Propagation and Rate-Aware Cell Switching Optimization in HAPS-Assisted Wireless Networks

This paper proposes a rate-aware cell switching optimization framework for HAPS-assisted 6G networks that integrates realistic propagation losses and multi-objective formulation to significantly reduce data rate degradation while maintaining energy efficiency and user connectivity, validated through both system-level simulations and Sionna-OAI emulation.

The Gist

Imagine a bustling city where the "internet" is delivered by a fleet of delivery trucks (Base Stations) and a giant, hovering drone (HAPS) that can cover the whole city from above.

Right now, these trucks are working hard to deliver data to everyone. But keeping all of them running 24/7 is expensive and wastes a lot of energy. The old way of saving energy was simple: "If a truck isn't delivering much, turn it off and let the others pick up the slack."

The problem? This old method was a bit clumsy. It didn't care who was left behind. If you turned off a truck near a tall building, the people inside might get their signal blocked by the walls. If you forced them to connect to the giant drone far away, the signal might get weak because of the rain or the atmosphere. The result? You saved money on electricity, but your video call froze, or you lost your connection entirely.

This paper proposes a smarter, more considerate way to manage these trucks. Here is the breakdown:

1. The New Rules of the Road (The Problem)

The authors realized that turning off a cell tower isn't just about math; it's about physics.

• The Wall Effect (Building Entry Loss): Imagine trying to shout to someone inside a concrete bunker. It's hard. If you turn off the truck right outside their door, they have to shout to the drone miles away, and the signal gets even weaker.
• The Weather Effect (Atmospheric Loss): The drone has to fly through rain and air, which eats up signal strength, just like fog makes it hard to see.

The old system ignored these "obstacles." The new system says, "We can't just turn off a truck if it means the people inside the building will lose their connection or get a terrible speed."

2. The Two New Strategies (The Solution)

To fix this, the authors created a "Traffic Control Center" that uses two different strategies to decide which trucks to turn off. They treat the goal like a three-legged stool: you need to balance Energy Savings, Keeping Everyone Connected, and Keeping Speeds High.

Strategy A: The "Weighted Score" (WSM)

Think of this like a restaurant menu where you can customize your order.

• You can tell the system: "I care 50% about saving energy, 30% about keeping people connected, and 20% about speed."
• Or, if it's a crisis, you can say: "Don't worry about energy; just make sure everyone has internet!"
• The Analogy: It's like a smart thermostat that adjusts the heat based on how many people are in the room and how cold it is outside, rather than just turning the heat off completely.

Strategy B: The "Safety Net" (ϵCM)

This is stricter. It says: "We want to save as much energy as possible, BUT we have hard rules we cannot break."

• Rule 1: No one can lose their connection.
• Rule 2: No one's internet speed can drop below what it was before.
• The Analogy: It's like a budget. You can spend money (turn off trucks) to save energy, but you cannot spend so much that you can't pay your rent (connectivity) or buy food (data speed). If turning off a truck breaks a rule, the system says, "Nope, keep that truck on."

3. The Real-World Test (The Emulation)

Most studies just run these ideas on a computer simulation, which is like playing a video game of traffic. The authors didn't stop there. They built a miniature, real-life testbed using actual software that mimics real cell towers.

• They tested their ideas in a tiny "city" with real radio waves, real building walls, and real weather effects.
• The Result: Their smart strategies worked!
  • For people inside buildings with thick walls (High-Loss), the new methods reduced bad internet speeds by up to 70% compared to the old "turn everything off" method.
  • For people with weaker walls (Low-Loss), they completely eliminated the drop in speed that usually happens when towers are turned off.

The Big Picture

This paper is about sustainability. In the future (6G networks), we need to save energy to save the planet, but we can't do it by making people's lives worse.

The authors showed that by being aware of the environment (walls, weather) and using smart balancing acts, we can turn off the lights to save energy without leaving anyone in the dark. It's the difference between a clumsy bouncer kicking everyone out of a club to save on electricity, and a smart manager who gently guides people to the right rooms so the lights can be turned off, but the party keeps going.

Technical Summary

Here is a detailed technical summary of the paper "Propagation and Rate-Aware Cell Switching Optimization in HAPS-Assisted Wireless Networks."

1. Problem Statement

The rapid growth of mobile traffic and the requirements for 6G networks (ubiquitous connectivity, ultra-reliability) necessitate network densification. However, deploying more Base Stations (BSs) significantly increases energy consumption, conflicting with 6G sustainability goals.

• Current Limitations: Existing cell switching strategies (turning off low-load BSs to save energy) rely on simplified models. They typically treat users as generic traffic loads, ignoring realistic propagation effects such as Building Entry Loss (BEL) and atmospheric losses (gas/rain attenuation) relevant to Non-Terrestrial Networks (NTN) like High-Altitude Platform Stations (HAPS).
• The Consequence: When a BS is switched off, users are offloaded to remaining BSs (often HAPS or distant terrestrial BSs). Without accounting for propagation losses, this can lead to:
  • Reduced Signal-to-Interference-plus-Noise Ratio (SINR).
  • Connection failures (unconnected users).
  • Significant data rate degradation (dissatisfied users).
• Objective: To reformulate the cell switching problem to jointly optimize energy efficiency, user connectivity, and data rate performance while explicitly modeling environmental propagation losses in a HAPS-assisted network.

