Source-linked AI summary
Energy Efficiency Challenges of 5G Small Cell Networks
Xiaohu Ge, Jing Yang, Hamid Gharavi, Yang Sun
TL;DR
Dense 5G small-cell deployment creates an energy-efficiency question about the relative importance of computation and transmission power. The paper estimates computation power using Landauer’s principle while examining antenna count and bandwidth, finding that computation consumes more than half of small-cell base-station energy. It concludes that both computation and transmission power should be considered in optimization.
Problem
Dense deployment and massive MIMO reduce transmission power but increase computation demands, leaving their relative importance for 5G small-cell energy efficiency unresolved.
Method
The paper models and estimates small-cell base-station computation power using Landauer’s principle while investigating antenna count and bandwidth effects.
Results
More than 50% of energy at 5G small-cell base stations is consumed by computation power.
Takeaways & Limitations
5G small-cell energy-efficiency optimization should consider computation and transmission power together.
Abstract
from arXiv · showhide
The deployment of a large number of small cells poses new challenges to energy efficiency, which has often been ignored in fifth generation (5G) cellular networks. While massive multiple-input multiple outputs (MIMO) will reduce the transmission power at the expense of higher computational cost, the question remains as to which computation or transmission power is more important in the energy efficiency of 5G small cell networks. Thus, the main objective in this paper is to investigate the computation power based on the Landauer principle. Simulation results reveal that more than 50% of the energy is consumed by the computation power at 5G small cell BS's. Moreover, the computation power of 5G small cell BS can approach 800 watt when the massive MIMO (e.g., 128 antennas) is deployed to transmit high volume traffic. This clearly indicates that computation power optimization can play a major role in the energy efficiency of small cell networks.
I. INTRODUCTION
5G small-cell networks must reassess energy efficiency because dense deployment and heavy traffic increase computation demands while reducing transmission-power requirements. The paper models computation power using Landauer’s principle and finds it can dominate energy consumption.
- Massive MIMO and millimeter-wave technologies increase the volume and complexity of signal processing.
- Ultra-dense 5G small-cell deployment can make base-station computation power larger than transmission power despite lower transmission requirements.
- Heavy anticipated traffic requires greater signal-processing computation at small-cell baseband units.
- The paper concludes that energy-efficiency optimization should consider computation and transmission power together.
- The study proposes a computation-power model for 5G small-cell networks based on Landauer’s principle.
- More than 50% of energy at 5G small-cell base stations is consumed by computation power.
II. POWER CONSUMPTION AT BSS
The paper analyzes small-cell base-station power consumption by incorporating 5G transmission technologies, including massive MIMO and millimeter wave.
- The analysis evaluates total base-station power consumption to compare the roles of computation and transmission power.
- Massive MIMO and millimeter-wave technologies are incorporated into the small-cell base-station power-consumption analysis.
A. BS Power Consumption Types
Base-station power consumption is divided into transmission, computation, and additional power, each covering distinct operational components.
- Base-station power consumption comprises transmission power, computation power, and additional power.
- Transmission power covers power amplifiers, RF chains, and feeder losses involved in wireless-signal transformation.
- Computation power is consumed at baseband units for digital processing, management, control, and core-network communication functions.
- Additional power covers base-station consumption other than transmission and computation power.
- Unlike macro-cell base stations, small-cell base stations normally do not have active cooling systems.
B. Total BS Power Consumption Model
The total base-station model combines transmission, computation, and power-loss components, while accounting for antenna count, technology parameters, and small-cell operating conditions.
- The EARTH framework divides base-station power into antenna interface, power amplifier, RF chains, BBU, mains supply, cooling, and DC-DC components.
- In macro-cell base stations, the power amplifier and antenna interface account for 57% of total consumption, while RF chains and BBUs account for about 10% and 13%.
- The model calculates input base-station power as PA and RF-chain power per antenna multiplied by antenna count, plus BBU computation power.
- Power-loss rates for small-cell base stations are configured as 8% for DC-DC conversion, 10% for mains supply, and 0% for cooling.
- The model treats RF-chain power per antenna as fixed constants for different base-station types and assumes BBU power is constant in traditional models.
- Small-cell deployment reduces transmission power through shorter base-station-to-user distances, making the BBU a dominant power-consumption source.
III. COMPUTATION POWER MODEL
The model evaluates computation power in 5G small-cell base stations by measuring BBU data-processing operations and applying Landauer’s principle. It also examines how massive MIMO and millimeter-wave technologies affect computation power.
- BBU data-processing volume is evaluated using operations per second at 5G small-cell BSs.The model targets extensive traffic processing at the base station.
- Landauer’s principle is used to estimate the computation power consumed by data processing.
- The analysis studies the effects of massive MIMO and millimeter-wave technologies on 5G small-cell computation power.
A. Computation Power Types
The BBU contains systems for baseband processing, control, transfer, and power management. In small-cell BSs, integrated semiconductor chips simplify the BBU architecture while increasing computation demands with massive traffic.
- Small-cell architecture: Massive traffic is expected to increase BBU power consumption as small-cell BSs replace macro-cell BSs for wireless data transmission.
- BBU systems: Traditional macro-cell BBUs include baseband, control, transfer, and power systems.The baseband system performs filtering, FFT/IFFT, modulation, demodulation, DPD, signal detection, and channel coding or decoding.
- BBU systems: The control system manages resource allocation, communication-control protocols, and the man-machine interface for local configuration.
- BBU systems: The transfer system connects the BS to the core network through S1 and forwards inter-BS control and management information through X2.
