Industry positioning
For research institutes and universities, we provide research validation environments for frontier directions such as AI-RAN, semantic communication and computing-network convergence, together with experimental teaching platforms for 5G and satellite internet.
Industry challenges
Frontier communication research shares common challenges: commercial terminals, base stations and core networks do not support the non-standard protocols and custom coding required for research; experiments with real devices are limited in scale and difficult to configure into scenarios; and end-to-end environments require integration of solutions from multiple suppliers, creating significant joint debugging risks.
Application scenarios
① AI-RAN testing
For research into the integration of AI and wireless networks, we provide full-lifecycle AI model validation and AI for RAN effectiveness testing.
- Full-lifecycle AI model validation: covers model training → testing → emulation → deployment → inference → monitoring and updates, supporting evaluation of new models’ impact on existing networks in near-real network conditions
- AI for RAN scenario validation: beam management (beam prediction), CSI feedback enhancement (two-sided model CSI compression and time-domain prediction), mobility optimization (handover success rate), load balancing, and network energy saving
- Control-plane and user-plane perception testing: accurately simulate control-plane UE behavioral characteristics and traffic models (high-volume registrations, high-volume calls, high-speed rail/highway models), construct characteristics of poor user experience, and quantitatively evaluate optimization results (call drop rate, handover success rate, video stalling, and web page opening time)
② Semantic communication testing
For semantic communication, a frontier 6G direction, we provide customizable protocol stacks and end-to-end validation environments.
- Flexible protocol stack customization: the terminal simulator uses SDR technology to flexibly customize RLC, SDAP and MAC layer protocol fields, provides modification of standard and non-standard DCI fields, and enables PHY layer coding schemes to be switched by service scenario, meeting semantic communication requirements for approaches with and without channel coding
- Large-scale semantic terminal simulation: a single terminal simulator supports hundreds of simulated semantic terminals, allowing transmission efficiency comparisons between semantic and conventional communication
- End-to-end environment: terminals, base stations, core networks and a semantic knowledge base work together; core networks and service traffic generators use virtualized deployment, enabling end-to-end validation with a small amount of equipment
- Knowledge base integration: API interfaces connect to the semantic communication knowledge base, updating core network QoS policies for different service types
- Field-level editing: thousands of signaling message field templates for RRC / NAS / HTTP2 / PFCP / NGAP, with field editing through a visual interface
③ Computing-network convergence testing
For computing network and computing-network convergence research, we provide computing constellation simulation and validation capabilities.
- Distributed simulation and validation of space-based computing networks, supporting joint simulation of computing nodes, network topology and service flows
- Computing network performance testing: collective communication efficiency, lossless Ethernet and ROCEv2 compatibility validation
- Protocol validation and scenario development for frontier computing network research
④ 5G + satellite internet teaching
For undergraduate and vocational institutions, we provide modular experimental teaching platforms for 5G and satellite internet.
- Complete closed-loop teaching process: teaching → Q&A → practical exercises → examinations → technical research
- End-to-end simulation environment: dual scenarios for 5G and satellite internet—on the terrestrial side, core networks (4G / 5G / NTN / 6G) + access networks + terminals; on the satellite side, spaceborne base stations, space-to-ground links and inter-satellite networking. Signaling procedures conform to 3GPP specifications, with packet capture and Wireshark analysis supported
- Satellite internet teaching modules: NTN access and handover experiments, space-to-ground link impairment demonstrations, and inter-satellite routing networking experiments, supporting satellite internet curriculum development
- White-box experiments: programmable protocol stacks supporting custom networking, signaling field modification, reverse import of PCAP packets, signaling research and custom experiments
- AI modules: automatically generate configurations from instructions, answer questions and automatically analyze PCAP packets
- Concurrent multi-user access: supports 200+ users online simultaneously, cloud and on-premises deployment, and API integration with campus systems
- Courses and teaching materials: supporting course content for 4G / 5G / NTN / 6G core networks, access networks, terminals, satellite internet, industrial internet and other topics
Verifiable facts
- Semantic communication: a single terminal simulator supports concurrent simulation of hundreds of semantic terminals; an end-to-end solution is provided
- 5G + satellite internet teaching: one physical server can host hundreds of virtual core network systems; concurrent multi-user access supports 200+ users online simultaneously
- The teaching platform has been deployed at multiple universities
Success stories
| Case | Content |
|---|---|
| National laboratory · 6G semantic communication end-to-end test environment | Development of an end-to-end test environment for 6G semantic communication |
| National key laboratory · Intrinsic-security programmable protocol stack PoC testing | Programmable protocol stack PoC testing for intrinsic security |
| Beijing University of Posts and Telecommunications · 5G simulation teaching environment | 5G simulation teaching environment (virtualization of all UE / gNB / 5GC network elements) |
| Zijinshan Laboratory · AIRAN simulation test platform | AI-RAN performance testing platform based on a high-performance multi-user simulator |
