfomring with USRPs.

Our software-defined framework, implemented in Python, allowed the central controller to automatically discover the installed sensors, obtain energy needs, and schedule charging tasks in an asynchronous and non- blocking manner that allowed the network to scale. The demonstration included the design and fabrication of RF energy harvesting circuits that interface with the TIEZ430 sensors, implementation of a software-defined control and data plane, as well as a real-time distributed beamforming algorithm on USRP radios that results in a battery-free network of sensors.

Fig. 2. Architecture of software-defined energy beamforming.

Furthermore, we developed a network architecture called HYDRA (from Harvesting energY from Direct RF and Ambient sources), a planning and network resource management framework for powering small form factor sensor nodes. HYDRA uses distributed ad hoc beamforming-capable energy transmitters (ETs) and also leverages ambient energy from cellular and TV spectrum bands (Fig. 3(a)).

Fig. 3(a) Network architecture of HYDRA where sensors are powered by joint ad­hoc beamforming from multiple energy transmitters (ETs) and ambient energy sources.

Fig. 3(b) Average received power density at various locations in the Boston area, from ambient sources as well as when the maximum energy harvesting yield source is intelligently chosen through the so called 'cognitive EH' approach.

Through modeling optimal radiating power levels and phases of the ETs, HYDRA determines the spatial scheduling of the energy beams mapping ET operations to changing network requirements and available ambient energy, while minimizing the energy expenditures of ETs. Through actual in-field experimentation, we showed that ambient RF energy harvesting is best performed by adaptively choosing the available spectrum bands, which in turn alters the energy contribution required from the ETs (Fig. 3(b)).

Publications:
U. Muncuk, S. Mohanti, K. Alemdar, M Y. Naderi, K. R Chowdhury, “Software-defined Wireless Charging of Internet of Things using Distributed Beamforming," ACM SenSys 2016 - The ACM Conference on Embedded Networked Sensor Systems, Stanford, CA, USA, November 2016. PDF

K. Chowdhury, and M.Y. Naderi, “Distributed Wireless Charging System and Method", Patent No. 62/308298.


2. Simultaneous Wireless Energy Transfer and Communication:
 When data communication and RF energy recharging occur in-band, sharing the RF medium and devoting separate access times for both operations raises architectural and protocol level challenges. We developed a method of concurrent transmission of data and energy to solve this problem, allowing ETs to transmit energy and sensors to transmit data in the sameband synchronously. Our key idea concerned devising a physical layer modulation scheme that allows the data transmitting node to introduce variations in the envelope of the energy signal at the intended recipient. We implemented a proof-of-concept receiver, modeled and validated through extensive experimentation. We also designed a new physical layer mechanism for guaranteed successful delivery of information in a point-to-point link.

Fig. 4. Energy management functions at the sensor, with the harvested energy budget feeding both the sensing as well as future transmission needs.

Publications:
R. G. Cid-Fuentes, M. Y. Naderi, S. Basagni, K. Chowdhury, A. Cabellos-Aparicio, E. Alarcon, "On Signaling Power: Communications over Wireless Energy", IEEE INFOCOM 2016, San Francisco, CA, USA. PDF

R. G. Cid-Fuentes, M.Y. Naderi, S. Basagni, K. R. Chowdhury, A. Cabellos-Aparicio and E. Alarcón, "An All-Digital Receiver for Low Power, Low Bit-Rate Applications Using Simultaneous Wireless Information and Power Transmission", IEEE ISCAS 2016, Montreal, Canada, May 2016. PDF


3. Wake-Up Receiver & Token Identification using Energy Harvesting:
The existing passive wake-up receivers (WuRxs) are radio frequency identification (RFID) tag based, which incur high cost and complexity. In this activity, we studied cost-effective and long-range WuRx solutions for range-based wake-up (RW) as well as directed wake-up (DW). In particular, we considered the case of an RF energy harvesting wireless sensor node and investigate how a low-cost WuRx can be built using an RF energy harvester available at the node.

Fig. 5. Prototype of EH-based passive wake-up receiver.

