Latency-Aware Offloading in Integrated Satellite Terrestrial Networks

Wiem Abderrahim, Osama Amin, Mohamed-Slim Alouini, Basem Shihada

Abstract

Next-generation communication networks are expected to integrate newly-used technologies in a smart way to ensure continuous connectivity in rural areas and to alleviate the traffic load in dense regions. The prospective access network in 6G should hinge on satellite systems to take advantage of their wide coverage and high capacity. However, adopting satellites in 6G could be hindered because of the additional latency introduced, which is not tolerable by all traffic types. Therefore, we propose a traffic offloading scheme that integrates both the satellite and terrestrial networks to smartly allocate the traffic between them while satisfying different traffic requirements. Specifically, the proposed scheme offloads the Ultra-Reliable Low Latency Communication (URLLC) traffic to the terrestrial backhaul to satisfy its stringent latency requirement. However, it offloads the enhanced Mobile Broadband (eMBB) traffic to the satellite since eMBB needs high data rates but is not always sensitive to delay. Our scheme is shown to reduce the transmission delay of URLLC packets, decrease the number of dropped eMBB packets, and hence improve the network's availability. Our findings highlight that the inter-working between satellite and terrestrial networks is crucial to mitigate the expected high load on the limited terrestrial capacity.

Introduction

As the telecommunications industry pivots toward 6G, terrestrial networks are encountering a significant "capacity crunch." The proliferation of data-intensive applications and the density of urban deployments are pushing existing ground-based backhaul links—whether fiber or microwave—to their physical limits. While Low Earth Orbit (LEO) satellite constellations (such as SpaceX's Starlink or OneWeb) offer a promising solution for global coverage and high capacity, their integration into the 6G fabric is complicated by the inherent propagation delays associated with space-to-ground links.

The fundamental challenge lies in the heterogeneous nature of modern data traffic. 6G must simultaneously support Ultra-Reliable Low Latency Communication (URLLC), which requires sub-millisecond response times for applications like autonomous driving, and enhanced Mobile Broadband (eMBB), which demands massive throughput for high-definition streaming. This paper proposes a smart offloading framework that leverages LEO satellites as a "relief valve" for terrestrial networks, strategically routing traffic based on specific Quality of Service (QoS) requirements.

System Model and Architecture

The proposed Integrated Satellite-Terrestrial Network (ISTN) architecture consists of multiple terrestrial small cells served by a single LEO satellite. Each cell contains a set of base stations and various user terminals generating both URLLC and eMBB traffic.

System architecture of the proposed Integrated Satellite-Terrestrial Network (ISTN) showing traffic offloading between terrestrial and satellite links.

As illustrated in the figure above, the system differentiates between traffic types at the source. URLLC traffic ($\lambda_i^u$) is prioritized for terrestrial backhaul links ($C_{Ter}$) to minimize latency. Conversely, eMBB traffic ($\lambda_i^e$) is offloaded to the satellite link ($C_{Sat}$) when terrestrial capacity is congested. This dual-path approach ensures that high-volume, delay-tolerant data does not block time-critical packets.

Proposed Offloading Scheme

The core of the proposed solution is a two-step optimization framework designed to manage resources across the ISTN:

  1. Terrestrial Resource Allocation and Puncturing: Within the terrestrial small cells, the system employs a "puncturing" technique. This allows urgent URLLC packets to preemptively overwrite portions of ongoing eMBB transmissions. To manage the resulting interference and resource distribution, the authors utilize Successive Convex Approximation (SCA). This method transforms the complex, non-convex resource allocation problem into a series of solvable convex sub-problems, ensuring efficient convergence.
  2. Integrated Offloading Logic: The system dynamically determines the volume of eMBB traffic to offload to the satellite. This decision is formulated as a multi-objective optimization problem that balances satellite bandwidth availability, terrestrial backhaul capacity, and the instantaneous load of individual base stations.
  3. Analytical Modeling: The terrestrial link is modeled as an M/G/1 queue. This mathematical foundation allows for the analytical derivation of mean waiting times and the probability distribution of delays for URLLC packets, providing a theoretical guarantee for latency performance.

Performance Evaluation

The framework was validated using industry-standard simulation parameters for three prominent LEO constellations: SpaceX (Starlink), OneWeb, and Telsat LEO. Terrestrial parameters were modeled after standard 5G small cell configurations with a 500m radius and 100MHz bandwidth.

Key Results

  • Latency Reduction: The ISTN scheme significantly outperforms traditional terrestrial-only benchmarks. The reduction in URLLC delay is most pronounced in scenarios where ground capacity is limited (e.g., below 30 Mbps). By moving heavy eMBB traffic to the "space lane," the terrestrial "fast lane" remains clear for urgent data.
  • Network Availability: Offloading to space drastically reduces the number of dropped eMBB packets. Network availability—defined as the percentage of successfully transmitted data—reaches 100% significantly faster as link capacity increases compared to conventional architectures.
  • Constellation Performance: Simulations indicated that SpaceX parameters provided the highest performance gains, attributed to its superior capacity and carrier-to-noise density ratios compared to the other modeled constellations.
  • Algorithmic Efficiency: The SCA-based optimization proved robust, typically reaching an optimal sum rate for eMBB traffic within approximately 70 iterations.

Conclusion and Future Directions

This research demonstrates that satellites should not be viewed merely as a solution for rural connectivity but as an essential component of urban 6G infrastructure. By acting as high-capacity bypasses for delay-tolerant data, LEO satellites can effectively mitigate terrestrial congestion and safeguard the performance of mission-critical URLLC applications.

Future work will address the complexities of LEO handovers and the dynamic nature of satellite constellations. Additionally, the integration of Machine Learning (ML) is proposed to enable proactive offloading. By predicting traffic spikes before they occur, ML-driven systems could further optimize latency and ensure seamless connectivity across the integrated network.

Summary updated

Abstract

Next-generation communication networks are expected to integrate newly-used technologies in a smart way to ensure continuous connectivity in rural areas and to alleviate the traffic load in dense regions. The prospective access network in 6G should hinge on satellite systems to take advantage of their wide coverage and high capacity. However, adopting satellites in 6G could be hindered because of the additional latency introduced, which is not tolerable by all traffic types. Therefore, we propose a traffic offloading scheme that integrates both the satellite and terrestrial networks to smartly allocate the traffic between them while satisfying different traffic requirements. Specifically, the proposed scheme offloads the Ultra-Reliable Low Latency Communication (URLLC) traffic to the terrestrial backhaul to satisfy its stringent latency requirement. However, it offloads the enhanced Mobile Broadband (eMBB) traffic to the satellite since eMBB needs high data rates but is not always sensitive to delay. Our scheme is shown to reduce the transmission delay of URLLC packets, decrease the number of dropped eMBB packets, and hence improve the network's availability. Our findings highlight that the inter-working between satellite and terrestrial networks is crucial to mitigate the expected high load on the limited terrestrial capacity.
BibTeX
@article{abderrahim2020latency,
  author = {Wiem Abderrahim and Osama Amin and Mohamed-Slim Alouini and Basem Shihada},
  title = {{Latency-Aware Offloading in Integrated Satellite Terrestrial Networks}},
  journal = {IEEE Open J. Commun. Soc.},
  year = {2020},
  url = {https://openreview.net/forum?id=xgC4bF9fibX},
}
Figure