Abstract
This paper proposes a novel downlink signal design and optimal uplink scheduling for Wireless Powered Communication Networks (WPCNs). Traditional resource allocation often assumes static channels because simple energy-harvesting users cannot receive dynamic scheduling updates without a dedicated data receiver. We introduce "Power Level Modulation" (PLM), where the Hybrid Access Point (H-AP) embeds uplink scheduling information into the downlink energy signal by varying power levels across subslots. Users determine their optimal uplink transmission windows based on the relative power levels of received signals. We formulate a sum-throughput maximization problem and solve it using convex optimization. Numerical results demonstrate that the proposed PLM scheme significantly outperforms conventional resource allocation methods.
Introduction
Wireless Powered Communication Networks (WPCNs) utilize Radio Frequency (RF) energy harvesting to power remote devices, which then transmit data back to a Hybrid Access Point (H-AP) using the "harvest-then-transmit" protocol. To maximize network efficiency, the H-AP must dynamically schedule uplink transmissions based on varying channel conditions. However, ultra-low-power IoT sensors often lack downlink data receivers to minimize circuitry and power consumption, creating a communication gap for scheduling commands.
This research addresses this gap by proposing Power Level Modulation (PLM), a technique that leverages the energy signal itself as a low-complexity control channel.
Proposed Power Level Modulation Scheme
The PLM scheme divides the downlink energy period into $K$ subslots. Instead of transmitting at a constant power, the H-AP modulates the power level in each subslot according to the desired uplink schedule.
Mechanism and User Extraction
The relative ratio of power levels in the downlink subslots corresponds to the ratio of uplink time allocated to each user. Because path loss is constant across these short subslots, the relative power ratios remain invariant at the receiver. This allows users to calculate their specific uplink time ($\\tau^U_i$) simply by measuring the energy harvested in each subslot, effectively creating a "zero-power" signaling method.
Optimization Framework
The sum-throughput maximization problem is formulated to determine the optimal time and power allocations. Although the initial formulation is non-convex due to coupled variables, the authors apply a logarithmic transformation to convert it into a convex optimization problem. This ensures that a global optimum can be found efficiently, even as the number of users increases.
Results and Discussion
The performance of the PLM scheme was validated through Monte Carlo simulations. The results indicate that PLM provides substantial throughput gains over conventional Equal Resource Allocation (ERA), particularly in networks with many users or high average transmit power.
A key practical consideration explored is the dynamic range of the power levels. While high dynamic range allows for more precise scheduling, it may exceed hardware peak-power limits. The study shows that a reduced dynamic range can still achieve significant performance gains while remaining compatible with realistic hardware constraints.
Conclusion
Power Level Modulation offers a robust solution for dynamic scheduling in WPCNs without requiring complex downlink receivers. By embedding control information in the energy signal, the proposed scheme nearly doubles throughput in dense sensor networks. Future research will focus on the impact of energy measurement noise and the extension of PLM to multi-antenna and mobile environments.
Summary updated