Citizens Broadband Radio Service (CBRS)

• CBRS is a spectrum sharing framework that offers dynamic, three-tiered access to 150 MHz mid-band frequencies while protecting incumbent systems.
• It enables high indoor spectral reuse with neutral-host deployments achieving up to 535× throughput gains and significant TX power reductions.
• Leveraging a centralized SAS, advanced ESC, and AI-driven detection, CBRS ensures real-time interference management and secure spectrum allocation.

Citizens Broadband Radio Service (CBRS) is a paradigm-shifting spectrum sharing framework introduced in the United States, enabling dynamic tiered access to 150 MHz of mid-band spectrum (3550–3700 MHz) for commercial wireless, private LTE/5G, and critical IoT applications while guaranteeing protection for federal incumbents. CBRS leverages a centralized Spectrum Access System (SAS), sophisticated Environmental Sensing Capability (ESC) networks, and a multi-participant ecosystem involving incumbent users, Priority Access Licensees (PAL), and General Authorized Access (GAA) operators. The platform is distinguished by its regulatory structure, high spectral efficiency through indoor neutral-host models, advanced dynamic management, and empirical performance data verifying robust indoor and outdoor coexistence.

1. Tiered Spectrum Access and Regulatory Architecture

CBRS spectrum access is structured by a three-tiered model:

• Incumbent Access (Tier 1): U.S. Navy shipborne radars and federal users, with absolute, preemptive access. Incumbents are protected through exclusion zones, guaranteed interference thresholds (e.g., $INR_{thr} = -6$ dB at radar receivers (Krishnan et al., 2017)), and real-time ESC-based detection mechanisms.
• Priority Access Licenses (Tier 2): Short-term, geography-specific licenses (10 MHz channels) granted via auction. PALs are protected from interference by Tier 3 but preempted by incumbents. Channel grants and revocations are dynamically managed by SAS. Optimal split between licensed (L) and unlicensed (W−L) spectrum, with a moderate L/W ≈ 0.3–0.5, maximizes social welfare (Ghosh et al., 2019).
• General Authorized Access (Tier 3): Unlicensed users opportunistically access any available spectrum not in use by higher tiers. All Tier 3 operations are coordinated via SAS, with no regulatory protection from co-tier interference.

The SAS maintains a real-time, geolocated database on CBSD (Citizens Broadband Radio Service Device) registrations, assigns grants, and manages channel and power allocations (Rochman et al., 23 May 2025). Integration with ESC networks ensures protection for incumbents by vacating or modifying grants in response to radar detection. Emerging privacy-preserving SAS solutions, e.g., TrustSAS, combine multi-server private information retrieval (BatchPIR), EPID anonymity, and BLS threshold signatures within a blockchain/BFT consensus framework, offering compliance and privacy guarantees at scale (Grissa et al., 2019).

2. Neutral-Host Indoor Deployments and Performance Metrics

CBRS enables high-efficiency indoor neutral-host (NH) network models. In this architecture:

• Low-power CBSDs (≤30 dBm/10 MHz EIRP) are ceiling-mounted, forming dense small-cell topologies.
• NH operators utilize Multi-Operator Core Network (MOCN) interfaces, supporting subscribers from multiple MNOs through PLMN-ID broadcasting and secure SIM-based tunnel establishment (Rochman et al., 23 May 2025).
• Empirical data from healthcare and retail store environments show a median building penetration loss exceeding 22 dB (typ. 26.6 dB), sharply attenuating outdoor leakage and preventing harmful interference to incumbents (Rochman et al., 23 May 2025, Palathinkal et al., 5 Jun 2025). This physical containment allows high indoor spectral re-use and robust coexistence.

Performance improvements in measured deployments include:

• Downlink median throughput: $$R_\mathrm{DL} = T_{\mathrm{neutral},DL}/T_{\mathrm{worst\,MNO},DL} = 535\times$$ (e.g., $\sim$45 Mbps CBRS-NH vs. 0.084 Mbps macro MNO-A).
• Uplink median improvement: $R_\mathrm{UL} = 33\times$.
• Median uplink TX power reduction: $\Delta P \approx 12$ dB.
• Spectrum efficiency: Six CBSDs cover $17,400\,\mathrm{m}^2$ area, compared to $65$ Wi-Fi APs, while achieving up to $2.08\times$ higher user-layer throughput than 5G macro with 5.6-fold less spectrum (Rochman et al., 23 May 2025, Palathinkal et al., 5 Jun 2025).

Capacity offload is quantifiable; median freed resource blocks in macro MNO slots $\Delta RB \approx 233$, directly improving macro user throughput outdoors.

3. Interference Management and Incumbent Protection

CBRS interference management rests on both architectural and operational controls:

• Protection distances: Monte Carlo analysis confirms that, for co-channel secondary operation, a $30$ km exclusion zone yields $P(INR \leq -6\ \mathrm{dB}) \geq 0.9$ for radar receivers; adjacent-channel deployment allows as little as $1$ km separation (Krishnan et al., 2017).
• Power control: Centralized or sectorized power control algorithms minimize aggregate EIRP, reducing exclusion zones by adapting per-device limits (down to $20$ dBm where required).
• Dynamic frequency assignment: SAS leverages real-time ESC sensing and CBSD self-reporting to rapidly exclude protected zones and enforce channel/power reassignment (Rochman et al., 23 May 2025).
• Privacy/security: Advanced cryptosystems and permissioned BFT blockchains report and log coexistence events, location data, and usage notifications in compliance with FCC and user privacy constraints (Grissa et al., 2019).

