Instructions to Use CIC demodulation on LoRa Packets

1- open param_configs file and set SF, BW, Receiver Sampling Frequency and num of data symbols in a LoRa pkt accroding to your LoRa pkts' and receiver configuration. set appropriate path to the binary file that contains I/Q samples of LoRa signal in the variable named 'path' & in the variable named 'fil_nm' set the appropriate binary file name to be read by matlab.

2- Open main.m file -> main.m file has been divided into following sections - %% Loading variables - %% Loading File - %% Active Sessions Detection using Dechirping Windows - %% DC correlations to find LoRa pkts out of collision - %% UC correlation to filter false positives and frequency offset & Packets' SNR estimation - %% Data Demodulation using CIC - %% Calculating SER for the File

3- Run whole main file, or run the file section by section avoiding for loop iterating over 'm' (choose an appropriate active session yourself) (while running each section, comment 'continue' statements in 'if' conditions in sections '%% DC correlations to find LoRa pkts out of collision' and '%% UC correlation to filter false positives and frequency offset & Packets' SNR estimation' because 'continue' statement can only be called in a for loop)

4- In the current folder, 'sym.mat' is supposed to contain 'sym' variable that contains true symbols against which output of CIC is to be compared in order to compute 'Symbol Error Rate' or 'Symbol Throughput'

5- The Symbol Error Rate/Symbol Throughput is computed and printed on command window. At the end, 'demod_sym_stack' (2D matrix) variable can be found. The number of rows indicate the packets detected in the file. At each row, lies a demodulated packet whose column contain corresponding demodulated symbols. The demodulated symbols are also saved in 'symbols.txt' file, in a format(each row has symbol and packets are separated by -1 identifier) that is needed to feed symbols to rpp0.

DATA********************* 3 DataSets have been provided and can be accessed via following link https://uwprod-my.sharepoint.com/:f:/g/personal/mshahid2_wisc_edu/Es8DfWbahW1CpBdjv2uSdA8BdVTdkf0BblOOB472O8XtOQ

Naming Convention: '# of LoRa Tx in network'_'individual node's pkt rate'lm_1 e.g. 20tx_1lm_1 -> 20 = '# of LoRa Tx in network' 1lm = 'individual node's pkt rate' Experimental Setup has been Described in the Experimental_setup.pdf.

-> Run pip install -r requirements.txt. -> Open param_configs.py and set SF, BW, Receiver Sampling Frequency and num of data symbols in a LoRa pkt accroding to your LoRa pkts' and receiver configuration. -> Open main.py and set appropriate path to the binary file that contains I/Q samples of LoRa signal in the variable named in_file_path. -> (Optional) In main.py, set symbols_ground_truth_path to the path to the text file containing the expected symbols in the input file. The ground truth file should contain rows of space-separated symbols, where each row corresponds to a packet. See symbols_ground.txt for an example of the format. -> Run python main.py. Demodulated symbols will be written to the file located at symbols_out_path, and peaks will be written to the file peaks_out_path.

-> For Pkt throughput, we have used rpp0/gr-lora that can be installed along with dependencies as mentioned in the following 'rpp0' repository https://github.com/rpp0/gr-lora

-> gr-lora contains two modules. 1- Standard LoRa Demodulator and 2- Decoder LoRa Demodulator takes in I/Q samples and demodulates these samples based on default LoRa to give symbols at the output, these symbols are then passed to Decoder that gives the decoded bits at the output. Since, We have our own CIC-based LoRa Demodulator, we run CIC to get the symbols and then pass our symbols output to gr-lora's decoder to get bits at the output (thus, bypassing gr-lora's demodulator). The Code to bypass the gr-lora's demodulator can be found in the 'Decoder' folder. Look for 2 files'decoder_impl_mod.cc' & 'decoder_impl_mod.h'.

-> After setting up the gr-lora move to folder /gr-lora/lib/. This contains 2 files 'decoder_impl.cc' and 'decoder_impl.h'. Copy the contents of 2 files in the Decoder folder of CIC and then paste it to the above mentioned files (copy contents of .cc to .cc and .h to .h files).

-> This modified version reads a text file 'symbols.txt' as an input that contains demodulated symbols from CIC (file format: each row of txt file contains a single symbol and packets are separated by -1 identifier) and outputs decoded bits corresponding to each packet on a row in a text file.

-> Add correct paths to the input and output text files in the LoRa-Class constructor.

-> Get the Decoded bits in the relevant bits output file 'bin_out.txt'.