Optimizing the Harvest Then Transmit Time Split

45 concepts40 papers1 proposal

In wireless networks composed of battery-free Internet of Things (IoT) sensors, medical implants, or remote relays, energy is the ultimate currency. Because these devices lack a permanent internal power supply, they must extract energy directly from ambient or dedicated radio frequency (RF) signals. This operational constraint introduces a fundamental scheduling conflict: a single device cannot typically harvest energy and actively transmit data over the same frequency at the same time.

To resolve this, networks rely on a time-switching protocol known as the harvest-then-transmit framework. Within a fixed communication block, a node must divide its time into two distinct phases: a harvesting interval and a transmission interval. This creates a zero-sum game of time allocation where maximizing throughput requires finding a precise mathematical "sweet spot."

The Linear-Logarithmic Asymmetry

To understand why an optimal time split exists, we must look at how data throughput behaves at the extremes.

Imagine a communication block of total duration $T$. If a device allocates a fraction $\alpha$ of this block to energy harvesting, the remaining fraction $(1-\alpha)$ is reserved for data transmission.

Total Time Block (T)
├───────────────────────────────┴───────────────────────────────┤
│    Harvesting Phase (α)       │     Transmission Phase (1-α)  │
│    (Accumulates Energy)       │     (Sends Data Bits)         │
└───────────────────────────────┴───────────────────────────────┘

If the device chooses either extreme, the system fails completely:

  • $\alpha = 0$ (All Transmission): The device spends no time harvesting, leaving it with zero operating power. No data can be sent.
  • $\alpha = 1$ (All Harvesting): The device accumulates maximum energy, but the remaining transmission window shrinks to zero. No data can be sent.

Between these two failures lies a strictly concave throughput curve. The shape of this curve is dictated by a fundamental physical asymmetry between time and power. As noted by Meixia Tao et al. in Utility-Based Wireless Resource Allocation for Variable Rate Transmission, instantaneous transmission rates scale linearly with allocated time but only logarithmically with transmit power ($\log(1 + \text{SNR})$).

When a node increases its harvesting time $\alpha$, the total energy collected grows linearly. Because this energy must be entirely spent during the remaining transmission window, the available transmit power spikes aggressively as $\alpha$ approaches 1. However, because the data rate only benefits logarithmically from this power boost, it cannot compensate for the linear collapse of the remaining communication window $(1-\alpha)$.

This interplay produces a predictable bell-shaped curve. As demonstrated in Wireless-Powered Communications: Performance Analysis and Optimization by Xiaoming Chen 0001 et al. and in On the Performance Optimization of Two-way Hybrid VLC/RF based IoT System over Cellular Spectrum by Sutanu Ghosh et al., this mathematical concavity guarantees a single, globally optimal time-switching coefficient $\alpha^*$ that maximizes throughput.

Compounding Complexity in Relays and Multi-User Networks

The trade-off becomes significantly more intricate when moving from a single point-to-point link to multi-hop relay networks or multi-user environments.

In a decode-and-forward relay system, the relay node does not just need energy to transmit; it also consumes energy merely decoding the incoming signal from the source. Yanxin Yao demonstrates in Optimizing Energy Harvesting Decode-and-Forward Relays With Decoding Energy Costs and Energy Storage that ignoring these internal processing costs leads to severe overestimations of network capacity.

Furthermore, because a relay system splits the transmission phase into two distinct hops (source-to-relay and relay-to-destination), any time dedicated to harvesting squeezes both communication windows. As shown by Caijun Zhong et al. in Optimum Wirelessly Powered Relaying and Yazid M. Khattabi et al. in On the Performance of Interference-Based Energy-Harvesting-Enabled Wireless AF Relaying Communication Systems, the optimal time split must dynamically adapt to the physical location of the relay. If the relay is close to the power source, it can harvest energy rapidly, allowing $\alpha$ to shrink and leaving more time for data delivery.

Hardware imperfections also warp this optimization landscape. For instance, Behrooz Makki reveals in On joint energy and information transfer in relay networks with an imperfect power amplifier that non-linearities in real-world power amplifiers cause a two-stage behavior: at low source powers, the optimal time split remains rigid, but it shifts aggressively toward data transmission once the source power crosses a threshold that saturates the amplifier.

When scaled to multi-user networks utilizing Non-Orthogonal Multiple Access (NOMA), the time split becomes coupled with power-domain multiplexing and beamforming vectors, yielding highly non-convex optimization problems that require distributed algorithmic solutions, such as those proposed by Zhongyu Wang et al. in Energy-Efficient Resource Allocation in Massive MIMO-NOMA Networks With Wireless Power Transfer: A Distributed ADMM Approach and Zhou Ni et al. in Analysis of RF Energy Harvesting in Uplink-NOMA IoT-based Network.

