Source-linked AI summary
Physiological Information Reliability: Cross-Layer Adaptive Resource Allocation for Cardiovascular Sensing
Navaneeth Krishnan Kamalakannan, Janakiraman Kamalakannan, Harinisri Velmurugan
TL;DR
Cardiovascular sensing must preserve clinically useful information under degradation, wireless loss, energy limits, and computation latency. PIR jointly models physiological information with wireless, energy, and compute states and uses a contextual bandit to adapt coding and placement; PIR-LinUCB reaches a low-power operating point while meeting latency constraints, but the evaluation relies on synthetic data and proxy utility.
Problem
Existing reliability approaches separate signal quality, delivery, energy, and computation, despite ECG/PPG windows carrying unequal physiological value.
Method
PIR combines physiological information value with wireless, energy, and compute states, then uses contextual-bandit control for coding redundancy and compute placement.
Results
11.2 mW: PIR-LinUCB achieves the lowest steady-state power while all policies attain 100% deadline-success; its reward is narrowly ahead of Heuristic-Rule.
Takeaways & Limitations
PIR-LinUCB reaches heuristic-competitive low-power decisions, while higher redundancy provides better raw accuracy at 34% more energy.
Takeaways & Limitations
The evaluation uses synthetic ECG/PPG data, simulated system parameters, and UPIR as a proxy rather than independently validated clinical information agreement.
Abstract
from arXiv · showhide
Cardiovascular sensing systems must preserve clinically useful information despite signal degradation, wireless losses, energy constraints, and edge-computation latency. We introduce Physiological Information Reliability (PIR), a cross-layer framework that represents physiological information value jointly with wireless, energy, and computation states and uses a contextual bandit to adapt sensing and communication decisions. We integrate multimodal ECG/PPG signal-quality estimation with physiological information value and an adaptive network-coding layer under burst-erasure conditions. Across controlled multiseed experiments, PIR-LinUCB demonstrates a promising low-energy operating point while maintaining medical latency constraints and competitive physiological estimation performance relative to fixed and heuristic policies. We analyze the resulting accuracy-energy-latency trade-offs and identify limitations of proxy PIV estimation and simulated communication dynamics. These results provide an initial computational demonstration of physiological-information-aware resource allocation and motivate future clinical and real-channel validation.
1. Introduction
PIR reframes cardiovascular sensing reliability around the physiological value of data, not merely packet delivery. It combines physiological, wireless, energy, and compute states in a learned cross-layer controller.
- Conventional reliability treats delivery as whether bits arrive, although ECG/PPG windows differ in clinical value.
- Prior work largely separates signal quality, transmission reliability, semantic relevance, energy, and computation.
- PIR allocates wireless, energy, and compute resources according to physiological information at stake rather than channel conditions alone.
- PIR represents medical, wireless, energy, and compute substates jointly as Xt and uses a contextual bandit to select coding redundancy and compute placement.
- The study compares PIR-LinUCB with four non-learned baselines using controlled multiseed experiments and reports both benefits and limits of synthetic proxy evaluation.
2. Related Work
Related work spans physiological signal quality, task-aware communication, network coding, and contextual-bandit allocation. PIR connects these strands into a physiological-information-aware control objective for one closed-loop sensing pipeline.
- Physiological signal quality: Multimodal signal-quality work evaluates ECG/PPG reliability across graded and complete-modality degradation regimes.
- Task-aware information preservation: Task-aware communications prioritizes decision-relevant meaning over raw bits, whereas PIR applies this principle to physiological sensing.
- Wireless medical sensing: RLNC provides algebraic resilience to packet erasures, and adaptive peer clustering extends it to Gilbert–Elliott burst-erasure channels.
- Adaptive, ML-based resource allocation: The framework combines ECG/PPG information estimation, adaptive network coding, and contextual-bandit decisions over sensing, coding, and placement.
- PIR's distinction: PIR treats physiologically valuable information as an explicit joint objective over sensing quality, wireless state, energy, and compute placement.
3. The PIR Framework
The PIR framework forms a joint state from physiological, wireless, energy, and compute information, then selects coding redundancy and compute placement using a contextual-bandit reward. Its utility values successful decoding according to physiological information while penalizing energy, latency, and power violations.
- Joint state: The medical substate contains per-modality SQI, modality-presence flags, and a scalar PIV, with missing modalities excluded from the quality average.
- Joint state: PIV combines presence-weighted signal quality with cross-modal heart-rate disagreement, which is considered only when both modalities are present.
- Joint state: Wireless state follows a Gilbert–Elliott channel, while energy tracks battery and packet costs and compute state includes local, edge, and transmission latency.
- Action and reward: The controller chooses among six actions combining RLNC redundancy R ∈ {2, 8, 20} with local or edge placement.
- Action and reward: The utility proxy UPIR = PIVt · Pdec estimates physiologically valuable information successfully decoded, but is not independently validated clinical-information agreement.
- Action and reward: Softplus penalties remain near zero within the 150 ms latency deadline and 100 mW power cap, then increase after violations.
- Controller: PIR-LinUCB maintains a ridge posterior for each action and updates it from observed contextual rewards.
4. Experimental Setup
The evaluation uses synthetic multimodal ECG/PPG streams, a burst-erasure wireless model, documented wearable energy and compute parameters, and continuous sequential sessions across five seeds. PIR-LinUCB is compared with fixed, placement-pinned, and heuristic policies using accuracy, energy, latency, and delivery metrics.
