Wireless network design is traditionally a quest for speed. Engineers measure success through Spectral Efficiency (SE)—the number of bits per second that can be squeezed into a single Hertz of bandwidth. However, as billions of battery-constrained devices join the network, a second metric has become equally vital: Energy Efficiency (EE), or the number of bits transmitted per Joule of energy.
The central challenge in modern communications is that these two goals are often in direct conflict. Maximizing SE usually requires cranking up transmit power, but as Jalal Jalali and Hina Tabassum demonstrate in Placement, Orientation, and Resource Allocation Optimization for Cell-Free OIRS-aided OWC Network, the relationship between power and efficiency is not linear. Instead, EE typically follows a bell-shaped curve. Beyond a certain "sweet spot," increasing power yields diminishing returns in data rate while rapidly draining batteries.
The Less is More Principle
Intuition suggests that high-speed modulation (like 64-QAM) is always better because it moves data faster. However, Farhad Meshkati and Andrea J. Goldsmith identified a "less is more" principle in A Game-Theoretic Approach to Energy-Efficient Modulation in CDMA Networks with Delay QoS Constraints. They found that to maximize bits-per-joule, users should actually use the lowest possible modulation order (such as QPSK) that still satisfies their delay requirements.
This works because higher-order modulations require a disproportionately higher signal-to-noise ratio to maintain reliability. By staying at a lower "gear," a device can operate much more efficiently, even if it takes slightly longer to send the same amount of data.
Architectural Solutions for the Tradeoff
To move the entire bell curve upward—improving both SE and EE simultaneously—researchers are looking toward new physical architectures.
- Intelligent Surfaces: Tianwei Hou and Yuanwei Liu explore Reconfigurable Intelligent Surfaces (RIS) in Reconfigurable Intelligent Surface Aided NOMA Networks. These "smart mirrors" steer wireless signals toward users, reducing the power a base station needs to blast in all directions to ensure coverage.
- Massive MIMO: By using hundreds of antennas, base stations can focus energy into narrow beams. X. Wang shows in Spectral Efficiency of the Cellular Two-Way Relaying with Large Antenna Arrays that this "beamforming" allows mobile users to drastically reduce their transmit power without losing signal quality, effectively decoupling high throughput from high energy consumption.
- Pinching-Antenna Systems (PASS): Hardware flexibility also plays a role. Guangyu Zhu and Naofal Al-Dhahir found that switching between different antenna waveguide modes creates a hard toggle: "Waveguide Switching" maximizes energy efficiency, while "Waveguide Multiplexing" pushes spectral efficiency to its limit, as detailed in SE-EE Tradeoff in Pinching-Antenna Systems: Waveguide Multiplexing or Waveguide Switching?.
Recycling Energy with SWIPT
One of the most radical approaches to the EE problem is to treat the communication signal itself as a power source. In Cooperative Non-Orthogonal Multiple Access with Simultaneous Wireless Information and Power Transfer, Yuanwei Liu and Zhiguo Ding 0001 introduce a protocol where "near" users harvest energy from the base station's signal (SWIPT) and use that harvested energy to relay data to "far" users. This turns the energy-intensive task of relaying into a self-sustaining process, boosting the network's overall reach without additional battery drain.
Algorithmic Balancing
Finding the optimal point on the SE-EE curve is a complex optimization problem. Yanxiang Jiang and Xiaohu You 0001 use non-cooperative game theory in Energy Efficient Non-Cooperative Power Control in Small Cell Networks to allow base stations to reach a "Nash Equilibrium"—a state where every node is operating at its most energy-efficient power level given the interference from others.
More recently, researchers like Xianzhong Xie have turned to machine learning. In User selection and dynamic power allocation in the SWIPT-NOMA relay system, they use Deep Reinforcement Learning to dynamically allocate power in real-time, allowing the network to adapt its efficiency strategy as users move and data demands shift.