Space–ground integration digital twin

Build a joint digital twin environment for satellite and terrestrial 5G networks, evaluating space–ground link delay, attenuation, Doppler and end-to-end service QoS assurance policies in one closed loop, so integrated space–ground resource scheduling and service quality assurance work in simulation before deployment in orbit.

One-line positioning

Build a joint digital twin environment for satellite and terrestrial 5G networks, evaluating space–ground link delay, attenuation, Doppler and end-to-end service QoS assurance policies in one closed loop, so integrated space–ground resource scheduling and service quality assurance work in simulation before deployment in orbit.

Solution overview

6G will integrate terrestrial networks, satellite communications, optical communications and other communication modes to deliver global coverage and seamless handover. LEO satellite communications are a mainstream direction for next-generation 6G satellite communications. Building on 6G multimodal characteristics, space–ground network integration unifies satellite and existing terrestrial communication technologies at network protocol architecture and management levels. Technical advantages include integration and handover between satellite and terrestrial 5G networks, high speed and low latency, high capacity, intelligent network resource management and unified service quality assurance.

The central engineering challenge is that terrestrial resource allocation models assume “stable links and controllable latency”, whereas space–ground links inherently have long delays, high dynamics, and periodic variations with weather and orbit. Their QoS assurance mechanisms must be jointly optimized to provide unified service quality assurance in one network.

The core deliverable is a joint space–ground integrated QoS assurance optimization simulation system, comprising control-plane, service terminal simulation, network infrastructure layer, and management/demonstration subsystems. Core network simulation, link simulation tools and network impairment emulators collaborate in a loop of “initial configuration → link simulation → impairment injection → probe feedback → policy optimization”. The solution has been delivered in a real engineering project.

Core capabilities

① Joint simulation architecture with four subsystems

  • Control-plane subsystem (core control unit): unified multidimensional network resource management and orchestration, QoS assurance policy control, adaptive intelligent resource adaptation and collaboration, intelligent multi-domain resource scheduling control, QoS monitoring and optimization mechanisms, and system management
  • Service terminal simulation subsystem (application simulation unit): bidirectional terminal-side and network-side simulation of 8 service scenarios and metrics
  • Network infrastructure layer subsystem (network environment simulation unit): simulation of satellite base stations, terrestrial base stations, space–ground links, terrestrial 5G air-interface links, transport networks and core network user planes
  • Management and demonstration subsystem (metric demonstration unit): monitoring, collection, comparison and graphical presentation of core functions and metrics

The control plane uses a hierarchical architecture of master control system + space-based control system + ground-based control system. The master creates global resource scheduling policies from overall heterogeneous space–ground network state and service needs, delivering control policies and QoS requirements to space-based and ground-based controllers through southbound interfaces. Domain controllers generate optimized local control policies from local network state and deliver them to domain network elements, enabling real-time, on-demand end-to-end resource scheduling across multiple users, services, access types, subnets and protocols.

② Multi-layer, multi-domain closed-loop QoS assurance

The system assures QoS through three closed-loop levels: global QoS — task QoS — real-time QoS:

  1. The master controller performs global QoS decomposition for differentiated services, extracting hierarchical, domain-specific task QoS for distributed controllers to form a “service needs → global QoS → task QoS” decomposition loop
  2. Each distributed controller matches task QoS with pooled domain resources on demand, generating optimized domain resource scheduling policies to form a “task QoS → pooled resources” allocation loop
  3. Scheduling policies adapt to physical network elements, and network nodes execute them to transmit tasks
  4. After transmission, domain monitoring points report network state and real-time QoS results to the master controller, forming a “real-time QoS → global QoS” monitoring loop
  5. The master dynamically adjusts QoS policies based on comparisons, delivering policies to domain controllers for resource reallocation

③ Space–ground link simulation models

  • Gaussian noise model: simulate statistical noise characteristics of space–ground links, using signal-to-noise ratio (SNR) to represent signal power relative to noise power
  • Signal attenuation model: simulate attenuation using free-space path loss, also considering atmospheric attenuation, multipath effects and other factors
  • Satellite displacement model: predict satellite position and velocity using orbit prediction models to calibrate link position information
  • Doppler shift model: calculate frequency shifts caused by relative satellite/terminal motion, covering fast-moving satellites, airborne platforms at 20 kilometers and satellites at 600 kilometers

④ Propagation chain of key simulation parameters

Input distance, weather, noise, phase shift, multipath and other parameters to derive, in sequence:

  1. Signal delay to the ground, SINR and BLER
  2. SINR and BLER → CQI and MCS
  3. CQI and MCS → maximum link transmission rate (MCS code rate × 1024bps)
  4. Number of carrier service users → available RB count → user service rate