In addition, we developed a touch energy harvester for smart interaction with capacitive touch–enabled devices by associating the token’s identity with its contact, or touch. The proposed token’s design features two key novel technical components: (1) a through–touch–sensor low–energy communication method for token identification and (2) a touch–sensor energy harvester design. The communication mechanism involves the token transmitting its identity (ID) directly through the touch–sensor by artificially modifying the effective capacitance between the touch sensor and token surfaces. By enabling the token to harvest energy from touch–screen sensors or touch–surfaces the token is rendered battery–free.

Fig. 6. (a) Touch energy harvester's schematic, (b) PCB design and token’s schematic.

Publications:
K. Kaushik, D. Mishra, S. De, K. R. Chowdhury, and W. Heinzelman, “Low-Cost Wake-Up Receiver for RF Energy Harvesting Wireless Sensor Networks,” IEEE Sensors Journal, vol. 16, no. 16, Aug. 2016. PDF

P. Nguyen, U. Muncuk, A. Ashok, K. R. Chowdhury, M. Gruteser, and T. Vu, “Battery-Free Identification Token for Touch Sensing Devices, ACM Conference on Embedded Networked Sensor Systems (SenSys), Stanford, CA, Nov. 2016. PDF


4. Medium Access for Energy Transmitters:
 We researched, what we call a 'Duty­ cycled random-phase Multiple Access (DRAMA)' scheme, for wireless RF power transmission. Our approach is based not only in handling the interferences of multiple ETs, but also to benefits from them to broaden the input power range of existing energy harvesters at the sensor nodes. For this, DRAMA relies on the fundamental assumption that efficiency is maximized when the input power varies in time as much as possible since the energy harvesters operate with increased efficiency as a function of the input power.

Publications:
R.G. Cid-Fuentes, M.Y. Naderi, R. Doost, K.R. Chowdhury, A. Cabellos-Aparicio, E. Alarcon, "Leveraging Deliberately Generated Interferences for Multi-sensor Wireless RF Power Transmission", in Proc. IEEE GLOBECOM 2015, San Diego, CA, USA, Dec. 2015. PDF


5. Multi-band Ambient RF Energy Harvesting for Self-powered Sensors:
We developed a multi-band adjustable circuit for harvesting from LTE 700MHZ, GSM 850MHZ, and ISM 900MHZ bands with one single circuit. To this end, we designed a tunable impedance matching network composed of off-the-shelf components, such as adjustable capacitors (i.e. trimmers) to adapt the impedance matching network configuration with the selected RF band using a single fabricated ambient RF energy harvesting circuit. Our circuit design is fabricated on the printed circuit board with comprehensive evaluations at each associated frequency to test the power conversion efficiency.

Fig. 7. (a) Overview of our adjustable ambient RF energy harvesting system, (b) Prototype of the circuit.

Additionally, we designed and implemented an RF-EH prototype that can constantly operate TI eZ430-RF2500 sensor at multiple city locations, with efficiencies of up to 45% in LTE 700, GSM 850, and ISM 900 bands. We conducted a comprehensive RF ambient survey study in Boston and demonstrated that more than 65% of the locations have enough RF power density to perpetually operate low-power sensor mote in battery-less operation mode.

Fig. 8. (a) Battery-free communications setup with TI eZ430-RF2500 sensors that are powered with LTE signals, (b) Levels of average ambient harvested power at different Boston locations.

Finally, we designed and developed a self-powered BLE network that operates with our ambient RF energy harvesting module and harvests available ambient energy. We integrated a customized network stack that implements adaptive duty-cycling within BLE nodes based on data rate and available power, and store storage to ensures smooth DC power delivery to BLE beacon mote.

