CAREER: Towards Scalable, Low-Power, Wide Area Networks
Synopsis :
Wireless data delivery is crucial for real-time data collection and analysis in the fast-growing areas of smart agriculture, livestock monitoring, and precision farming among others. Such large-scale monitoring networks currently use cellular and satellite networks to collect data. Due to the high infrastructure costs for solar panels and batteries, such networks are deployed only at limited locations and have delays ranging from a day up to a month. Real-time data collection will provide an understanding of the spatio-temporal dynamics of an environment. Such monitoring systems typically deploy sensors in large fields and require long battery life, long communication range to reduce infrastructure costs, share the medium with many more devices, and low cost. However, they have less stringent demands on communication speed and delay. Existing strategies for addressing long battery life and long-range communication do not co-exist efficiently in a large-scale network. Technologies that allow long-range communication typically drain the battery energy faster, while battery-efficient communication technologies are limited to short distances. In this work, novel communication algorithms, protocols, and architectures are proposed to simultaneously achieve long-range communication and long battery life in large-scale deployments. The outcomes of this work will have a direct impact on the Wisconsin farming community through real-time soil health monitoring. It will impact high-school, undergraduate, and graduate education through community-driven wireless courses incorporating prototypes developed in this work.
The proposed research takes the following three directions to overcome the barriers for scalability and meet the connectivity needs of long-range, low-power, wide-area networks (LPWAN): 1) A novel modulation technique that minimizes energy consumption by encoding information in the time interval between transmissions is proposed. Error correction and source codes to address the challenges of a timing channel will be studied in this thrust. 2) In the second thrust, interference cancellation and distributed interference avoidance to decode concurrent transmissions from collisions are proposed. Interference cancellation in existing LPWANs such as LoRa, as well as cross-layer protocols to avoid collisions, will be studied in this thrust. Novel time-frequency analysis to decode packet collisions across multiple gateways will also be explored. 3) In the third thrust, a unified cloud-based LPWAN receiver that can receive and demodulate a variety of LPWAN technologies is proposed. Compression algorithms to transmit radio samples to a cloud server will be designed. A cloud-based LPWAN receiver enables us to coherently combine signals from multiple gateways, in turn facilitating cooperative decoding.
Personnel
Bhuvana Krishnaswamy
Muhammad Osama Shahid
Shiza Shakeel
Publications :
Shahid, Muhammad Osama, Millan Philipose, Krishna Chintalapudi, Suman Banerjee, and Bhuvana Krishnaswamy. “Concurrent interference cancellation: Decoding multi-packet collisions in LoRa.” In Proceedings of the 2021 ACM SIGCOMM 2021 Conference, pp. 503-515. 2021.
Sangar, Yaman, Yoganand Biradavolu, and Bhuvana Krishnaswamy. “A novel time-interval based modulation for large-scale, low-power, wide-area-networks.” ACM Transactions on Sensor Networks 18, no. 4 (2022): 1-30.
Code
Concurrent Interference Cancellation : https://github.com/UW-CONNECT/CIC
Educational activities
Updated an existing undergraduate course to introduce wireless specific projects and updated curriculum and student evaluation to be more project oriented in Fall 2022. In spring 2023, Krishnaswamy is teaching a graduate course on advanced wireless communications where students get to use software repository on CIC.
Outreach and other broader impact outcomes
Results from the proposed work on TIM is being used to build a sensor module to be deployed in ag\ricultural fields in Wisconsin. We are currently working with soil scientists to deploy the sensor modules in the field. Our experience in LoRa networks is used to deploy solar harvesting modules across Madison, WI to understand power management in microgrids. LoRa is used as the backhaul network.