Performance analysis of LoRa networks with Successive Interference Cancellation.
The increase of the connected Internet-of-Things (IoT) devices, necessitates the development of low-power wide-area networks (LPWANs). LoRa technology is a popular and promising solution for communication in LPWANs. This thesis studies the use of Successive Interference Cancellation (SIC) to decode received signals in LoRa networks. The proposed method utilises the capture effect and designs a new receiver structure, capable of retrieving collided packets using SIC. A complete theoretical analysis is provided and closed-form expressions are derived for the successful decoding of packets via SIC taking path loss, fading, noise, and interference into account. The theoretical results are validated by Monte Carlo simulations.
A circular coverage region is considered (R km radius) where there is only one gateway and multiple end devices. The baseline system's model performance is given in the figures below for two different cell radius, 6 km and 12 km. The number of end devices remains constant.
In the figures solid lines are obtained via numerical evaluation, whereas the markers represent simulation results.
- Connection Probability: A node is connected to the gateway if the SNR (Signal-to-Noise Ratio) is above a default threshold.
- Capture Probability: For a successful reception in the presence of interference, the SIR (Signal-to-Interference Ratio) must exceed the capture threshold.
- Coverage Probability: A node is in coverage if it is connected to the gateway and there is no collision.
The SIC technique is used to decode only one interfering message, even if there are more and the system's performance is evaluated under specific circumstances. Since SIC is a method for mitigating interference, the connection probability will not be affected.
Applying SIC increases the capture probability up to 23.6%.
SIC increases the coverage probability up to 20% when the system radius is 6 km. A smaller improvement is observed at 12 km, since the noise (connection probability) is the primary cause of drop in performance.
The basic_script.m implements the Monte Carlo simulation of the proposed system which validates the mathematical analysis. Executing the script generates the aforementioned plots for a specific radius.