The design of wireless networks has historically been driven by a single imperative: maximizing throughput. To meet the exponential growth in mobile data traffic, traditional architectures have scaled by densifying base stations and increasing transmission power. However, as highlighted by Stefano Buzzi et al. in A Survey of Energy-Efficient Techniques for 5G Networks and Challenges Ahead, continuing down this throughput-centric path leads to unsustainable energy consumption, escalating operational costs, and an unacceptable environmental footprint.
To prevent a telecommunications energy crisis, the field is undergoing a fundamental paradigm shift. Instead of treating energy as an infinite resource to maximize data rates, next-generation networks treat energy as a strict constraint. The goal is to maximize energy efficiency—measured in bits per Joule—rather than raw capacity. Achieving the 10- to 100-fold efficiency improvements required for 6G networks, as outlined by Xiaohu You et al. in Towards 6G wireless communication networks: vision, enabling technologies, and new paradigm shifts, requires rethinking how networks allocate power, route data, and interact with the physical environment.
Cooperative Transmissions and Joint Power Allocation
In a traditional cellular network, base stations operate as isolated islands. Each station blasts power to reach its designated users, often creating inter-cell interference that forces neighboring stations to increase their power even further. This competitive approach is highly inefficient.
Cooperative communication transforms this dynamic by allowing multiple base stations or nodes to coordinate their transmissions. By treating the network as a joint optimization problem, systems can eliminate redundant transmissions and minimize total power consumption. For instance, Qimei Cui et al. developed a closed-form mathematical solution in Closed-form solution for minimizing power consumption in coordinated transmissions that allows coordinated base stations to cooperatively allocate power across subcarriers. Instead of relying on traditional, localized "water-filling" algorithms, their approach ensures that data rate requirements are met globally with the absolute minimum aggregate power expenditure.
Similarly, Xueqing Huang et al. demonstrated in Joint Spectrum and Power Allocation for Multi-Node Cooperative Wireless Systems that jointly optimizing user association, spectrum allocation, and power levels analytically limits the number of base stations a single user needs to connect to, drastically reducing the operational overhead of cooperative networks. To handle these highly dynamic environments without needing perfect future channel state information, online learning frameworks—such as the "no-regret" learning algorithms proposed by Panayotis Mertikopoulos et al. in Learning to be green: robust energy efficiency maximization in dynamic MIMO-OFDM systems—allow transceivers to adapt their policies "on the fly" to maximize energy efficiency despite measurement noise and user mobility.
Proactive Caching: Trading Storage for Transmission Energy
Wireless channel quality fluctuates constantly due to mobility and environmental fading. Transmitting data when a channel is in a "deep fade" requires a massive expenditure of power to maintain connection quality. Proactive edge caching solves this problem by exploiting a fundamental trade-off: memory is cheap and energy-efficient, while real-time wireless transmission over poor channels is expensive.
Instead of pulling content through power-hungry backhaul networks every time a user requests it, proactive caching architectures store popular content directly at the network edge (e.g., at base stations or local routers).
Traditional Reactive Delivery (High Energy Cost):
User Request ──> Base Station ──[Backhaul Network]──> Core Cloud (High Power)
User <──[Poor Channel / High Power]── Base Station <── Content Delivered
Proactive Caching (Low Energy Cost):
Off-Peak / Good Channel: Core Cloud ──> Base Station Storage (Cached)
User Request ──> Base Station Storage ──[Short Distance / Low Power]──> User
By utilizing stochastic geometry to model dense network deployments, Bhanukiran Perabathini et al. provided a rigorous mathematical proof in Caching at the Edge: a Green Perspective for 5G Networks demonstrating that local storage significantly reduces area power consumption under realistic spatial conditions.
Furthermore, this strategy can be optimized dynamically based on content lifetimes. In Energy-Efficient Wireless Content Delivery with Proactive Caching, András György 0001 et al. introduced the Longest lifetime In–Shortest lifetime Out (LISO) policy. The intuition behind LISO is to "pre-download" content during periods of exceptionally high-quality channel conditions before the user explicitly requests it. By matching transmission bursts to peak channel states, the network avoids the energy spikes associated with real-time, reactive delivery over degraded channels.