For GAA devices, SAS limits mutual awareness and does not perform co-tier pairwise interference checks, requiring local dynamic channel selection for optimal performance (Tusha et al., 2024).

4. Secondary and Adjacent Channel Interference

Secondary interference in CBRS arises from both co-channel (CCI) and adjacent-channel (ACI) mechanisms:

• Co-channel interference (CCI): Occurs when multiple GAA CBSDs utilize the same frequency in close proximity, leading to measured throughput collapses (e.g., median drops from $25$ Mbps to $15$ Mbps in overlapping regions) (Tusha et al., 2024).
• Adjacent-channel interference (ACI): High-power 5G C-band adjacent deployments (3700–3980 MHz) with no guard band lead to significant leakage and unsynchronized TDD schedule conflicts. Overlapping 4G CBRS and 5G C-band uplink/downlink slots can cause up to $60\%$ and $43\%$ throughput loss, respectively (Rochman et al., 2023). Implementing a mere $20$ MHz guard band reduces worst-case CBRS downlink loss from $60\%\to21\%$.
• TDD desynchronization: Uncoordinated frame structures in adjacent bands exacerbate cross-interference.

To mitigate these effects, recommended strategies include:

• Static/dynamic guard bands.
• Edge PRB power back-off.
• TDD synchronization.
• Application-aware network slicing for latency protection (Rochman et al., 2023).

Table 1: Throughput Reduction Metrics under ACI

| Scenario | CBRS DL Loss | C-band DL Loss |

|:-----------|:------------:|:--------------:|

| No guard | 60% | 43% |

| 20 MHz gap | 21% | 30% |

5. Spectrum Sensing, ESC, and AI-Driven Detection

Environmental Sensing Capability (ESC) is central to FCC compliance, mandating ≥99% detection probability for naval radar at SINR ≥ 20 dB. Recent ML-based systems dramatically surpass this threshold:

• ViT-based spectrogram classifiers achieve 99% radar detection accuracy down to SINR = $-5$ dB, extending prior art by 25 dB (Khan et al., 11 Oct 2025). Accurate LFM waveform classification (6 types) at 93% is achieved even under high 5G interference.
• RadYOLOLet combines spectrogram CNN and Wavelet-CNN to push detection accuracy to 100% for ≥16 dB SINR/SNR, with sub-1% false alarm, fast inference (<16 ms per decision window), and robust parameter extraction (Sarkar et al., 2023).
• Decentralized alternatives leveraging dApps and O-RAN integration enable sub-frame spectrum sensing and real-time gNB adaptation, potentially eliminating external SAS/ESC dependencies (Gangula et al., 2024). This approach achieves $P_{\rm D} \geq 0.95$ and $P_{\rm FA} \leq 0.02$ at SNR ≥ $-5$ dB, enabling agile protection.

6. Distributed Sharing, Private CBRS, and Application Scenarios

CBRS supports:

• Distributed Spectrum Sharing: Advanced MIMA physical layers enable simultaneous multichannel monitoring, slot-level “fast rendezvous” ($T_r\approx$ 1 LTE slot), and high efficiency ($80$–$90\%$ of upper bound at high SNR) for opportunistic SUs (Cai et al., 2017).
• Private and IoT-Focused Deployment: Private CBRS networks utilize domain proxies for real-time channel grant/status mediation with SAS. The Maximum Transmission Continuity (MTC) scheduler dynamically assigns channels to maximize data continuity for prioritized IoT flows, with simulation-based improvements in packet delivery (96% PDR), median latency (22 ms), and continuity (98%) over baseline schemes (Kuo et al., 2023).
• Heterogeneous Application: CBRS, in conjunction with Wi-Fi and macro LTE/5G, supports smart grid advanced metering infrastructure (AMI) via a duty-cycled LTE-U/WiFi overlay. Time-division duty cycles (e.g., $\alpha=0.6$) reliably partition resources without LBT, maintaining both LTE and Wi-Fi throughput ($\sim$36 Mbps each in 20 MHz) (Parvez et al., 2017).

7. Empirical Measurements and Best Practices

Ground and aerial measurement campaigns reveal:

• Urban deployments exhibit strong CBRS signal power increases (up to 10 dB) with altitude due to line-of-sight expansion. Occupancy is highest (up to 80%) in the upper 50 MHz (3650–3700 MHz), indicating heavy PAL/GAA contention (Raouf et al., 2023).
• Dynamic channel (re-)allocation, SAS-driven frequency/power assignments, and detailed site surveys for building loss optimization are essential for optimal indoor deployment (Rochman et al., 23 May 2025, Palathinkal et al., 5 Jun 2025).
• Smart reconfigurable surfaces (e.g., WaveFlex) can provide 8.5 dB average SNR gain and multi-Mbps throughput improvements in private indoor cellular networks, adapting to fast-changing channel/frequency allocations (Yi et al., 2023).

---

In sum, Citizens Broadband Radio Service (CBRS) operationalizes a robust, scalable, and efficient spectrum sharing ecosystem. Its field-validated architectures and control frameworks reconcile regulatory incumbency, dense user demand, and robust coexistence, setting precedents for mid-band sharing worldwide. Ongoing research underscores the importance of rigorous interference management, ML-driven spectrum awareness, privacy-preserving architecture, and flexible application-layer adaptation for next-generation wireless systems (Rochman et al., 23 May 2025, Krishnan et al., 2017, Tusha et al., 2024, Kuo et al., 2023, Palathinkal et al., 5 Jun 2025, Khan et al., 11 Oct 2025, Sarkar et al., 2023, Grissa et al., 2019, Gangula et al., 2024).