Publications:
U. Muncuk, K. Alemdar, J. D. Sarode and K. R. Chowdhury, "Multi-band Ambient RF Energy Harvesting Circuit Design for Enabling Battery-less Sensors and IoTs," IEEE Internet of Things Journal, vol.5, no: 4, Aug. 2018 PDF


6. Over-the-air frequency and phase synchronization for energy beamforming.
We designed and implemented TX and Rx circuits for real-time over-the-air synchronization called RFClock that is the core enabler of real-time distributed energy beamforming. We designed a Tx circuit that generates a two-tone signal at different frequencies (915 MHz, 925 MHz) separated by desired reference clock frequency (10 MHz). Additionally, we devised the Rx circuit that is capable of extracting the 10 MHz from the two-tone signal transmitted over the air. The received signal is then amplified by using low-noise-amplifier (LNA) without significantly degrading its signal-to-noise ratio. A bandpass filter, filters out the band of the two-tone signal and this filtered signal is divided in two, in order to perform the multiplication step of the signal by itself. The desired reference clock is extracted from the output of the mixer by filtering with a bandpass filter centered at the desired reference frequency. These Tx and Rx over-the-air synchronization circuits enable real-time beamforming with the elimination of on-host frequency and phase synchronizations that add significant time overheads.

Fig. 9. (a) Schematic of RFClock, (b) Prototype of the RFClock.


7. Wearable sensor charging with software-defined magnetic resonance-based transfer.
There has been increased interest in the use of wearable sensors in everyday activities from hearing aids to the smartwatch. We aimed to leverage the existing surfaces as energy transmitter to charge these sensors and implemented a multi-coils system (both Tx and Rx) for charging wearable sensor using magnetic resonance energy transfer. We first built a software-controlled surface with 27-coils in 1-square foot that has only one power management circuit. This powers any inductive compatible wearable sensor as soon as it place of the surface. Then we built a distributed magnetic resonance beamforming system. In particular, we designed and fabricated a multi-coils current sensing circuit based on 1:1 PFD3215 transformer as shown in Figure 10. The current measured would be the one flowing through a power inductor inserted in series with the amplifier; the transformer acts as an insulator, being connected in parallel with the sensing inductor. Our configuration provides a lower sensor’s equivalent impedance and higher resolutions of the real-time current measurement.

Fig. 10. (a) first (b) second versions of magnetic resonance-based sensing circuits.

Furthermore, we built a prototype for beamforming that estimates channels based on the mutual inductance between the coil and accordingly calculates the weights of amplitude and phase for each transmitting coil so the transmitted power is maximized at the wearable sensor as Rx. Finally, we designed and fabricated a small-form-factor Rx based on a stack of coils in the form of multi-layer PCB, and integrated into a wearable sensor.

Fig. 11. COMSOL-based magnetic flux intensity results for (a) beamforming toward the surface center, and (b) beamforming toward the surface side.


8. City-scale WakeUp Radio Design using LTE Signals.
We designed and demonstrated a wakeup radio control plane that allows fine-grained signaling for city-scale IoT deployments without installing any additional infrastructure, and called FreeIoT. We developed a novel encoding scheme that changes the spatial positioning of Almost Blank Subframes (ABS) within a standard LTE frame to convey control information. ABS was originally defined in the standard to allow coexistence between the macro-cell eNB and nearby small cells, which in our system leveraged as a side channel for IoT signaling. In addition, we implemented a session management protocol to maintain contextual information of the control signaling. This allows our system to handle situations where the control message may span multiple frames, or when the LTE operator temporarily reduces the number of ABS. Furthermore, we incorporated an error detection and correction mechanism to counter channel and fading errors. Finally, we demonstrated a proof of concept testbed to validate the operation of our system using a software-defined LTE eNB and custom-designed RF energy harvesting circuit interfaced with TI eZ430-RF2500 sensors.

Fig. 11. (a) LTE network with the eNB and a given small cell, with spectrum sharing through ABS subframes, (b) FreeIoT framework: The control/wake-up information is conveyed by the position of ABS within a standard LTE frame, (c) circuit schematic of FreeIoT receiver.

We validated our wakeup radio design performance through extensive indoor & outdoor experiments, as well as simulations. Through experiments, we show that our approach can support an effective communication rate of up to 375 bps, and yet limit the symbol error rates in the range of 1-6% for typical outdoor deployments. Also, we provided quantitative simulation results to demonstrate the viability of system performance, offering in three city environments a) residential area, b) stadium and c) city downtown each of area 250×250m2 with the different distribution of received input power and location of 200 IoT sensors. We found the lowest throughput rate 180bps at stadium configuration and highest rate 370bps at downtown setting.