2. Methodology

The authors propose a comprehensive framework combining mathematical modeling, multi-objective optimization, and hybrid validation (simulation + emulation).

A. System & Propagation Models

• Network Topology: A hybrid network consisting of one HAPS (Super-Macro BS), one Macro BS (MBS), and multiple Small BSs (SBSs).
• User Classification: Users are categorized into three types based on propagation conditions:
  1. High-loss indoor: Experiences high BEL and additional attenuation.
  2. Low-loss indoor: Experiences reduced BEL.
  3. Outdoor: No BEL.
• Propagation Modeling:
  • Terrestrial (TN): Uses 3GPP Urban Macro (UMa) models for Line-of-Sight (LoS) and Non-LoS (NLoS), explicitly adding BEL ($L_{BEL}$) and user-type specific losses.
  • Non-Terrestrial (NTN/HAPS): Uses 3GPP models incorporating Free Space Path Loss (FSPL) and Atmospheric Losses ($L_a$, including gas and rain attenuation).
• Power Model: Uses the EARTH model to calculate instantaneous power consumption based on load factors, distinguishing between active and sleep modes.

B. Optimization Framework

The conventional energy-focused problem (minimizing power only) is reformulated into a Multi-Objective Optimization Problem (MOP) that minimizes:

1. Total Network Power Consumption ($P$).
2. Number of Unconnected Users ($U$).
3. Number of Dissatisfied Users ($D$) (those whose data rate drops after switching).

Two complementary solution approaches are developed:

1. Weighted Sum Method (WSM): Combines the three objectives into a single function using adaptive weights ($\alpha, \beta, \upsilon$). This allows dynamic prioritization (e.g., QoS-dominant vs. Power-dominant scenarios).
2. $\epsilon$-Constraint-Inspired Method ($\epsilon$CM): Converts the connectivity and rate objectives into hard constraints ($R_a \geq R_b$ and $U_a \leq U_b$) within the primary power minimization problem. This ensures sustainability gains are achieved without compromising user experience.

C. Validation Strategy

Unlike prior works relying solely on simulations, this study employs a hybrid validation framework:

• System-Level Simulations: Used for large-scale network analysis.
• Sionna–OpenAirInterface (OAI) Emulation: A smaller network emulation where Sionna handles channel modeling/SINR calculation and OAI handles the radio access network (RAN) association. This validates the concept under realistic protocol conditions.

3. Key Contributions

1. Propagation-Aware Modeling: First attempt to mathematically derive the impact of BEL and atmospheric losses specifically within the context of HAPS-assisted cell switching.
2. Multi-Objective Reformulation: Shifts the paradigm from single-objective (energy-only) to a multi-objective framework that balances energy, connectivity, and data rates.
3. Dual Solution Approaches: Development of WSM (flexible weighting) and $\epsilon$CM (constraint-based) to solve the complex MOP, offering different trade-off mechanisms for network operators.
4. Realistic Validation: Integration of system-level simulations with Sionna-OAI emulation to bridge the gap between theoretical optimization and practical implementation.

4. Key Results

The study evaluated performance across varying Building Entry Loss (BEL) values (0–30 dB) and user densities.

• Data Rate Performance:
  • Conventional (EFM): Caused severe rate degradation for high-loss indoor users as BEL increased. It also caused a 44% drop in rates for low-loss indoor users in certain scenarios.
  • WSM: By tuning weights, WSM reduced rate degradation for high-loss indoor users by up to 70% and eliminated the drop for low-loss indoor users entirely in balanced scenarios.
  • $\epsilon$CM: Successfully maintained data rates for indoor users up to ~10–25 dB of BEL by keeping necessary BSs active, though this slightly reduced outdoor rates due to increased interference.
• Power Consumption:
  • EFM: Achieved the lowest power consumption but at the cost of connectivity and rate.
  • WSM & $\epsilon$CM: Showed that significant power savings are still possible (especially in low-to-moderate user densities) while maintaining QoS. For example, $\epsilon$CM reduced power for up to 400 users while strictly adhering to connectivity constraints.
• Algorithm Comparison:
  • Exhaustive Search: Found the optimal solution but has exponential complexity ($O(2^\Gamma)$), making it impractical for dense networks.
  • Greedy & Genetic Algorithms: Provided near-optimal solutions with significantly lower complexity, suitable for real-time deployment.

5. Significance

This paper addresses a critical gap in 6G network sustainability research. By demonstrating that ignoring propagation physics leads to suboptimal and potentially unusable cell switching strategies, the authors provide a robust framework for future networks.

• Sustainability: It proves that energy efficiency does not have to come at the expense of user experience (connectivity and speed).
• HAPS Integration: It provides a concrete methodology for integrating HAPS into terrestrial networks, accounting for the unique atmospheric and path-loss challenges of non-terrestrial links.
• Practicality: The use of Sionna-OAI emulation moves the research from theoretical abstraction toward deployable solutions, validating that these optimization strategies can function within realistic 5G/6G protocol stacks.

In conclusion, the proposed Propagation and Rate-Aware Cell Switching ensures that 6G networks can achieve their sustainability goals without compromising the quality of service required for immersive and reliable communications.