Next-generation wireless networks (6G) aim to transition from pure data pipelines into dual-purpose engines capable of both transmitting information and mapping their physical surroundings. This paradigm, known as Integrated Sensing and Communication (ISAC), promises to turn base stations and access points into radar systems.
However, a fundamental physical and mathematical tension underpins this integration: a radio wave optimized to deliver maximum data throughput looks radically different from a wave optimized to precisely locate an object. When these two functions share the same spectrum, hardware, and power budget, they enter a zero-sum competition. Understanding why these systems cannot perfectly serve both masters requires looking at the structural, geometric, and mathematical mismatches between information delivery and environmental radar sensing.
The Metric Disparity: Information vs. Estimation
The core reason why ISAC optimization is challenging stems from a fundamental mismatch in what each function tries to maximize.
Communication theory relies on Shannon capacity and mutual information—metrics that quantify how reliably random data symbols can be decoded over a noisy channel. Radar sensing, conversely, is an estimation problem. It cares about minimizing errors when calculating physical parameters like distance, velocity, and angle. This accuracy is typically bounded by the Cramér-Rao Bound (CRB) or Mean Squared Error (MSE).
As Lei Xie et al. point out in Sensing Mutual Information with Random Signals in Gaussian Channels, communication signals are inherently random, whereas traditional radar waveforms are deterministic and carefully structured to avoid range ambiguities. To bridge this gap, the authors utilize Sensing Mutual Information (SMI) as a cross-domain metric to optimize shared precoders.
The mathematical boundary of this conflict is formalized by Xinyang Li et al. in An Achievable Rate-Distortion Region for Joint State and Message Communication over Multiple Access Channels. By defining an achievable "rate-distortion" region, the authors mathematically demonstrate that any attempt to push data transmission rates closer to the theoretical maximum inherently forces a corresponding degradation in the accuracy of environmental state reconstruction.
Communication Performance (Data Rate)
^
| +-------------------+
| / \ <- Pareto Frontier
| / \
|/ \
+-----------------------------> Sensing Performance (1 / CRB)
The Geometric Conflict: Spatial Diversity vs. Angular Precision
Beyond mathematics, ISAC systems face severe physical constraints regarding how antenna arrays are designed and where hardware is deployed.
In Generative AI for Integrated Sensing and Communication: Insights From the Physical Layer Perspective, Jiacheng Wang et al. highlight a stark physical layer contradiction in antenna spacing:
- Communication tasks favor widely spaced antenna elements. Large spacing provides high spatial diversity, which helps decorrelate channels and maximize multi-user data multiplexing.
- Sensing tasks require tightly spaced antenna elements (typically half a wavelength or less). If the spacing is too wide, the radar algorithms suffer from spatial aliasing, creating "grating lobes" that introduce severe ambiguity into Direction of Arrival (DoA) estimations.
This spatial dilemma extends to the macro-deployment of the network itself. In Access Point Deployment for Localizing Accuracy and User Rate in Cell-Free Systems, Yongming Huang et al. demonstrate that communication throughput is maximized when access points are placed as close to users as possible to minimize path loss. Radar localization, however, requires access points to be arranged in specific angular configurations around a target area to maintain geometric precision (low Geometric Dilution of Precision). Moving an access point closer to a user to boost their data rate often destroys the angular geometry needed to track a target.
Beampattern and Power Allocation Dilemmas
When a base station shapes its wireless signal via beamforming, it must choose how to distribute its spatial degrees of freedom and power.
Communication beamforming aims to distribute energy toward multiple scattered users or nullify interference to maximize the Signal-to-Interference-plus-Noise Ratio (SINR). Radar beamforming requires concentrating energy into narrow, high-power beams directed at specific target directions, or sweeping a broad area uniformly to discover unknown objects.
Prioritizing one shape over the other carries heavy penalties. In Revealing the Impact of Beamforming in ISAC, Chongjun Ouyang et al. reveal that forcing a dual-functional waveform to prioritize sensing performance drops the communication diversity order down to one. This means that while a sensing-centric beam provides excellent environmental awareness, it strips the communication link of its resilience, leaving it highly vulnerable to signal fading.
To navigate this trade-off, researchers use joint optimization frameworks. As demonstrated by Xiaowei Qian et al. in Sensing-based Beamforming Design for Joint Performance Enhancement of RIS-Aided ISAC Systems and Yue Xiu et al. in CRB Minimization for RIS-Aided mmWave Integrated Sensing and Communications, systems must dynamically adjust a tunable trade-off factor. These frameworks minimize the sensing CRB while treating the users' minimum data rate requirements as hard mathematical constraints, mapping out a strict Pareto frontier between the two functions.
The Physics of Propagation: One-Way vs. Two-Way Channels
The fundamental trade-off is further exacerbated by how wireless signals decay over distance.
Communication (One-Way): [Base Station] ------------R------------> [User] (Power ~ 1/R²)
Sensing (Two-Way Roundtrip): [Base Station] ----R----> [Target] ----R----> [Receiver] (Power ~ 1/R⁴)
Communication is a one-way channel where signal power attenuates proportional to the square of the distance ($1/R^2$). Radar sensing is a two-way round-trip channel where the signal must travel to the target, reflect off its radar cross-section, and travel back to the receiver, causing power to drop off at a rate of $1/R^4$.
In Robust beamforming design for UAV communications based on integrated sensing and communication, Haolin Zhang et al. explain that this scaling disparity creates a severe resource drain. As a target moves further away, maintaining a constant level of tracking accuracy requires a disproportionate, exponential allocation of beamforming power and time-frequency resources toward the sensing task. This rapidly starves the communication functionality, causing user data rates to plummet.
Hardware and Metasurface Partitioning
The zero-sum nature of ISAC is also prominent in hardware architectures that utilize Reconfigurable Intelligent Surfaces (RIS) or Simultaneously Transmitting and Reflecting Surfaces (STARS). These surfaces contain a fixed number of meta-elements that alter or reflect wireless waves.
When deploying these surfaces for ISAC, engineers face a hardware allocation dilemma. In STARS-ISAC: How Many Sensors Do We Need?, Yuanwei Liu et al. show that a surface's elements must be split between passive reflection units (which steer data signals toward communication users) and active sensing units (which receive and process target echoes).
Converting a passive element into an active sensor improves the echo sampling resolution, lowering the sensing error. However, it simultaneously reduces the surface's spatial degrees of freedom for passive beamforming, degrading the communication link's quality of service. The authors prove that the maximum density of active sensors on a shared surface is strictly bounded by the network's minimum communication throughput requirements, illustrating that hardware partitioning is a direct physical manifestation of the sensing-communication trade-off.