Lazy is timely: Status updates by an energy harvesting source

TLDR

A source submits status updates to a service facility for delivery to a monitor, and each update consumes energy from a stochastic energy harvesting system. We evaluate updating policies for a source with a large battery using a status age timeliness metric. The source uses knowledge of the service facility state to avoid queue delays by sending a new update only after the previous one completes. An optimal policy is lazy, leaving the service facility idle after completion even when energy is available.

Abstract

A source submits status updates to a service facility for delivery to a monitor. Each update requires energy and the source is powered by a stochastic energy harvesting system. With knowledge of the service facility state, the source avoids queue-induced delays by submitting a fresh update only after the service completion of a prior update. For a source with a large battery, we evaluate updating policies using a status age timeliness metric. We show that an optimal policy is lazy; following a service completion, the service facility is frequently left idle even though the server may have sufficient energy to submit an update.