MWC Shanghai 2026: Luowave Unveils 6G & AI-Native SDR Solutions for the "IQ Era"

From June 24 to 26, 2026, MWC Shanghai, the annual flagship event of the global mobile communications industry, was successfully held at the New International Expo Centre.

Themed "The IQ Era," this year's conference focused on three core agendas: the commercial scale-up of 5G-Advanced, the evolution toward AI-native networks, and the formulation of 6G standards. Wuhan Nuoguang Electronic showcased its flagship platforms, including the LW-X410 PLUS, DEEPSDR 1 & 2, and the SDR-LW 4940 Note all-in-one terminal. Alongside these, the company exhibited its comprehensive SDR product portfolio—comprising the LW-B205, B210, N310, N210, N321, E320-G, and X310—engaging with industry peers on cutting-edge RF hardware solutions tailored for 6G and edge AI.

Event Background: Key Takeaways from MWC Shanghai

MWC Shanghai 2026 delivered a clear signal: wireless communication is undergoing a fundamental shift from "connecting devices" to "orchestrating intelligence."

As hybrid AI penetrates devices, telecom RAN edges, and the cloud, 6G is emerging as the core enabling technology driving this transformation—elevating connectivity into a collaboratively orchestrated network of interconnected intelligence. The newly debuted 6G Zone and Mobile AI Innovation Frontiers exhibition areas were packed with attendees, reflecting a strong industry consensus on the convergence of "AI-Native + 6G."

This year's 6G Zone highlighted the co-evolution of AI and 6G, exploring frontiers such as intelligent edge collaboration, semantic communications, distributed computing, and embodied AI systems. Satellite-terrestrial integration and Non-Terrestrial Network (NTN) architectures were also prominent on the agenda.

However, no matter how advanced the concepts, the RF performance and computing power of the underlying hardware remain the prerequisites for validating any breakthrough. This is exactly what Nuoguang Electronic aims to deliver at this exhibition: concrete, production-ready hardware and solutions that work.

Booth Highlight: LW-X410 PLUS—Engineering Practice of Multi-Channel Coherent Synchronization

The centerpiece of our booth is the recently launched LW-X410 PLUS (Coherent Edition)—a four-channel software defined radio platform purpose-built for high-order coherent applications such as phased arrays and Massive MIMO.

Empowering Massive MIMO

The core advantage of the LW-X410 PLUS lies in its innovative LO Sharing architecture. Utilizing a proprietary RF link synchronization mechanism, the system supplies a unified local oscillator signal to all four RF channels. Combined with a two-stage superheterodyne architecture, it guarantees a phase stability of better than $1^\circ \text{ RMS}$ within a single unit or across cascaded systems. This metric directly determines the beamforming accuracy of phased array radars, spatial multiplexing in Massive MIMO, and the sensing resolution of Integrated Sensing and Communication (ISAC) systems.

Furthermore, the LW-X410 PLUS supports external LO cascading, enabling flexible deployment of distributed synchronization systems. Cascading multiple devices expands the coherent channel count, providing a cost-effective framework for building large-scale Massive MIMO testbeds.

Target Applications

• 5G-Advanced & 6G Candidate Band Verification: Featuring a 4T4R full-duplex architecture and wideband coverage from 1MHz to 7.2GHz (tunable up to 8GHz), the platform supports a maximum instantaneous bandwidth of 400MHz per channel to meet 6G pre-research testing demands.

• Integrated Sensing and Communication (ISAC) Research: The combination of wideband multi-channel design and coherent synchronization makes it ideal for verifying algorithms in merged sensing and communication scenarios.

• Millimeter-Wave Band Extension: Paired with Nuoguang’s custom MMW-44 millimeter-wave extension module, IF signals can be upconverted to 24–44GHz, supporting research in 5G FR2 and 6G mmWave directions.

• Spectrum Monitoring and Data Streaming: Equipped with dual 100GbE QSFP28 high-speed interfaces, the platform supports ultra-fast data streaming and playback, enabling precise signal monitoring and analysis across the entire frequency band.

DEEPSDR 2: Introducing Edge AI Computing Power to RF Platforms

While the LW-X410 PLUS focuses on resolving multi-channel coherent synchronization challenges, the DEEPSDR 2 represents Nuoguang’s exploration in a different direction—integrating a high-compute GPU module directly into the SDR platform to enable both RF acquisition and AI inference within a single device.

Hardware Architecture: RFSoC + Blackwell GPU for Localized Inference

The DEEPSDR 2 integrates the LW-X410 PRO software defined radio with the NVIDIA Jetson AGX Thor edge AI module inside a single industrial-grade chassis. Interconnected via a 100GbE QSFP28 link, they form a complete, seamless pipeline: RF acquisition $\rightarrow$ data transmission $\rightarrow$ edge inference.

Why Choose This Integration?

A key discussion point at this year’s MWC Shanghai centered on offloading more AI workloads to the network edge to minimize the round-trip latency of cloud-bound data. In RF scenarios, this challenge is particularly acute: conventional setups rely on an SDR for acquisition, upload the data to a server or cloud for AI inference, and then return the results. In time-critical applications like low-altitude security and real-time spectrum sensing, this round-trip delay is often unacceptable.

The DEEPSDR 2 addresses this by embedding computing power right at the acquisition point:

• Massive Local Compute: The Blackwell-architecture GPU delivers 2070 FP4 TFLOPS of sparse compute power, enabling the local execution of intensive models such as modulation recognition, protocol parsing, and UAV feature identification.

• Zero-Bottleneck Throughput: The 100GbE QSFP28 interconnect ensures real-time IQ data transfer between the RFSoC and the GPU, eliminating throughput bottlenecks.

• Rich Software Ecosystem: Fully compatible with mainstream frameworks including GNU Radio, PyTorch, TensorFlow, and TensorRT, it comes with a comprehensive SDK to facilitate secondary development.

This architectural direction garnered significant interest during its MWC showcase and represents a technology pathway we are committed to investing in long-term.

Application Implementation: Closed-Loop Verification from MWC to Real-World Scenarios