Reducing the Cost of Implementing Filters in LoRa Devices

Area of Science:

• Wireless Communication
• Signal Processing
• Embedded Systems

Background:

• LoRa (Low-Power Long-Range) devices require efficient implementation of pulse shaping filters.
• Multiplier-less filter designs offer memory savings but face challenges in LoRa systems.

Purpose of the Study:

• To present methods for optimizing LoRa devices for economical multiplier-less pulse shaping filter implementation.
• To reduce the memory footprint of LoRa devices through filter optimization.

Main Methods:

• Chirp segmentation for efficient generation of basic chirp waveforms, reducing ROM storage requirements.
• Quantization of basic chirp samples to decrease unique input values and lookup table size for filters.

Main Results:

• Chirp segmentation reduces sample storage to a quarter.
• Quantization significantly reduces lookup table size for multiplier-less filters.
• Simulated LoRa system tests show minimal performance degradation (occupied bandwidth, FFT, bit-error rates) even with high quantization levels.

Conclusions:

• Chirp segmentation and quantization effectively reduce memory requirements in LoRa devices.
• These methods improve the feasibility of implementing multiplier-less filters in LoRa systems.
• Optimized filters maintain system performance, making LoRa more memory-efficient.