Then calculate:

  • Total space–ground end-to-end link delay = space–ground link delay (two segments) + terrestrial transmission delay
  • Total terrestrial end-to-end link delay = terrestrial 5G air-interface link delay + terrestrial transmission delay
  • Space–ground end-to-end bit error rate = maximum bit error rate of the space–ground links (two segments) and terrestrial transmission
  • Terrestrial end-to-end bit error rate = maximum bit error rate of the terrestrial 5G air-interface link and terrestrial transmission

⑤ Eight space–ground integrated service categories and QoS assurance

The eight categories span ultra-long-distance real-time communications through highly reliable guaranteed transmission. Each has independent 5QI, PDB, PER and guaranteed-rate parameters:

Service category Typical characteristics QoS metrics (design values) Rate parameters
Ultra-long-distance real-time Transoceanic voice calls, remote real-time news video and voice broadcasting; distances over 1000 kilometers across borders, oceans and continents, without terrestrial support and largely dependent on inter-satellite transmission PDB≤100ms, PER≤10⁻² GFBR 10Mbit/s, MFBR 25Mbit/s
Massive connectivity Space-based LEO IoT and terrestrial IoT terminal applications, such as narrowband IoT and large-scale robot control; connection density up to 1 million/km² 100ms≤PDB≤300ms, PER≤10⁻³ Session-AMBR 1Mbit/s, UE-AMBR 10Mbit/s
High-mobility broadband Broadband services from terminals moving at 300~1000 kilometers/hour, such as consumer applications on high-speed trains and aircraft PDB≤300ms, PER≤10⁻³ Session-AMBR 10Mbit/s, UE-AMBR 50Mbit/s
Real-time high-capacity transmission Real-time multimedia on demand, VR/AR, live broadcasting in remote areas and transoceanic real-time video conferencing PDB≤100ms, PER≤10⁻³ Space-based Session-AMBR 10Mbit/s / UE-AMBR 50Mbit/s; terrestrial Session-AMBR 100Mbit/s / UE-AMBR 1Gbit/s
Non-real-time high-capacity transmission Large file transfers, content-based video delivery and other non-real-time bulk data services with burst traffic Terrestrial and space-based LEO PDB≤300ms / PER≤10⁻³; space-based MEO/GEO PDB≤1100ms Space-based Session-AMBR 10Mbit/s / UE-AMBR 50Mbit/s; terrestrial Session-AMBR 20Mbit/s / UE-AMBR 100Mbit/s
Strong security assurance Confidential services such as dedicated encrypted government/enterprise communications and financial data transfers 100ms≤PDB≤300ms, PER≤10⁻⁶ GFBR 1Mbit/s, MFBR 50Mbit/s
Deterministic guaranteed transmission Industrial automation control (industrial robots and PLC control systems), medical assistance and remote control PDB≤150ms, PER≤10⁻³ Space-based GFBR 25Mbit/s / MFBR 50Mbit/s; terrestrial GFBR 50Mbit/s / MFBR 100Mbit/s
Highly reliable guaranteed transmission Aerospace communications (satellite communication system control channels, flight control systems and aircraft communications) PDB≤50ms, PER≤10⁻⁶ GFBR 10Mbit/s, MFBR 50Mbit/s

QoS parameters for the 8 services are determined from QoS metrics and corresponding 5QI values in 3GPP TS 23.501 R18 and satellite terminal test validation rates obtained by CCSA under laboratory conditions. A new non-delay-sensitive service for satellite scenarios uses 5QI 10 (maximum end-to-end PDB 1100ms), applying to medium-orbit and high-orbit geostationary satellite access.

⑥ Three-role collaborative simulation loop

QoS assurance advances through three levels: single-service optimization (mainly one-step 1RB adjustments), mixed optimization of heterogeneous multiple-service scenarios (mainly automatic optimization), and dynamic balancing of space–ground resources. Both single-service and multi-service optimization center on ARP adjustment (initial value 8 or the same value).

Simulation loop:

  1. The QoS assurance system sends initial QoS metrics, weather conditions and service user counts to core network simulation and link simulation tools
  2. The link simulation tool allocates radio resources using user counts and initial RB counts per service, estimates transport network resource usage, and calculates service rate, delay and bit error rate
  3. The link simulation tool reports radio and transport resource usage to the QoS system, while the core network reports its resource usage
  4. The link simulation tool sends simulated service rates, delays and bit error rates to the network impairment emulator
  5. The impairment emulator generates traffic from simulation parameters, sends impaired traffic to the core network UPF and starts the user plane
  6. The core network reports QoS measurements to the QoS system
  7. The QoS system displays current service QoS metrics and resource usage, completing one cycle

