Software Defined Radio (SDR) plays a critical role in modern wireless systems are designed by shifting radio functionality from fixed hardware implementations to flexible, software-driven architectures. This article presents an overview of the SDR framework and highlights iWave’s SDR solutions based on Zynq UltraScale+ MPSoC and RFSoC platforms. It compares two key offerings ZU7 MPSoC with JESD-based converters and ZU48DR RFSoC and outlines their architectural differences, advantages, and application suitability.
What is an SDR framework?
An SDR framework is the overall hardware + software architecture that enables users to build, control, and scale a radio system mostly in software instead of fixed hardware. It minimizes analog components near the antenna, pushing processing to digital domains for adaptability and reduced costs
An SDR framework defines,
- How RF data is captured?
- How data moves between blocks?
- Where signal processing runs?
Key Requirements of an SDR Framework for building SDR solution
- SDR framework must support high-throughput I/Q data streaming with lowest possible latency
- SDR framework should include a rich set of signal-processing building blocks (DDC/DUC, FFTs, filters, modulation, synchronization) that can be composed into pipelines
- SDR framework should support for real-time execution, scalability, and determinism
- SDR framework should provide debug, visualization, and deployment tools so users can develop and test the solutions
SDR Framework Architecture
An SDR framework typically consists of three major building blocks:
- RF Data Conversion
- High-speed ADCs convert RF signals to digital domain
- DACs reconstruct RF signals from processed data
- Programmable Processing
- FPGA fabric enables real-time DSP operations
- Supports functions such as filtering, modulation/demodulation, beamforming
- Implements the logic to move the large set of data between the blocks
- Software Control
- Embedded processors manage configuration and control
- Enables protocol stack implementation, analysis and visualization
SDR Framework Based on Zynq RFSOC
ZU48DR RFSoC Platform: Ideal for high-performance, wideband systems requiring tight integration and reduced latency. Features of this framework includes,
- FPGA Chipset: AMD RFSOC
- Channels: Up to 16 × 16 MIMO
- Tuning Range: DC up to 2.3 GHz Direct sampling
- Instantaneous Bandwidth: Up to 2 GHz
- ADC sampling rate: Up to 9 GSPS, 14 bits
- DAC sampling rate: Up to 10 GSPS, 14 bits
- PCIe Bandwidth: Up to 128Gbps with Gen 3, x8 lane
- Form Factor: Full-Height Half-Length PCIe
- Multi-tile synchronization (MTS) with SYSREF inputs
- Configurable number of channels
- Dynamic tuning of the decimation and interpolation
- Burst mode or continuous modes data streaming
- GNU Radio based visualization
- Frequencies, channels, and signal chains configuration through host drivers
- Robust Linux and Windows** host OS support
- Custom drivers, open-source SDR frameworks, GPU acceleration
More details on the Unified Software and Acceleration software stack is available here.
SDR Based on MPSOC with JESD Data converter
Zynq MPSoC Platform is best suited for applications requiring RF flexibility and custom front-end configurations. Features of this framework includes,
- FPGA Chipset: AMD ZU7 MPSOC
- RF Transceiver: AD9361
- Channels: RF 2 × 2 transceiver with integrated 12-bit DACs and ADCs
- TX band: 47 MHz to 6.0 GHz
- RX band: 70 MHz to 6.0 GHz
- Tuneable channel bandwidth: <200 kHz to 56 MHz
- PCIe Bandwidth: 64Gbps with Gen 3, x8 lane
- Form Factor: Full-Height Half-Length PCIe
- Multi-tile synchronization (MTS) with SYSREF inputs
- Configurable number of channels
- Dynamic tuning of the decimation and interpolation
- Burst mode or continuous modes data streaming
- GNU Radio based visualization
- Frequencies, channels, and signal chains configuration through host drivers
- Robust Linux and Windows*** host OS support
- Custom drivers, open-source SDR frameworks, GPU acceleration
More details on the Unified Software and Acceleration software stack is available here.
SDR Framework based on the Versal RF
The SDR framework on the iWave Versal RF development kit, currently in its final stages of development, represents an advanced evolution of earlier RFSoC-based SDR platforms, building on their tightly integrated ADC/DAC, FPGA, and processor architecture. While RFSoC platforms enabled compact single-chip SDR solutions with high sampling rates and low-latency processing, the Versal RF platform further enhances this capability by integrating AI Engines, increased DSP compute, and next-generation programmable logic to support more sophisticated and compute-intensive signal processing applications.
This translates into an SDR framework that supports,
- More complex and adaptive algorithms such as beamforming, spectrum analysis, and AI-driven signal classification within the same platform.
- The availability of heterogeneous compute (Arm cores + FPGA + AI Engines) enables partitioning of SDR functions across hardware and software more efficiently than RFSoC-based designs.
- The framework benefits from higher bandwidth data paths, multi-GSPS RF converters, and high-speed interfaces, enabling wider instantaneous bandwidth and improved real-time performance compared to earlier RFSoC solutions.
Highlights of iWave SDR Framework
iWave’s Software Defined Radio ecosystem is designed around a modular System-on-Module (SoM) approach, enabling flexibility, scalability, and reduced time-to-market.
- Modular Design
- SoM + carrier board architecture
- Pre-validated high-speed interfaces (JESD204, PCIe, Ethernet)
- Simplified hardware design for end users
- Scalability
- Pin-compatible SoMs across multiple device families
- Easy migration between MPSoC and RFSoC platforms
- High-Performance Signal Chain
- Multi-GSPS ADC/DAC support
- FPGA-based DSP acceleration
- ARM processors for control and applications
- Development Ecosystem
- Linux/PetaLinux BSP support
- FPGA design using Vivado
- ODM support for customization
Comparison: RFSoC vs MPSoC
| Feature | Zynq MPSoC + JESD | Zynq RFSoC |
|---|---|---|
| RF Conversion | External | Integrated |
| Flexibility | High | Moderate |
| Integration | Medium | Very High |
| Latency | Higher | Lower |
| Bandwidth | Converter dependent | Multi-GHz |
| Complexity | Higher | Lower |
Conclusion
The evolution of SDR platforms toward higher integration and performance is enabling next-generation wireless applications. iWave’s SDR framework leverages both MPSoC and RFSoC technologies to address diverse requirements from flexible prototyping to high-performance deployment. By adopting a modular SoM-based approach, iWave ensures scalability, reduced development risk, and faster time-to-market, making it a strong foundation for modern SDR solutions.
For more information about our SDR Framework Platforms, contacts us at mktg@iwave-global.com
