Sensing Assisted Resource Allocation for ISAC
Introduction
Resource allocation in OFDM-based cellular systems involves distributing time-frequency resources (RBs), transmit power, spatial layers (MIMO ranks), and MCS across users to maximize system throughput, fairness, or other objectives. In conventional systems, the scheduler relies on CSI reports, buffer status reports, QoS requirements, and traffic patterns to make allocation decisions.
In an ISAC system, the scheduler must additionally allocate resources for sensing operations while ensuring that communication performance is not degraded. Sensing-assisted resource allocation uses information derived from sensing — such as target locations, velocities, and environmental awareness — to make smarter allocation decisions for both communication and sensing.
Sensing-Assisted Resource Allocation Framework
┌────────────────────────────────────────────────────────────────────────┐
│ │
│ ┌───────────────────────────────┐ ┌──────────────────────────────┐ │
│ │ Sensing Subsystem │ │ Communication Subsystem │ │
│ │ ┌──────────┐ ┌────────────┐ │ │ ┌──────────┐ ┌──────────┐ │ │
│ │ │ Target │ │Environment │ │ │ │ CSI │ │ QoS / │ │ │
│ │ │ Detection│ │ Map │ │ │ │ Reports │ │ Traffic │ │ │
│ │ └────┬─────┘ └─────┬──────┘ │ │ └────┬─────┘ └────┬─────┘ │ │
│ │ │ │ │ │ │ │ │ │
│ └───────┼─────────────┼─────────┘ └───────┼────────────┼───────┘ │
│ │ │ │ │ │
│ ▼ ▼ ▼ ▼ │
│ ┌───────────────────────────────────────────────────────────────┐ │
│ │ Joint Resource Scheduler │ │
│ │ ┌────────────┐ ┌────────────┐ ┌────────────┐ │ │
│ │ │ Time-Freq │ │ Power │ │ Spatial │ │ │
│ │ │ Allocation │ │ Allocation │ │ Allocation │ │ │
│ │ └────────────┘ └────────────┘ └────────────┘ │ │
│ └───────────────────────────────────────────────────────────────┘ │
│ │
└────────────────────────────────────────────────────────────────────────┘
Key Concepts
Joint Sensing-Communication Resource Partitioning
ISAC systems must share spectrum between sensing and communication. Several strategies exist:
Resource Partitioning Strategies
┌────────────────────────────────────────────────────┐
│ │
│ Strategy 1: Time Division │
│ ┌──────┬──────┬──────┬──────┬──────┬──────┐ │
│ │ Comm │Sense │ Comm │ Comm │Sense │ Comm │ │
│ └──────┴──────┴──────┴──────┴──────┴──────┘ │
│ Slot 1 Slot 2 Slot 3 Slot 4 │
│ │
│ Strategy 2: Frequency Division │
│ ┌──────────────────────────────────────┐ │
│ │ Comm RBs │ Sensing RBs │ │
│ └──────────────────────────────────────┘ │
│ │
│ Strategy 3: Fully Shared (Dual-Function) │
│ ┌──────────────────────────────────────┐ │
│ │ ISAC Waveform (Joint Design) │ │
│ │ Comm + Sensing on same resources │ │
│ └──────────────────────────────────────┘ │
│ │
└────────────────────────────────────────────────────┘
Sensing-Informed Scheduling
Sensing provides environmental awareness that can improve scheduling decisions:
User proximity detection: Sensing can detect UEs or objects near the cell edge, informing inter-cell coordination or scheduling priority adjustments.
Clutter-aware allocation: Sensing identifies resource blocks heavily affected by clutter; the scheduler can avoid allocating these RBs to interference-sensitive UEs.
Velocity-aware MCS selection: Sensing-derived velocity information helps the scheduler choose an appropriate MCS that accounts for channel aging within the TTI.
Spatial reuse: Sensing-derived angular information enables the scheduler to spatially multiplex users in non-overlapping angular regions.
Dynamic Sensing Resource Adaptation
The sensing resource budget should adapt based on the sensing task requirements:
Target tracking mode: When targets are being tracked (already detected), fewer sensing resources are needed compared to the initial detection phase.
Event-triggered sensing: Sensing resources can be allocated on-demand when triggered by communication events (e.g., handover, beam failure) rather than periodically.
QoS-aware trade-off: The scheduler balances sensing accuracy requirements against communication QoS constraints, allocating sensing resources only when communication headroom permits.
Benefits
Benefit |
Description |
|---|---|
Improved spectral efficiency |
Environmental awareness leads to better-informed scheduling decisions. |
Adaptive sensing overhead |
Sensing resources dynamically scaled based on actual sensing needs. |
Interference-aware scheduling |
Clutter and interference maps derived from sensing improve frequency assignment. |
Fairness in ISAC |
Balanced resource sharing prevents sensing from starving communication users. |
Challenges
Optimization complexity: Joint sensing-communication resource allocation is a multi-objective optimization problem that is generally NP-hard.
Real-time requirements: Scheduling decisions must be made within sub-millisecond timescales, limiting the complexity of optimization algorithms.
Conflicting objectives: Maximizing sensing performance (e.g., wide bandwidth, repeated measurements) often conflicts with maximizing communication throughput.
Standardization: No existing 3GPP framework for joint sensing-communication scheduling; new MAC layer procedures and signaling may be required.