Projects

Please download all the projects from here: Projects

• Learning to Demap: Database Generation, Preprocessing, Postprocessing, Training, Validation and Inferences from the LLRNet
  • Table of Contents
  • Import Libraries
    • Import Python Libraries
    • Import 5G Toolkit Modules
  • Learning to Demap the Symbols
    • Input Output Mapping for M = 4
    • Input Output Mapping for M = 6
    • Input Output Mapping for M = 8
  • Throughput and BER Performance of LLRnet
    • Import Libraries
    • Simulation Parameters
    • PDSCH Parameters
    • LLRnet Parameters
      • Training Framework
      • Deployment Framework
  • Simulation Section
  • Performance Evaluation
    • Throughput vs SNR (dB) for 16-QAM, 64-QAM and, 256-QAM
    • Bit Error rate (BER) vs SNR (dB) for 16-QAM, 64-QAM and, 256-QAM
    • Block Error Rate (BLER) vs SNR (dB) for 16-QAM, 64-QAM and, 256-QAM
  • Performance Evaluation: 10000 batches and 64000 training samples for LLRNet
    • Throughput vs SNR (dB) for 16-QAM, 64-QAM and, 256-QAM.
    • Bit Error rate (BER) vs SNR (dB) for 16-QAM, 64-QAM and, 256-QAM.
    • Block Error rate (BER) vs SNR (dB) for 16-QAM, 64-QAM and, 256-QAM.
    • Complexity Analysis
  • Conclusion
    • Positives of the LLRnet:
    • Limitations of the LLRnet:
  • References:
• Blockage Probability Analysis for RedCap Devices in 5G Networks
  • Analysis of Blocking Probability for different Coverage Conditions
    • Import Python Libraries
    • Import 5G Toolkit Libraries
    • Simulation Parameters
    • PDCCH Scheduling Parameters
    • PDCCH Scheduling for Good Coverage Scenarios
    • PDCCH Scheduling for Medium Coverage Scenarios
    • PDCCH Scheduling for Extreme Coverage Scenarios
    • Plotting the results
      • References
  • Variation in Blocking Probability with Different Aggregation Levels (ALs)
    • Python Libraries
    • 5G-Toolkit Libraries
    • Simulation Parameters
    • PDCCH Scheduling Parameters
      • Impact of AL 1
      • Impact of AL 2
      • Impact of AL 4
      • Impact of AL 8
      • Impact of AL 16
    • Plot the Variation in Blocking Probability with number of UEs for different Aggregation levels.
    • References
  • Analyzing the effect of Number of Candidates on Blocking Probability
    • Simulation Parameters
    • PDCCH Scheduling Parameters
    • Plot the Variation in Blocking Probability with number of PDCCH candidates
    • References
  • Analyzing the Impact of Scheduling Strategy on Blocking Probability
    • Python Libraries
    • 5G-Toolkit Libraries
    • Simulation Parameters
    • PDCCH Scheduling Parameters
    • Simulation for Scheduling Strategy-I
      • Blocking probability vs number of UEs to be scheduled.
    • Simulation for Scheduling Strategy-II
    • Plotting Blocking Probability vs Number of UEs for Scheduling Strategy
      • References
  • Analyze the Impact of UE Capability on Blocking Probability
    • Python Libraries
    • 5G-Toolkit Libraries
    • Simulation Parameters
    • PDCCH Scheduling Parameters
    • Simulating the Reference Case
      • Plot Blocking Probability for Different CORESET Sizes for Different UEs
    • Simulating Reduced Blind Decoding Case-A
    • Simulating Reduced Blind Decoding Case-B
    • Plot Blocking Probability for Different CORESET Sizes for Different UEs
      • References
  • Selection of minimum CORESET Size for a Given Target Block Probability
    • Python Libraries
    • 5G-Toolkit Libraries
    • Simulation Parameters
    • PDCCH Scheduling Parameters
    • Compute minimum coreset size for numUEs = 5.
    • Compute minimum coreset size for numUEs = 10.
    • Compute minimum coreset size for numUEs = 15.
    • Display Minimum CORESET size required to meet the Target Blocking Probability for different number of UEs.