Reconfigurable Intelligent Surfaces: Reshaping the Channel Passively
When a wireless signal travels from a transmitter to a receiver, it bounces off walls, buildings, and obstacles, leading to destructive interference and path loss. Historically, networks overcame this by deploying active relays or small cells to intercept, amplify, and retransmit the signal. However, active hardware requires dedicated power supplies, RF chains, and amplifiers, which increases the network's overall energy footprint.
Reconfigurable Intelligent Surfaces (RIS)—also conceptualized as passive intelligent mirrors—offer a way to control the wireless environment without injecting additional active power. An RIS is an artificial metasurface embedded with hundreds of low-cost, passive scattering elements. Each element can independently alter the phase and amplitude of the incident electromagnetic wave.
Active Relay (High Power):
Transmitter ───> [Active RF Electronics / Power Supply] ───> Receiver
Passive RIS (Near-Zero Power):
Transmitter ───> [Passive Phase-Shifting Elements] ───> Receiver
(Constructive Reflection)
As demonstrated by Chongwen Huang et al. in Achievable Rate maximization by Passive Intelligent Mirrors, by intelligently aligning the phase shifts of the reflecting elements, the surface can focus the scattered signals directly onto the receiver, creating constructive interference. This dramatically boosts the achievable data rate and signal-to-noise ratio (SNR) without requiring any active RF power at the mirror.
The standardization and deployment scenarios for this technology, detailed by Ruiqi Liu et al. in Sustainable Wireless Networks via Reconfigurable Intelligent Surfaces (RISs): Overview of the ETSI ISG RIS, emphasize its role as a highly cost-effective and green alternative to traditional infrastructure scaling. Whether deployed on building facades or integrated into complex air-to-ground drone communications—as explored by Yuanyuan Yao et al. in UAV Trajectory and Energy Efficiency Optimization in RIS-Assisted Multi-User Air-to-Ground Communications Networks—RIS acts as a passive lens that optimizes spatial signal distribution, proving that software-controlled environments can substitute for raw power consumption.
Incremental and Hybrid Transmission Protocols
Energy efficiency can also be engineered directly into link-layer protocols and hybrid physical mediums. Traditional error-correction protocols often transmit data with a high power margin to guarantee delivery on the first attempt. In contrast, energy-centric design favors an incremental, adaptive approach.
In Green Communication via Power-Optimized HARQ Protocols, Behrooz Makki et al. analyzed a "gambling" strategy using Hybrid Automatic Repeat Request (HARQ) protocols. Instead of blasting data at maximum power initially, the transmitter starts with a low-power attempt. If the receiver fails to decode it, it stores the received samples and requests a retransmission. The transmitter then incrementally ramps up its power across successive rounds. Because the receiver combines the energy from all attempts, the average power consumed across many packets drops significantly—achieving up to 11 dB in energy savings while still meeting strict reliability and delay constraints.
On a broader architectural level, energy consumption can be mitigated by offloading traffic from the radio frequency (RF) spectrum entirely when alternative mediums are available. For indoor environments, Abdallah Khreishah et al. proposed a co-designed architecture in A Hybrid RF-VLC System for Energy Efficient Wireless Access that combines Wi-Fi with Visible Light Communication (VLC). By routing data through existing LED illumination infrastructure when lights are already active, the network can shut down or idle power-hungry RF transceivers, demonstrating that sustainability is ultimately achieved through a combination of spatial, temporal, and physical-layer adaptability.
Key Foundational Work
Closed-form solution for minimizing power consumption in coordinated transmissions (2012) — contributed 1 concept elements to this direction.
Editorial for the Special Issue: Green Cognitive and Cooperative Communication and Networking (2013) — contributed 1 concept elements to this direction.
Model-driven energy-aware rate adaptation (2013) — contributed 1 concept elements to this direction.