Sensing Assisted Power Control for ISAC

Introduction

Power control in cellular systems adjusts the transmit power of the gNB (downlink) and UE (uplink) to manage interference, satisfy QoS targets, and maximize energy efficiency. In 5G NR, downlink power control is managed by the gNB scheduler, while uplink power control uses open-loop (path-loss based) and closed-loop (TPC commands) mechanisms.

Conventional power control relies on UE measurements (RSRP, path loss estimates) and CQI/SINR feedback. These measurements are periodic and subject to reporting delays. Sensing-assisted power control uses real-time environmental information from sensing — such as target distances, angular positions, blockage detection, and interference source locations — to make faster and more accurate power control decisions.

 Sensing-Assisted Power Control Architecture
┌──────────────────────────────────────────────────────────────────────┐
│                                                                      │
│   ┌───────────────┐    ┌────────────────────┐    ┌────────────────┐  │
│   │   Sensing     │    │   Environment       │    │   Power        │  │
│   │   Echoes      │───▶│   Awareness         │───▶│   Control      │  │
│   │               │    │                      │    │   Engine       │  │
│   └───────────────┘    │  ● Distance to UE   │    │                │  │
│                        │  ● Blockage status   │    │  ● DL Tx power│  │
│   ┌───────────────┐    │  ● Interferer locs   │    │  ● UL TPC     │  │
│   │   UE Reports  │───▶│  ● Reflector map     │    │  ● Per-beam   │  │
│   │   (CQI, RSRP) │    │                      │    │    power      │  │
│   └───────────────┘    └────────────────────┘    └────────────────┘  │
│                                                                      │
└──────────────────────────────────────────────────────────────────────┘

Key Concepts

Distance-Based Path Loss Estimation

Sensing provides a direct range estimate to the target. Combined with a path loss model, this gives an instantaneous path loss estimate without waiting for UE RSRP reports:

\[PL(d) = PL(d_0) + 10 \cdot n \cdot \log_{10}\left(\frac{d}{d_0}\right) + X_\sigma\]

where \(d\) is the sensing-derived distance, \(n\) is the path loss exponent, and \(X_\sigma\) is shadow fading. The sensing-derived distance bypasses the need for UE-reported RSRP to estimate path loss for open-loop power control.

Blockage-Aware Power Adaptation

Sensing can detect blockage events — sudden appearance of an obstacle between the gNB and UE — before they fully impact the communication link:

 Blockage Detection and Power Adaptation
┌──────────────────────────────────────────────────────────────────┐
│                                                                  │
│   Time t₀: No blockage                                          │
│   ┌──────┐ ════════════════════════════════════ ┌──────┐         │
│   │ gNB  │  LOS path (normal power)             │  UE  │         │
│   └──────┘                                      └──────┘         │
│                                                                  │
│   Time t₁: Sensing detects approaching blocker                   │
│   ┌──────┐ ═══════ ▓▓▓(blocker)══════════════ ┌──────┐          │
│   │ gNB  │  Echo from blocker detected          │  UE  │          │
│   └──────┘  → Increase power proactively        └──────┘          │
│                                                                  │
│   Time t₂: Without sensing (reactive)                            │
│   ┌──────┐ ═══════ ▓▓▓══X (link drop!) ═════ ┌──────┐          │
│   │ gNB  │  CQI drops → power up (too late!)   │  UE  │          │
│   └──────┘                                      └──────┘          │
│                                                                  │
└──────────────────────────────────────────────────────────────────┘

Interference-Aware Power Setting

When sensing identifies the location and direction of interference sources, the gNB can:

• Reduce power towards interferer directions to minimize interference caused to neighboring cells.

• Increase power towards the UE when sensing confirms no interference risk in that direction.

• Per-beam power control: In multi-beam systems, each beam’s power can be independently set based on the sensing-derived interference map.

Uplink Power Control Enhancement

For uplink, sensing assists in:

• Path loss estimation without RSRP: The gNB estimates the UE’s path loss from the sensing-derived range, reducing the dependency on DL reference signal measurements.

• Fractional power control optimization: The fractional power control parameter \(\\alpha\) can be adapted based on sensing-derived inter-cell interference geometry.

• Timing advance from range: Sensing-derived range directly maps to timing advance, eliminating the need for RACH-based TA estimation.

Benefits

Table 12 Benefits of Sensing-Assisted Power Control

Benefit	Description
Proactive power adaptation	Blockage and path changes detected before communication is affected.
Faster convergence	Sensing-derived path loss avoids the slow convergence of closed-loop power control.
Reduced interference	Interference-aware per-beam power control reduces inter-cell interference.
Energy efficiency	Accurate path loss knowledge avoids over-provisioning transmit power.

Challenges

• Path loss model accuracy: Sensing provides distance but the path loss depends on the propagation environment (LOS/NLOS, frequency, clutter). The mapping from distance to path loss requires a well- calibrated model.

• Blockage detection latency: The sensing processing pipeline must be fast enough to detect blockage and trigger power adaptation before the communication link degrades.

• UE identification: Associating a sensing-detected target with a specific UE for per-UE power control requires additional context (e.g., beam correspondence, timing advance matching).

• Regulatory constraints: Increasing power proactively based on sensing predictions must still comply with regulatory EIRP limits and SAR constraints.