      • References
• Artificial Intelligence and Machine Learning (AI-ML) for CSI Compression and Reconstruction in 5G Networks
  • CSI Compression and Reconstruction using CSINet for TDD Massive MIMO 5G Networks
    • Import Libraries
      • Import Python Libraries
      • Import 5G Toolkit Libraries
    • Simulation Parameters
    • Wireless Channel Generation: CDL-A
    • Reconstrunction Performance of CSI-Net
    • PDSCH Parameters
    • PDSCH: Transmitter
    • SVD Based Beamforming: Perfect CSI
    • Pass through Channel
    • Link Level Simulation: SVD based Beamforming using Perfect CSI
    • SVD Based Beamforming: CSI Reconstructed using CSINet
    • Pass through Wireless Channel
    • Link Level Simulation: SVD based Beamforming using Imperfect CSI
  • Performance Evaluations
    • Throughput Evaluations
    • BLER Evaluations
    • References
  • Wireless Channel Dataset Generation for Training the AI based Models
    • Import Python Libraries
      • Import Basic Python LIbraries
      • Import 5G Toolkit Libraries
    • Simulation Parameters
    • Set Channel Parameters and Generate Common Parameters
    • Generate the Wireless Channels Databases and Preprocess it before storage.
    • Aggregate all the Datasets into a single Dataset
    • Display Sparsity of Wireless Channels
  • Training the CSINet
    • Import Libraries
      • Import Python Libraries
    • Important AI-ML Libraries
    • Load Datasets
    • Set Training Parameters
• Comparative Study of Reed Muller codes, Polar Codes and LDPC codes
• Link Level Simulations and Lnk budget Analysis for 5G Non Terrestrial Networks
  • Coverage Evaluation of Physical Broadcast Channels (PBCH) in 5G Non-Terrestrial Networks
    • Evaluation Methodology
      • Import Libraries
    • Import Python Libraries
    • Import 5G Libraries
      • Simulation Parameters
      • Generate NTN Channel
      • Generate MIB and PBCH Configurations for NTN
      • PSS, SSS, PBCH, PBCH-DMRS and SSB Generation
      • Transmission OFDM Resource Grid
      • Pass through the Wireless Channel
      • Heatmap of Received Grid
      • Link Level Simulation: PBCH
      • Displaying the Received Noisy Resource Grid
      • Displaying Noisy SSB Grid
      • Performance Evaluations: SNR vs BLER
  • Link Level Simulation for Physical Downlink Control Channels in 5G Non-Terrestrial Networks (NTN)
    • Import Libraries
      • Import Basic Python Libraries
    • Import 5G-Toolkit Libraries
    • Simulation Parameters
    • CORESET Parameters
    • Generate Wireless Channel: NTN-TDL-C
    • Link level Simulation: For each Aggregation level and Each SNR value
    • Reliability Performance: BER/BLER vs SNR
  • Link Level Simulation for Physical Downlink Shared Channel (PDSCH) in 5G Non Terrestrial Networks (NTN)
    • Import Python Libraries
    • Import 5G-Toolkit Libraries
    • Simulation Parameters
    • Generate Channel
    • PDSCH Configurations
    • PDSCH Implementation
    • Pass through the Wireless Channel
    • Recevier Side Processing
    • Simulation Results: Reliability
    • Simulation Results: Throughput
    • Simulation Results: Reliability (Averaged over 32000 batches)
    • Simulation Results: Throughput (Averaged over 32000 batches)
• Hybrid Automatic repeat Request in 5G and Beyond
• Constellation Learning in an AWGN Channel
  • PHY layer as AutoEncoder
  • Steps
    • Importing Libraries
    • Parameters of AutoEncoder
    • Training Data
    • Testing Data
    • Normalization Functions
    • Defining AutoEncoder Model
    • Training AutoEncoder
    • Defining Tx, Channel and Rx from Trained AutoEncoder
    • Block Error Rate (BLER) performance
    • Hamming Codes
  • Transmitter
  • BLER plot : comparison of AutoEncoder BLER with base line (n,k) Hamming Code BLER
    • Constellation Learning
  • learned constellation plot
    • References
• Downlink Synchronization using SSB in 5G systems
• Uplink Synchronization using PRACH in 5G systems
• Performance comparison between different Positioning Methods for millimeter wave 5G Networks
  • Import Libraries
    • Import Python Libraries
    • Import 5G Toolkit Libraries
  • Simulation Parameters
  • Generate Wireless Channels
  • SRS Configurations
  • Slot by Slot Simulation
  • Position Estimation: Based on UL-ToA
  • Visualization of Estimated Position
  • Performance Analysis of Positioning Error for ToA based method
  • Positioning Results Averaged over 2000 UEs