Source-linked AI summary
Lightweight CFR-Based Modulation Adaptation in a Real-Time MIMO-OFDM SDR Testbed
Luca Borst, Maryam Ansarifard, Ankith Vinayachandran, Kishor C. Joshi, George Exarchakos
TL;DR
Scalar link-quality indicators simplify link adaptation but omit frequency-selective channel structure, while richer CSI increases complexity. This paper evaluates compact BS-side CFR magnitude descriptors for modulation classification in a real-time MIMO-OFDM SDR testbed and finds higher accuracy than SNR-only features, with a lightweight decision tree deployed for real-time inference.
Problem
Scalar channel-quality indicators provide low-complexity adaptation but do not explicitly retain the frequency-selective structure of MIMO-OFDM channels, motivating a compact intermediate representation between scalar metrics and richer CSI.
Method
The paper evaluates compact BS-side CFR magnitude descriptors, alone and with SNR, using DT, RF, and KNN classifiers on 87,817 OTA observations from a USRP-based real-time MIMO-OFDM SDR testbed.
Results
CFR-based classifiers achieve 73.6%–81.4% test accuracy versus 35%–42% for SNR-only classifiers, with performance retained near the 10% BLER thresholds and RF reaching 82.8%.
Takeaways & Limitations
A depth-7 DT with 123 leaves provides a practical performance–complexity trade-off and is integrated into the LabVIEW C Node for real-time inference at 10 Hz.
Takeaways & Limitations
The evaluation is limited to stationary indoor line-of-sight conditions and magnitude-only CFR features.
Abstract
from arXiv · showhide
Conventional link adaptation typically relies on scalar link-quality indicators such as signal-to-noise ratio (SNR), while richer channel state information (CSI) can improve adaptation at the cost of higher processing complexity. This paper investigates a compact alternative for modulation selection in a real-time multiple-input multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) system using channel frequency response (CFR) magnitude descriptors. A dataset of 87,817 over-the-air (OTA) samples is collected using a USRP-based testbed, with CFR measurements extracted at the base station (BS) from received uplink pilots. Decision tree (DT), random forest (RF), and k-nearest neighbours (KNN) classifiers are evaluated using BS-side SNR, CFR features, and their combination. SNR-only classifiers achieve 35%-42% test accuracy, whereas CFR-only features achieve 73.6%, 81.4%, and 80.0% for DT, RF, and KNN, respectively. CFR-based performance is maintained near the 10% BLER reliability thresholds, with RF reaching 82.8%. A depth-7 DT with 123 leaves is further integrated into the LabVIEW C Node for real-time inference. The results show that compact BS-side CFR descriptors provide more discriminative information than the available scalar BS-side SNR while remaining suitable for lightweight SDR implementation.
I. INTRODUCTION
The paper targets the gap between low-complexity scalar link-quality adaptation and richer, more complex CSI-based approaches. It evaluates compact CFR magnitude descriptors for lightweight ML-based modulation selection in a real-time MIMO-OFDM SDR testbed.
- Motivation: Scalar SNR-based adaptation is efficient but does not retain the frequency-selective structure of MIMO-OFDM channels.Different channel realisations can have similar average SNR while differing across subcarriers and antenna pairs.
- Motivation: Full complex CSI can improve link adaptation but increases learning dimensionality and implementation complexity.Prior work using a full complex channel matrix with CNN-LSTM achieved 92.5% test accuracy.
- Approach: CFR provides an intermediate representation by compressing frequency-dependent channel information into mean, standard deviation, range, and slope descriptors.The descriptors represent average channel magnitude and its variation across frequency.
- Evaluation: The study compares BS-side SNR, CFR descriptors, and their combination using DT, RF, and KNN classifiers on OTA measurements near modulation reliability boundaries.Modulation adaptation is evaluated independently of coding-rate optimisation.
- Results: CFR-based classifiers outperform SNR-only classifiers, reaching 73.6%–81.4% test accuracy versus 35%–42%.Performance is maintained or improved near the 10% BLER reliability thresholds, reaching 82.8% for the RF CFR+SNR classifier.
- Implementation: A depth-7 DT with 123 leaves is selected for real-time deployment because its inference can be implemented as compact conditional C code.The classifier was integrated into the LabVIEW C Node without online retraining or GPU-based inference.
II. SYSTEM MODEL
This section introduces the system overview, signal model, and OTA experimental setup used to evaluate the proposed modulation-selection approach.
- II. SYSTEM MODEL: The section defines the system overview, signal model, and OTA measurement configuration for the study.The detailed hardware and measurement conditions are described in subsequent subsections.
A. System Overview
The testbed estimates the downlink channel at the BS from uplink pilots under TDD reciprocity, then uses CFR magnitude descriptors for downlink modulation selection.
- System Overview: The real-time testbed uses NI USRP-2954 SDR hardware and LabVIEW Communications in a 14-BS-antenna, 1-MS-antenna configuration.The resulting frequency-dependent channel matrix is H[k] ∈ C^Nr×N_t.
- Modulation Selection: Under channel reciprocity, the uplink-derived CFR is used to extract magnitude descriptors for the ML-based modulation selector.The selected modulation index is passed to the downlink frame scheduler and OFDM modulator.
- Channel Acquisition: Uplink pilots transmitted by the MS are processed at the BS to estimate the uplink CFR using a least-squares algorithm.The BS uses this estimate as the corresponding downlink channel representation under TDD reciprocity.
- Measurement Roles: The architecture provides CFR-derived features to the deployed decision-tree classifier, while MS-side SNR and BLER are reserved for measurement analysis and evaluation.This separates BS-side inference inputs from MS-side performance measurements.
C. OTA Measurement Setup
The OTA evaluation uses a controlled indoor line-of-sight setup with a stationary mobile station positioned 60 cm from the testbed.