⑦ Space–ground resource scheduling and reservation

Considering time-varying space–ground network characteristics, the system schedules and reserves resources in advance for service continuity:

  • The space-based network displays satellite IDs and reserved bandwidth (RB counts) for the current service period and satellite IDs and bandwidth for continued service in the next period, showing 5 service periods over 20 minutes (4 minutes each)
  • The terrestrial network displays bandwidth RB requirements for current and handover-serving base stations
  • Core and transport networks display service QoS metrics and core-allocated user-plane bandwidth (BW), with graphical transport and core transmission paths
  • Use this information to predict resource usage and perform satellite-link and subsequent resource switching in advance to assure QoS metrics

⑧ Space–ground routing policies

Support next-hop selection and path visualization for space–ground routing. The system selects next-hop satellite nodes based on calculated routing metrics and displays the complete forwarding path as green dotted flowing curves in a flow diagram, making multi-hop space–ground path selection visible and reviewable.

⑨ Results presentation

High-definition 3D imagery and three-role graphical interfaces present satellite orbits, service user counts, network node counts, virtualized network resource management, service quality management and network optimization evaluation in an overall view of the 3D Earth + terrestrial base stations + satellite links + terminals. Key technical presentations include functional parameters and policy templates. Before/after service assurance comparisons cover four triggers: priority adjustments, congestion assurance, resource or bandwidth limits, and space–ground resource scheduling.

Key technical specifications

Item Parameters / description
Simulated satellite scale 200 satellites (example scenario configuration)
Simulated base station scale 1,000 base stations (example scenario configuration)
Service terminal simulation scale 1 million simulated service terminal UEs online; core network simulation supports 60,000 simulated base stations; up to 1,000 users per base station (design targets)
Service grouping 1 million UEs divided into 8~10 groups of approximately 100,000~125,000 UEs, each corresponding to one service type (subdividable into space-based / terrestrial services)
Service scenario categories 8 (ultra-long-distance real-time, massive connectivity, high-mobility broadband, real-time high-capacity, non-real-time high-capacity, strong security, deterministic assurance and high-reliability assurance)
End-to-end QoS parameters 5QI, ARP (priority level / pre-emption capability / pre-emption vulnerability), GFBR, MFBR, Session-AMBR, UE-AMBR, PDB, PER
Optimization levels Single-service / heterogeneous multi-service mixed / dynamic space–ground resource balancing
Resource scheduling display period 4 minutes per period, displaying 5 periods (20 minutes)
Weather variables Clear / light rain / heavy rain (to create resource shortages and trigger QoS rebalancing)
Simulation loop latency reference Fixed core network simulation latency of 20ms (transport network simulation considers delay and bit error effects under congested and uncongested conditions)
Frequency references IF centers 2.16GHz and 2.34GHz; Ka band 26.5~40GHz; Q/V band 30~300GHz; 5G millimeter wave (ITU) 24.25~27.5GHz, etc.
Interface protocols N5 (PCF policy delivery), N6 (DN probe QoS result reporting), and simulation tool parameter input interfaces, all using HTTP/2

Target customers

  • Satellite operators (space–ground integrated network planning and QoS policy design)
  • Terrestrial telecom operators (integrated space–ground service transport validation and joint QoS assurance research)
  • Research institutes and universities (teaching, research and demonstration of 6G satellite/terrestrial network integration and interaction)
  • Satellite network system design and standards research organizations

Verifiable facts

  • Engineering deployment: the core system has been delivered for a real national science and technology project; 18 test cases were designed and executed, and all 18 passed
  • Joint simulation of satellite base stations, terrestrial 5G base stations, terrestrial transport networks and 5G converged core networks is available
  • Modeling methods for four space–ground link models (noise, attenuation, displacement and Doppler) and decomposed end-to-end delay and bit error rate calculations are available
  • QoS parameter templates for eight integrated service categories and three-level optimization procedures are available
  • Space–ground routing metric calculation and forwarding path visualization are available
  • 3D Earth graphical demonstrations cover satellite orbits, base station distribution, resource usage and before/after QoS optimization comparisons

Deliverables

Deliverable Description
Joint space–ground integrated QoS assurance optimization simulation system Control-plane, service terminal simulation, network infrastructure layer and management/demonstration subsystems
QoS policy templates for 8 service categories Default 5QI, PDB, PER, GFBR / MFBR / AMBR and ARP configurations per category
Simulation scenario configurations and optimization scripts Multi-round configurations for weather, user counts, RB allocation and ARP adjustments
Test report Report issued by a third-party testing body (available to customers after redaction; the original includes project commissioning information and is not publicly disclosed)
Demonstration materials 3D graphical demonstration interface and before/after optimization comparison videos