Resource Allocation Under Energy Causality Constraints

37 concepts25 papers1 proposal

In traditional wireless communication systems, devices operate under static power constraints. A transmitter might be limited by a maximum peak power or an average power budget over a transmission block, drawing from a steady grid or a pre-charged battery. Under these assumptions, if a device encounters an exceptionally clear channel early in its transmission window, it can choose to blast data at high power, knowing it can balance its energy budget by throttling down later.

However, the rise of Internet of Things (IoT) devices and autonomous sensor networks has shifted the paradigm toward Energy Harvesting (EH) systems. These devices scavenge energy from their surroundings—such as solar radiation, thermal gradients, or stray radio frequency (RF) signals—as reviewed by Elza Erkip et al. in Energy Harvesting Wireless Communications: A Review of Recent Advances. Because environmental energy arrives stochastically and intermittently, these systems must obey a strict physical law: the energy causality constraint.

Energy causality dictates that a node cannot consume energy before it has been harvested. This introduces a fundamental temporal asymmetry. While a traditional device can "borrow" from its future budget by spending power now and recharging later, an energy-harvesting device can only shift energy forward in time. As noted by Shuyan Hu et al. in Modeling and Analysis of Energy Harvesting and Smart Grid-Powered Wireless Communication Networks: A Contemporary Survey, this one-way arrow of time completely alters resource management, requiring new optimization frameworks to maximize throughput and minimize data distortion.

The Mechanism of Directional Water-Filling

To understand how communication systems adapt to this constraint, we must look at how traditional power allocation changes under energy causality.

In standard information theory, maximizing data capacity over a time-varying fading channel relies on Shannon’s water-filling algorithm. Imagine the time slots of a transmission window as adjacent bins, where the bottom of each bin represents the noise level of the channel (lower bottoms mean better channel conditions). To maximize throughput, you pour a fixed volume of "water" (your total energy budget) across all bins. The water naturally flows to equalize its surface level, resulting in more power allocated to high-quality channels and less to poor ones.

Under energy causality, this fluid model requires a crucial modification: directional water-filling. As explained by Yaming Luo et al. in Training Optimization for Energy Harvesting Communication Systems, we must place "one-way valves" or "directional walls" between the time slots.

Standard Water-Filling:
[ Slot 1 ] <---> [ Slot 2 ] <---> [ Slot 3 ]  (Energy flows freely)

Directional Water-Filling:
[ Slot 1 ] ----> [ Slot 2 ] ----> [ Slot 3 ]  (Energy can only flow forward)

Energy harvested at time slot $t$ can flow forward to fill current and future slots ($t, t+1, t+2, \dots$), but it is physically blocked from flowing backward to past slots. If a spectacular channel opportunity occurs in slot 1, but the energy to exploit it does not arrive until slot 3, that energy cannot be pulled back in time.

Consequently, the optimal water level is no longer perfectly flat across the entire horizon. Instead, as demonstrated by Rajeev Gangula et al. in Optimization of Energy Harvesting MISO Communication System with Feedback, the water level forms a staircase structure that can only step up or remain constant as time progresses.

This behavior is further regulated by the physical capacity of the device's battery. Zhe Wang et al. show in Iterative Dynamic Water-filling for Fading Multiple-Access Channels with Energy Harvesting that a finite battery introduces an upper bound on the water level. If a massive burst of energy arrives when the battery is already full, the excess energy overflows and is lost. The optimal allocation strategy must therefore navigate a dual hazard: it must prevent the battery from hitting zero (violating causality) while ensuring it does not overflow (wasting harvested resources).

Circuit Costs and Bursty Glue Pouring

The basic directional water-filling model assumes that every milliwatt of harvested energy is converted directly into radiated RF signals. In practice, however, a wireless transceiver consumes a fixed baseline "processing tax" just to keep its internal circuitry turned on.

Deniz Gunduz et al. integrated this reality into the causality framework in Throughput Maximization for an Energy Harvesting Communication System with Processing Cost. If an energy-harvesting node transmits continuously at a very low power level, the fixed processing cost will quickly deplete its scarce energy reserves without delivering meaningful data throughput.

To overcome this, the authors introduced the directional glue-pouring algorithm, a concept expanded by Oner Orhan et al. in Energy Harvesting Broadband Communication Systems with Processing Energy Cost.

In this framework, before any energy can be allocated to transmission (the "water"), a solid foundation of "glue" (the fixed processing cost) must be poured into a time slot to activate it. If the harvested energy available up to that point is insufficient to cover both the glue and a meaningful amount of water, the optimal strategy shifts from continuous transmission to bursty communication. The transmitter chooses to stay completely silent during certain epochs, allowing its battery to pool energy over multiple harvesting intervals. Once enough energy accumulates to overcome the circuit tax efficiently, the node turns on and transmits at high power for a short, concentrated burst.

Dual Causality in Multi-Hop Networks

The complexity of managing energy causality intensifies when moving from a single point-to-point link to multi-hop relay networks, where a source node transmits data to an intermediate relay, which then forwards it to the final destination.

When the relay node relies entirely on environmental energy harvesting, the system becomes governed by a dual causality constraint. As formulated by Deniz Gunduz et al. in Two-hop communication with energy harvesting, the relay's transmission schedule is bound by two distinct arrows of time:

  1. Energy Causality: The relay cannot spend more energy than it has harvested up to the current moment.
  2. Data Causality: The relay cannot forward more data bits than it has already received from the source node up to the current moment.

This dual constraint tightly couples the transmission strategies of both nodes. If the source node has abundant energy and floods the relay with data early on, but the relay is currently starved of harvested energy, the data sits uselessly in the relay's buffer, risking delay or buffer overflow. Conversely, if the relay experiences a surge of harvested energy but the source has been starved, the relay's transmission capacity is wasted because it has no data to forward.

To resolve this bottleneck, Yaming Luo et al. developed joint optimization policies in Optimal Scheduling and Power Allocation for Two-Hop Energy Harvesting Communication Systems. By tracking the intersection of the data arrival profile and the energy harvesting profile, these algorithms dynamically adjust the source's transmission rate to match the relay's anticipated energy states. The resulting framework ensures that the "water levels" across both hops are synchronized, preventing energy depletion at the relay while respecting the strict physical boundaries of both data and energy availability.

Key Foundational Work

Optimal Packet Scheduling in an Energy Harvesting Communication System (2012) — contributed 3 concept elements to this direction.
Training Optimization for Energy Harvesting Communication Systems (2012) — contributed 2 concept elements to this direction.
Two-hop Communication with Energy Harvesting (2011) — contributed 2 concept elements to this direction.
History & citationVersion 2 · 2026-10-04

Current · Permanent link

  1. Version 2 (latest) · 2026-10-04 · 21 papers
  2. Version 1 · 2026-10-04 · 32 papers
Cite this direction

Latest version (a living page):

@misc{lacuna_energy_causality_constraint_dabe1f93,
  title = {energy causality constraint},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/energy-causality-constraint/rdr_dabe1f93f743453b9262283fc8a53687},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-22},
}

This exact version (version 2, frozen text):

@misc{lacuna_energy_causality_constraint_dabe1f93_v2,
  title = {energy causality constraint},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/energy-causality-constraint/rdr_dabe1f93f743453b9262283fc8a53687/v/2},
  note = {Version 2, frozen 2026-10-04. Accessed 2026-11-22},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake