ASN Filter Benchmarks.pdf

• Infinite Impulse Response (IIR): chosen for applications where linear phase is not crucial and memory is limited, with lower implementation costs.
• Finite Impulse Response (FIR): chosen for applications where linear phase is important and memory/computational performance are available.

The ASN Filter Designer's ANSI C SDK framework provides a comprehensive code base for deploying filtering applications to any C/C++ embedded platform.
Benchmark testing was conducted on microcontrollers, including:
1. STM32F103C8 – Arm Cortex-M3
2. STM32F469NI – Arm Cortex-M4F
3. STM32F769NI – Arm Cortex-M7F
4. ESP32-WROOM-32 – Espressif (Dual-core Xtensa 32-bit LX6)

The ASN-DSP SDK framework offers several advantages, including:
1. Rapid development, testing, and deployment of DSP filtering applications.
2. Open-source and agnostic code base for maximum performance.
3. Support for complex filters and low memory requirements.
4. Signal analysis tools for testing filter performance and accuracy.
5. Extra filtering functions, such as median, moving average, and non-linear functions.
6. Support for single and double precision floating-point arithmetic.

The ASN Filter Designer software generates C code for benchmarking filter performance, and the SDK is compared to Arm's CMSIS-DSP library.
An 8th-order lowpass Biquad IIR filter was designed using the ASN Filter Designer for benchmark testing.
The results show the performance of the ASN-DSP SDK framework on different Arm cores and the ESP32 microcontroller.