- OTA Measurement Setup: Measurements are collected at 4.2 GHz over a 20 MHz channel bandwidth with the MS held stationary.The stationary configuration supports repeatable channel conditions during data acquisition.
- OTA Measurement Setup: The OTA setup uses an NI USRP-2954 testbed with 14 BS antennas and 1 MS antenna separated by 60 cm.The configuration creates a line-of-sight propagation scenario.
III. MODULATION ADAPTATION METHODOLOGY
The study collects synchronized BS-side and MS-side measurements under controlled OTA conditions, then fits BLER relationships to derive reliability regions for offline analysis.
- The testbed sweeps BS transmit power across 100 levels for each modulation and records measurements during fixed 30 s intervals.
- BS-side CFR and SNRBS are sampled at 10 Hz, while MS-side SNRMS and BLERMS are recorded at 1 Hz and aligned by timestamp.
- BS-side measurements supply classifier inputs, whereas MS-side measurements support BLER fitting, threshold derivation, and offline analysis.
- CFR magnitude is smoothed before extracting features that represent average channel quality and frequency-selective variation.
- A logistic sigmoid is fitted to each modulation’s BLERMS–SNRMS relationship, and the 10% BLER SNR defines reliability and transition regions.
B. Feature Engineering
The feature-engineering stage compresses smoothed CFR magnitude into four descriptors and compares CFR-only, SNR-only, and combined BS-side inputs.
- Four CFR descriptors capture mean magnitude, standard deviation, range, and first-order least-squares slope across subcarriers.
- The CFR descriptors are combined with BS-side SNR to form candidate feature vectors for modulation classification.
- Three feature configurations are evaluated to isolate the contribution of channel representation from classifier architecture.
- MS-side SNRMS and BLERMS are excluded from classifier inputs and reserved for threshold derivation and offline performance analysis.
- Feature importance analysis tests whether performance is associated mainly with average CFR magnitude or its variation across frequency.
C. Machine Learning-Based Modulation Selection
Modulation selection is posed as three-class supervised classification and evaluated with DT, RF, and KNN models under multiple BS-side feature configurations.
- QPSK, 16-QAM, and 64-QAM are used as the three output classes, with transmitted modulation serving as the training label.
- DT, RF, and KNN provide complementary comparisons of interpretable partitioning, ensemble robustness, and non-parametric classification.
- The dataset uses an 80% training and 20% test split, with 5-fold cross-validation for model selection and tuning.
- The classifiers are evaluated with SNR-only, CFR-only, and CFR+SNR BS-side feature configurations.
- Real-time inference runs at 10 Hz, and modulation updates require three consecutive identical predictions, corresponding to 0.30 s.
IV. EXPERIMENTAL RESULTS
The OTA evaluation uses a large filtered dataset and examines modulation reliability through BLER thresholds and BS-side SNR distributions.
- The final dataset contains 87,817 OTA observations after constraint-mask filtering.
- The measured BLERMS–SNRMS relationship is summarized with fitted sigmoid curves and 10% BLER reliability thresholds.
- The thresholds are 13.0 dB, 20.4 dB, and 27.0 dB for QPSK, 16-QAM, and 64-QAM, respectively.
- Only 0.1% of QPSK samples fall below its threshold, while 5.3% of 64-QAM samples fall above its threshold.
- BS-side SNR class means cluster at 28.2, 28.4, and 28.9 dB with within-class standard deviations of 1.7–2.3 dB, producing strong overlap.
B. Offline Classification Performance
CFR-derived features substantially improve modulation-classification accuracy over SNR-only inputs, while adding BS-side SNR to CFR provides little further benefit. RF achieves the highest offline accuracy, whereas DT offers a compact real-time deployment option.
- CFR-derived features substantially improve classification accuracy compared with SNR-only inputs across DT, RF, and KNN.
- Adding SNRBS to CFR features provides little additional benefit, indicating that CFR captures most useful discriminative information under the considered conditions.
- RF achieves the highest offline accuracy among the evaluated classifiers.
- A depth-7 DT with 123 leaves is selected by 5-fold cross-validation as a suitable performance–complexity operating point for real-time deployment.
C. Feature Importance
Feature-importance analysis attributes the classification gains primarily to compact CFR descriptors rather than scalar BS-side SNR. The mean CFR magnitude is dominant, while transition-region evaluation tests whether this advantage persists near reliability thresholds.
- The CFR+SNR feature set is analysed with DT and RF to identify which BS-side measurements drive classification improvement.
- SNRBS contributes only 1.0% for DT and 9.5% for RF, substantially less than the CFR-derived features.
- Mean CFR magnitude |H[k]| is the most influential feature, accounting for 63.3% of DT importance and 38.0% of RF importance.
- The remaining importance is distributed across CFR range, standard deviation, and slope, capturing both average magnitude and frequency-domain variation.
- Near the 10% BLER thresholds, CFR-based models maintain or improve performance, while adding SNRBS again provides limited additional benefit.
V. CONCLUSION
The study finds that compact BS-side CFR magnitude descriptors outperform SNR-only inputs for modulation classification in a real-time MIMO-OFDM SDR system. A depth-7 DT provides a practical deployment trade-off, but evaluation remains limited to stationary indoor LoS conditions and magnitude-only CFR features.
- CFR-based features consistently outperform the SNR-only baseline across 87,817 OTA observations, with little additional benefit from adding SNRBS.
- CFR-based performance is retained near the measured 10% BLER reliability thresholds, and mean CFR magnitude is the dominant input.
- A depth-7 DT with 123 leaves is integrated into the LabVIEW C Node for real-time inference at 10 Hz as a practical performance–complexity trade-off.
- The evaluation is limited to stationary indoor LoS conditions and magnitude-only CFR features.