Publications:
K. Sankhe, U. Muncuk, M. Y. Naderi, and K. R. Chowdhury, "Talking When No One is Listening: Piggybacking City-scale IoT Control Signals Over LTE,'' IEEE INFOCOM 2018, Hawaii, USA, Apr. 2018. PDF

K. Chowdhury, and M.Y. Naderi, "Method And Apparatus For Overlaying City-Scale IoT Control Signals Over LTE Cellular Networks." Patent No. 62/577,509.


9. WiFi Friendly Energy Delivery with Distributed Beamforming.
High power radiation within the ISM band interferes with the packet reception for existing WiFi devices. We devised the first system that merges the RF energy transfer functions within a standards compliant 802.11 protocol to realize practical and WiFi-friendly Energy Delivery (WiFED). We designed and implemented a software-defined architecture composed of a centralized controller that coordinates the actions of multiple distributed energy transmitters (ETs), and then integrated specific features of 802.11 within sensor control plane for requesting and monitoring energy. Additionally, we devised a controller-driven bipartite matching-based algorithmic solution that assigns the appropriate number of ETs to energy requesting sensors for an efficient energy transfer process. We demonstrated our WiFi-friendly distributed energy transfer through a practical system in software-defined radio testbed and extensive simulations.

Fig. 12. WiFED architecture for energy delivery with distributed beamforming over the existing 802.11ac network.

To characterize the effects of energy transmission in the ISM band of 2.4GHz during 802.11ac WiFi data communication, we created a real testbed. We found that receiver only gets the energy signals at 0.4 Watts through continuous energy beamforming, while the data transmission gets completely interrupted by the higher power energy signal. Additionally, our experiments showed random energy transfer causes the WiFi receiver to attain almost half of the best achievable throughput with around 40% PER. We observed that the throughput achieves near-optimal rates in the absence of any interfering non-cooperating protocol, with negligible PER. Finally, our performance evaluation results in NS3 show 15% improvement in network lifetime and 31% reduction in the charging delay compared to the classical nearest distance-based charging schemes that do not anticipate future energy needs of the sensors and are not designed to co-exist with WiFi systems.

Publications:
S. Mohanti, E. Bozkaya, M. Y. Naderi, B. Canberk and K. R. Chowdhury, "WiFED: WiFi Friendly Energy Delivery with Distributed Beamforming,'' IEEE INFOCOM 2018, Hawaii, USA, Apr. 2018. (best paper award) PDF


Broader Impacts:
We engaged in building an outreach program, where high school freshmen, sophomores, and juniors get early exposure to engineering in a day-long visit to the NU campus. For this event, we have developed an educational game that runs on the Google Nexus-`S' Android phone. Here, the students search for so-called "radio eggs" spread throughout the campus with the phone, which are actually beacon signals transmitted by pre-selected access points. The phone displays an "egg-found" message only when the signal strength is between -65 to -70 dBm. For lower magnitudes of signal strength, the phone hints at a nearby presence of the egg by changing the color to blue e.g., Bike stand. All identified eggs switch permanently to green with a strike-out. At the end of the activity, students have explained concepts of the exponential fall of signal strength, path loss in the wireless channel, device addressing through beacon exchanges, and factors impacting correct packet reception.

Fig. 13. Various events from interactive demonstrations to educational games.

Additionally, we partnered with Brooke Charter Schools to host 50-60 students for a half day event with interactive demonstrations, presentations, and discussions with faculty and students. This school system has five branches and academic levels up to K-8 with 94% of the students in attendance being of African-American and Latino background. Previous demonstrations have included showing wireless signal energy distribution through our RF energy harvesting sensors that are connected to a buzzer. The audio tones indicate how much energy has been harvested at a location, followed by explanations on why tones (and hence, energy levels) differ by location.