- Data: Each seed contains 1,200 sequential 8-second windows across six ECG/PPG degradation regimes generated at 125 Hz.
- Data: Graded degradation adds scaled noise, baseline wander, muscle artifacts, and motion bursts, while missing modalities become structureless noise floors.
- Wireless: All policies share a two-state Gilbert–Elliott channel with common random numbers, ensuring identical wireless randomness across comparisons.
- Energy / compute: The setup models a BLE-class radio, on-body MCU, edge target with +12 ms round-trip time, a 150 ms deadline, and a 100 mW power cap.
- Baselines: Baselines include Fixed-Rate, Always-Local, Always-Edge, and Heuristic-Rule alongside PIR-LinUCB.
- Protocol and metrics: Policies run continuous sessions per seed, with steady-state results reported as means and 95% confidence intervals across five seeds.
5. Results
PIR-LinUCB achieves a low-power operating point while meeting the deadline constraint, and its Pareto position reflects a trade-off between accuracy, energy, and latency. Its reward advantage over Heuristic-Rule is small, while higher redundancy improves raw accuracy at greater energy cost.
- 100% deadline-success was achieved by all five policies, with PIR-LinUCB using the lowest steady-state power at 11.2 mW.Fixed/Heuristic used 12.0 mW, while Always-Local/Edge used 15.0 mW.
- PIR-LinUCB and Heuristic-Rule lie on the accuracy–power and energy–latency Pareto frontiers, while Fixed-Rate is dominated on both axes.
- R=20 redundancy improves raw accuracy but consumes 34% more energy, illustrating the intended accuracy–energy trade-off.
- PIR-LinUCB’s mean reward is 0.205 versus 0.204 for Heuristic-Rule, a gap smaller than either policy’s 95% CI half-width.The comparison is reported as an effect size rather than a significance claim.
6. Findings
The findings support PIV as a cross-layer signal and show that adaptive allocation can reduce energy without violating latency. However, the learned policy remains heuristic-competitive rather than dominant, and reactive evaluation leaves predictive extensions motivated.
- PIV tracks reward across all six degradation regimes, consistent with decision-relevant information about downstream task difficulty.
- PIR-LinUCB has the lowest steady-state energy and power of all five policies while meeting the 150 ms deadline on 100% of windows.
- PIR-LinUCB’s reward edge over Heuristic-Rule is within noise, and its raw accuracy trails redundancy-maximizing baselines.The reported advantage is heuristic-competitive with less hand-tuning, not dominant.
- Because the controller uses only the last-observed channel state, it cannot exploit short-horizon structure in the bursty Gilbert–Elliott channel.This motivates model-predictive or forecasting extensions.
7. Limitations
The evaluation is an initial computational demonstration limited by synthetic data, proxy measures, simulated system parameters, limited statistical power, and missing oracle, ablation, and real-world validation.
- Scope and validation: The bandit-facing pipeline is evaluated on synthetic ECG/PPG waveforms rather than recorded patient data and is not itself validated against real recordings.The underlying signal-processing components were separately validated against real ICU and fetal-ECG recordings, but that validation does not extend to this pipeline.
- Proxy measures: UPIR is a proxy defined as PIV multiplied by decode probability, not an independently validated measure of clinical information agreement.The event-F1 metric also uses a bradycardia/tachycardia threshold rather than clinically adjudicated arrhythmia labels.
- System modeling: Wireless, energy, and compute constants are physically motivated and documented but are not fitted to a specific real radio or MCU.This limits direct interpretation of the simulated communication and hardware dynamics.
- Statistical power: With 5 seeds, several pairwise differences, notably PIR-LinUCB versus Heuristic-Rule, do not clear a strict significance threshold.The study reports effect sizes with confidence intervals rather than significance claims for these comparisons.
- Unresolved validation: The submission does not evaluate an oracle upper bound or state ablations, and it leaves real-channel and hardware validation for future work.The proposed next steps include dropping PIV, wireless, energy, or compute context in turn and testing on real channels and hardware.
Appendix A. Extended Configuration
The configuration fixes LinUCB, reward, compute, and PHY parameters used to instantiate the PIR controller and system model.
- Bandit configuration: LinUCB uses α=0.3, ridge λ=1, and a 10-dimensional context.The context includes per-modality SQI/presence, PIV, normalized SNR, EWMA erasure estimate, link state, battery fraction, and bias.
- Reward configuration: Reward weights are λE=4.0 J−1, λL=0.05 ms−1, and λP=0.02 mW−1, with softplus β=0.5.These weights define the relative energy, latency, and power terms in the reward configuration.
- System model: The compute model assumes 8 GF(256) cycles/op, 2×10^6 inference cycles, 6.4× that for edge execution, and a 250 kbps PHY rate.All parameters and their physical justification are documented in the released configuration file.
Appendix B. Per-Regime Breakdown
Across degradation regimes, PIR-LinUCB’s reward and heart-rate estimation are best in cleaner or single-missing-modality conditions and worst when both modalities are degraded, paralleling mean PIV.
- PIV by regime: 0.739 mean PIV in both-clean conditions falls to 0.543 in both-degraded conditions across policies.The reward ordering tracks this PIV ordering across every policy, not only the bandit.
- Heart-rate estimation: 3.5–3.8 bpm HR-MAE in clean and missing-modality regimes rises to 14.4 bpm under both-degraded conditions.In this synthetic protocol, graded noise is harder for heart-rate estimation than modality absence.