SoC FPGA for IoT Edge Computing
One of the reasons for the explosive growth of IoT is that embedded devices with networking capabilities and sensor interfaces are cheap enough to deploy them at a plethora of locations.
However, network bandwidth is limited. Not only that, but also, the latency of the network can be of seconds or minutes. By the time the sensor data is acquired by the centralized computers, its value for decision making could be lost. In other words, for the IoT solution to be effective, it should not only deliver meaningful data securely (and filter it as much as possible to avoid network congestion), it should also analyze it and act upon it at the origination point of the data. At the very edge of the network.
Applications for edge computing that call for intelligent sensors include:
- Smart buildings
- Autonomous transportation
- Machine control
- Healthcare
- Augmented reality
- Voice, image and video recognition
IoT edge applications can be implemented based on ASICs, FPGA and CPUs. SoC FPGAs combine the advantages of the last two, namely, CPU and FPGA.
A typical SoC these days include a powerful processor and FPGA. Both Xilinx and Altera are offering SoCs based on ARM Cortex multicore processors and hundreds of thousands of processing logic elements, embedded memory and DSP blocks.
The combination of CPU and FPGA brings the following advantages:
- Increased flexibility and reconfigurability
- CPU offloading of data intensive processing where FPGAs excel:DSP algorithms parallel implementationData filtering
- FFT analysis
- Software Defined Radio - SDR
- GPS
- Slow and fast sensors data acquisition, response, filtering and analysisSPI
- I2C
- RS-232, RS-485
- Bluetooth
- Physical and low layers implementation of diverse networking protocolsEthernet
- WiFi
- Security implementation (data encryption and data access validation)
In the last ISDF (Intel SoC Developer Forum), in 2016, Fujisoft presented its current solution for Edge Computing based on Intel/Altera SoC FPGA, as well as their vision for the future of SoC FPGAs for IoT Edge applications. In the future, their plan is to use the growing capabilities of FPGAs with embedded powerful CPUs and high data bandwidth memory access to implement, manage and reconfigure the Edge solutions based on Genetic and Deep Learning algorithms.
For additional reading:
Edge computing (Wikipedia)
Edge computing, the door to IoT data kingdom (GE)
Why edge computing is crucial for the IoT (RTI)
For additional articles about the FPGA field, including technical news and hands-on projects, please visit my site