Network-Wide Side Effects: Outages and Collisions

When an entire network of devices independently optimizes its harvest-then-transmit ratio, severe macro-level bottlenecks can emerge.

The primary culprit is temporal compression. If a cluster of sensors experiences weak RF signals, every device must increase its harvesting coefficient $\alpha$ to accumulate enough operating voltage. Consequently, the remaining window for data transmission $(1-\alpha)$ is compressed into a narrow sliver of time.

As Kaiqiang Lin details in Toward Sustainable Subterranean mMTC: Space-Air-Ground-Underground Networks Powered by LoRaWAN and Wireless Energy Transfer, forcing a dense deployment of sensors to transmit within a highly compressed timeframe causes a dramatic spike in packet collisions. Thus, an individual node's attempt to achieve energy sustainability can inadvertently trigger a network-wide throughput collapse.

This temporal compression is exacerbated when networks must support security protocols. For example, when a swarm of unmanned aerial vehicles (UAVs) acts as both data relays and friendly jammers to block eavesdroppers, they require massive amounts of energy to power active interference signals. As explored by Hanh Dang-Ngoc et al. in Secure Swarm UAV-Assisted Communications With Cooperative Friendly Jamming and Lifeng Wang et al. in Secure D2D Communication in Large-Scale Cognitive Cellular Networks: A Wireless Power Transfer Model, this dual-energy requirement forces the optimal harvesting duration to widen significantly, leaving an even narrower margin of error for successful data delivery.

Bypassing the Trade-off with Ambient Backscatter

To break free from the strict zero-sum constraint of harvest-then-transmit scheduling, modern architectures often introduce a third operational mode: ambient backscatter communication.

Instead of generating an active RF signal—which demands significant power and dedicated transmission time—a device can communicate by cleanly reflecting or absorbing an ongoing primary signal (such as a nearby TV or Wi-Fi broadcast). This transforms the bimodal time split into a trimodal optimization problem: a device can harvest energy, backscatter data immediately, or save its energy for active transmission later.

Trimodal Hybrid Scheduling
┌───────────────────────────────┬───────────────────────────────┐
│     Primary Channel Busy      │     Primary Channel Silent    │
├───────────────┬───────────────┼───────────────────────────────┤
│ Harvesting    │ Ambient       │ Active Transmission           │
│ Phase         │ Backscatter   │ Phase                         │
└───────────────┴───────────────┴───────────────────────────────┘

This hybrid approach solves a notorious bottleneck in cognitive radio networks. As explained by Dinh Thai Hoang et al. in The Tradeoff Analysis in RF-Powered Backscatter Cognitive Radio Networks and Wenbo Wang et al. in Stackelberg Game for Distributed Time Scheduling in RF-Powered Backscatter Cognitive Radio Networks, when a primary transmitter is highly active, a standard harvest-then-transmit device accumulates plenty of energy but suffers from massive latency because it cannot find a silent interval to transmit actively.

By allowing nodes to backscatter data while the primary channel is busy, the network maintains a steady data flow. Because backscattering and harvesting are mutually exclusive during these busy periods, devices must dynamically adjust their schedules. Using online reinforcement learning frameworks, such as those developed by Nguyễn Văn Huynh et al. in Optimal and Low-Complexity Dynamic Spectrum Access for RF-Powered Ambient Backscatter System with Online Reinforcement Learning and Kevin Shen Hoong Ong et al. in Cognitive Radio Network Throughput Maximization with Deep Reinforcement Learning, nodes can continuously adapt their time-switching ratios to match fluctuating channel conditions, maximizing throughput without requiring prior environmental knowledge.

Key Foundational Work

Simultaneous Wireless Information and Power Transfer Under Different CSI Acquisition Schemes (2014) — contributed 1 concept elements to this direction.
Utility-Based Wireless Resource Allocation for Variable Rate Transmission (2007) — contributed 1 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_time_switching_resource_allocation_03f817a4,
  title = {time-switching resource allocation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/time-switching-resource-allocation/rdr_03f817a419a749db9e7d84af6d435c95},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}

This exact version (version 1, frozen text):

@misc{lacuna_time_switching_resource_allocation_03f817a4_v1,
  title = {time-switching resource allocation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/time-switching-resource-allocation/rdr_03f817a419a749db9e7d84af6d435c95/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake