In classical wireless communication optimization, energy is often treated as a fluid that can be stored indefinitely in an infinitely deep reservoir. Under this idealized assumption, an energy-harvesting transmitter can save up power during periods of high environmental availability and wait arbitrarily long for perfect channel conditions to execute an optimal transmission strategy.
However, real-world Internet of Things (IoT) sensors, edge devices, and supercapacitors do not possess infinite reservoirs. Their physical storage units introduce hard constraints that fundamentally break classical communication models. To design realistic autonomous systems, researchers must account for two primary physical limitations: energy overflow and energy leakage.
The Geometry of Finite Storage: The Tight String
To understand how a finite battery changes transmission scheduling, we can look at an elegant geometric framework introduced by Bertrand Devillers et al. in [A General Framework for the Optimization of Energy Harvesting Communication Systems with Battery Imperfections].
Imagine plotting the cumulative energy harvested from the environment over time as a continuous curve. In an infinite-battery model, the cumulative energy spent by the transmitter simply cannot cross above this harvesting curve—you cannot spend energy you have not yet captured.
Energy
^
| /-- Upper Boundary (Harvested + Capacity C)
| /
| /------- Lower Boundary (Harvested Energy)
| /
| / <-- Transmission Strategy (The "Tight String")
| /
+------------------------------------------------------------> Time
When the battery has a finite capacity $C$, a second boundary emerges: an upper curve exactly $C$ units above the harvesting curve. If the cumulative energy spent falls below this upper boundary, it means the battery has filled to its maximum capacity. Any subsequent environmental energy arrivals will overflow and be permanently lost.
The authors demonstrate that finding the optimal throughput policy is mathematically equivalent to stretching a "tight string" through the corridor bounded by these two curves. The finite capacity forces the string to bend more frequently. The transmitter cannot simply wait for the best channel conditions; it must consume energy aggressively enough to keep the cumulative consumption curve above the upper boundary, actively preventing energy overflow.
The Saturation Phenomenon: Why Small Batteries Are Good Enough
Given that finite storage restricts our ability to smooth out stochastic energy arrivals, a natural hardware question arises: how large must a battery be to avoid significant performance degradation?
Work by Yishun Dong et al. in [Near Optimal Energy Control and Approximate Capacity of Energy Harvesting Communication] and Dor Shaviv et al. in [Capacity of the Energy Harvesting Channel with a Finite Battery] reveals a surprising and comforting truth: the relationship between battery capacity and channel capacity exhibits a sharp saturation point.
The authors demonstrate that once a battery's capacity is large enough to hold the maximum energy arriving in a single time slot, further increases in battery size yield rapidly diminishing returns. Once the storage can absorb a single typical burst of environmental energy, the system's performance bottleneck shifts entirely. It is no longer limited by storage constraints (the inability to buffer energy), but rather by the fundamental energy arrival rate of the environment itself. Consequently, hardware designers do not need massive, expensive batteries; a modest storage unit matched to the scale of local energy bursts is sufficient to achieve near-optimal data rates.
The Pressure of Leakage and Inefficiency
Real batteries do not just overflow; they also leak. Spontaneous energy dissipation, self-discharge, and charging/discharging conversion losses mean that stored energy effectively has an expiration date.
As explored by Deniz Gunduz et al. in [Throughput Maximization for an Energy Harvesting Communication System with Processing Cost] and Ramachandran Rajesh et al. in [Information capacity of energy harvesting sensor nodes], leakage introduces a temporal discount factor on stored power. If a device saves energy to wait for a better channel state in the future, a fraction of that energy vanishes into thin air while sitting in the buffer. This physical reality shifts the optimal transmission strategy away from smooth, long-term planning toward more immediate, bursty transmissions.
System-Level Consequences
These physical constraints ripple up from the hardware layer to alter higher-layer network protocols and optimization objectives:
- Data Freshness (Age of Information): In applications focused on keeping data fresh, Ahmed Arafa et al. demonstrate in [Timely Status Updating Over Erasure Channels Using an Energy Harvesting Sensor: Single and Multiple Sources] that finite battery constraints prevent simple uniform transmission rates. Instead, optimal policies must be threshold-based. When the battery approaches its finite limit, the sensor must transmit an update immediately—even if the channel is poor—simply to convert endangered energy into a chance of successful communication before it overflows or leaks away.
- Network Saturation: In multi-node networks, finite storage creates a hard ceiling on scalability. As more nodes compete for channel access, individual nodes are forced to buffer energy longer while waiting for a scheduling slot. Because of finite capacities, these extended waiting periods lead directly to widespread battery overflow, causing network throughput to saturate despite an abundance of ambient energy.
Key Foundational Work
Throughput Maximization for an Energy Harvesting Communication System with Processing Cost (2012) — contributed 2 concept elements to this direction.
Training Optimization for Energy Harvesting Communication Systems (2012) — contributed 1 concept elements to this direction.
A General Framework for the Optimization of Energy Harvesting Communication Systems with Battery Imperfections (2011) — contributed 1 concept elements to this direction.