- Small-cell architecture: Small-cell BS functions are integrated into a few semiconductor chips, making their BBU systems simpler than macro-cell BBUs.
B. Computation Power Model
The computation-power model combines Landauer-based energy estimates with hardware-processing characteristics and a reference base station. It uses a technology coefficient and system-parameter scaling to estimate computation power across different BBU implementations.
- The model addresses the difficulty of calculating computation power across the four BBU systems and their semiconductor hardware.BBU computation power is summed across the hardware parts or semiconductor chips.
- The model estimates computation power for DPD, filtering, CPRI, OFDM, frequency-domain processing, FEC, and CPU hardware.These components cover processing functions including sampling, serial links, FFT/OFDM, MIMO equalization, and channel coding or decoding.
- Landauer’s principle provides a lower theoretical computation-power bound based on the energy cost of logically irreversible information operations.Erasing one bit consumes more than kT ln(2) energy.
- A power coefficient ε represents semiconductor technology by relating transistor active-switching power to Landauer’s limit.The coefficient captures the distance between practical chip techniques and the theoretical limit.
- For 22-nanometer chips, the model configures ε = 10^3 and approximates transistor switching power as εkT ln(2) per operated bit.
- The model converts instructions per second and GOPS into semiconductor information throughput, then multiplies throughput by active-switching power.It assumes a 64-bit semiconductor-chip architecture and uses configured parameters ω = 0.1 and γ = 0.64.
- A reference base station enables computation-power estimates for real BBUs by comparing their system parameters with the reference configuration.The parameters include bandwidth, antenna count, modulation, coding rate, and time- and frequency-domain duty cycling.
- Scaling factors S_i encode whether each hardware component’s relationship with a BS system parameter is linear, nonlinear, or independent.The model sets S_i to 1, 2, or 0 for these three cases, respectively.
IV. EVALUATIONS OF COMPUTATION POWER
The evaluation examines how antenna count and bandwidth affect base-station computation power in 5G networks. Computation power increases with both factors and can reach hundreds or thousands of watts under massive-MIMO and millimeter-wave settings.
- Model validation: 7.22 W total and 3.6 W computation power are obtained with the proposed model, compared with 6.2 W and 2.4 W, respectively, for a reference small cell BS.The proposed model’s results agree with real wireless-network measurements and are shown capable of estimating 5G small-cell network power consumption.
- Computation-power ratio: Computation-power ratios increase with antenna count, and small-cell BSs have ratios larger than 50% and higher than macro-cell BSs.For bandwidth, the ratio also increases; with millimeter-wave technology and bandwidth at least 20 MHz, the small-cell ratio exceeds 50%.
V. FUTURE CHALLENGES
The paper identifies future energy-efficiency challenges involving computation power, transmission power, and their tradeoff in 5G small-cell networks. It calls for models and joint optimization methods that account for computation, transmission, storage, and cooling-related energy.
- Role of computation power: Computation power plays a more important role than other power components in 5G small-cell energy-efficiency optimization.The paper attributes this partly to reduced transmission power with massive MIMO and millimeter-wave technologies.
- Research gaps: Existing studies usually fix base-station computation power, while the effects of massive MIMO and millimeter-wave technologies on it are often ignored.The paper reports that these technologies have a greater impact on computation power at 5G small-cell BSs.
- Resource optimization: Scheduling antenna numbers and bandwidths is identified as a route to optimize computation power at 5G small-cell BSs.The paper frames this as a challenge arising when massive MIMO and millimeter-wave transmission are adopted.
- Power tradeoff: The computation–transmission-power tradeoff requires further investigation because the technologies can affect the two energy-efficiency components contradictorily.The paper proposes investigating their relationship and optimizing the tradeoff for small-cell BSs.
- Potential directions: Future work should develop energy-efficiency models and joint schemes that save computation and transmission power across BBUs and RF chains.The paper also suggests using SDN with cloud/fog computing functions to trade off computation and transmission powers.
- Cooling and heat recovery: Higher computation power can require additional cooling, motivating energy-cycle approaches that convert BBU heat into electrical energy.The paper specifically mentions the pyroelectric effect as a potential technology.
VI. CONCLUSIONS
The paper argues that computation power, previously ignored or treated as a small constant, must be evaluated jointly with transmission power in 5G small cell energy efficiency. Using Landauer-based estimation, it finds computation power grows with antennas and bandwidths and can become the more important component.
- Computation power was previously ignored or fixed as a small constant in traditional base-station energy-efficiency evaluations.
- The study estimates 5G small-cell base-station computation power using Landauer’s principle under massive MIMO and millimeter-wave technologies.
- Computation power increases as the number of antennas and bandwidths increases.
- Computation power plays a more important role than transmission power in 5G small-cell energy-efficiency optimization.
- Energy-efficiency optimization should consider computation and transmission power together.
- Converging computation and transmission technologies for 5G energy-efficiency optimization remains an open issue.
BIOGRAPHIES
The paper includes biographies of its authors and presents figures and a configuration table covering base-station architecture, BBU hardware, chip-technique power coefficients, computation power, computation-power ratios, and BBU parameters.
- BIOGRAPHIES: The authors include researchers from Huazhong University of Science and Technology, the University of Technology Sydney, and the National Institute of Standards and Technology.
- BIOGRAPHIES: The listed research interests include green communications, wireless networks, smart grids, mobile communications, and visual communications.
- FIGURES: The figures cover eNodeB logistical architecture, BBU hardware architecture, chip-technique power coefficients, computation power, and computation-power ratios.
- TABLE I: Table I presents configuration parameters of the BBU.