Enhancing Information Freshness in Energy-Harvesting Sensors: Advanced Optimization, Analysis, and Scheduling Strategies
Open Access DepositedDownloadable Content
In Internet of Things (IoT) networks, particularly in remote sensing applications,maintaining the freshness of status updates is crucial for ensuring timely and accurate monitoring. The inherent challenge lies in efficiently delivering information from a source to a monitoring receiver while operating under stringent resource constraints. Energy limitations at the source or receiver can significantly impact the timeliness of updates, making it essential to develop energy-efficient strategies for minimizing the Age of Information (AoI). This dissertation is motivated by the growing need for realtime status updating in energy-harvesting (EH) systems and focuses on optimizing AoI under energy constraints. Specifically, we analyze and design optimization frameworks for various EH-based status updating systems, with a particular emphasis on energy harvesting communication systems. First, we investigate the problem from the receiver’s perspective by considering an energy-harvesting receiver that incrementally recharges its battery while receiving updates from a single information source over discrete time slots. The receiver dynamically determines its ON-OFF state, where remaining OFF saves energy at the cost of discarding incoming updates, while switching ON enables reception at a deterministic energy cost per slot. We explore power-down policies, formulating age-threshold-based ONOFF schemes to optimize information freshness while meeting the energy causality constraints. The proposed schemes adapt the ON-OFF thresholds based on the source’s update generation behavior—whether operating in a prolific mode or a lazy mode with different update generation frequency. We analyze both single-unit and unlimited battery cases, deriving optimal policies that minimize average AoI. Next, we shift focus to the transmitter side, where an energy-harvesting source is responsible for both Sensing/Computation (S/C) and Transmission (Tx) tasks. In such a system, energy must be allocated strategically between generating new updates and transmitting previously generated updates. We study this intermittent status updating problem considering the scenario where an EH source with on-demand update generation transmitting over a memoryless Gilbert-Elliott (GE) erasure channel, where it must decide whether to generate a new update or transmit an existing one. We analyze the impact of different status updating policies, including age-threshold-based, window-based, and probabilistic retransmission schemes, deriving closed-form expressions for the average peak AoI (PAoI). Depending on the availability of channel state information (CSI), we demonstrate that threshold-based AoI optimization and window-based retransmission schemes outperform alternative approaches in minimizing information staleness. Finally, we examine remote sensing scenarios where update messages consist of multiple data packets. We study information freshness in an energy-harvesting sensor that broadcasts large update messages, each comprising K packets, to multiple users with heterogeneous channel erasure probabilities. We derive closed-form expressions for the average AoI under two transmission schemes: (i) retransmitting each packet until successful reception by all users, and (ii) employing Random Linear Network Coding (RLNC) to encode and transmit K packets collectively. Our analysis accounts for the computational energy required for RLNC encoding. We show that RLNC-based transmission significantly reduces AoI compared to uncoded retransmissions, provided the computational overhead remains within the sensor’s energy budget. Overall, this dissertation develops novel optimization methodologies for minimizing AoI in energy-harvesting sensor networks. By analyzing both receiver-side and transmitter-side constraints, as well as multi-packet update transmissions in broadcast settings, we provide a comprehensive framework for enhancing information freshness in resource-constrained IoT systems. Our findings contribute to the design of more efficient status updating mechanisms, improving the real-time performance of EH-powered networks in various application domains.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.