LTE and NR Dynamic Spectrum Sharing (DSS)

The fundamental mechanism of DSS is a super-fast, software-based scheduler in the base station. Unlike static spectrum refarming, which dedicates a block of spectrum permanently to one technology, DSS uses a Multiplexing Unit to interleave LTE and NR transmissions in the time domain.

• Resource Block Allocation: The scheduler assigns individual Resource Blocks (RBs) to either LTE or NR users every 1-millisecond Transmission Time Interval (TTI).
• Traffic-Driven: If more 5G devices are active, the scheduler instantly allocates more RBs to NR, and vice-versa, preventing stranded assets.

To transmit 5G NR signals without disrupting legacy LTE devices, DSS leverages Multicast-Broadcast Single-Frequency Network (MBSFN) subframes. This is a clever backward-compatibility trick.

• The Mechanism: The network configures specific LTE subframes as MBSFN, which legacy LTE user equipment (UE) ignores after reading the control region.
• NR Transmission: The base station then uses the data region of these 'empty' subframes to transmit NR synchronization signals (SSB) and system information (SIB1), making 5G visible to new devices without confusing older ones.

A core technical challenge is the LTE Cell-Specific Reference Signal (CRS), which is always-on and broadcast continuously across the entire carrier. 5G NR must work around this interference.

• CRS Rate Matching: The NR scheduler is explicitly informed of the exact time-frequency positions of the LTE CRS. It then 'punctures' or rate-matches its NR data transmissions to avoid these resource elements, preventing catastrophic interference.
• Overhead Impact: This avoidance reduces the raw capacity available for 5G NR, meaning DSS provides a smooth migration path but does not deliver the full spectral efficiency of a pure NR carrier.

DSS relies on intelligent user equipment (UE) capability reporting to steer traffic. The network doesn't just broadcast both technologies blindly; it actively sorts devices.

• Capability Exchange: When a 5G-capable phone connects, it reports its NR support. The network immediately recognizes it and can schedule it on NR resources.
• Seamless Handover: Legacy 4G-only devices remain on the LTE resources. This ensures zero service disruption for the existing subscriber base while 5G users benefit from the new radio access technology on the same carrier.

DSS functions as a zero-downtime bridge for spectrum refarming. The traditional method of shutting down a 4G carrier and re-launching it as 5G causes a hard service interruption.

• Soft Migration: Operators can activate NR on a live LTE carrier without a 'flash cut.' As the mix of 5G devices grows over time, the scheduler naturally allocates more capacity to NR.
• Traffic-Triggered Reallocation: Eventually, when LTE traffic drops below a threshold, the operator can seamlessly decommission the LTE side, and the entire carrier becomes a pure, high-efficiency NR channel without a disruptive re-launch event.

DSS requires tight coordination of the Physical Downlink Control Channel (PDCCH) for both technologies. The control signaling must be kept separate and intelligible.

• CORESET Configuration: For NR, the network configures a Control Resource Set (CORESET) that avoids the LTE control region. This ensures the 5G device knows exactly where to look for its scheduling grants.
• Search Space Design: The NR search space is carefully designed to not overlap with LTE's PDCCH, allowing both technologies to independently schedule their respective users in the same subframe without control channel collisions.

An intelligent wireless communication system that forms the theoretical basis for DSS. A cognitive radio is environment-aware, autonomously adapting its transmission parameters—such as frequency, power, and modulation—based on real-time sensing. It executes a cognitive cycle of observing, orienting, planning, deciding, and acting to optimize spectrum use and avoid interference without manual operator intervention.

The overarching real-time spectrum management strategy that DSS implements. DSA allows secondary users to opportunistically access temporarily unused licensed frequency bands without causing harmful interference to primary users. Unlike static allocation, DSA operates on a per-millisecond basis, exploiting spectrum holes in time, frequency, and space to dramatically increase spectral efficiency.

A machine learning technique that elevates DSS from reactive to proactive operation. By applying models like Long Short-Term Memory (LSTM) networks to historical spectrum usage data, the system forecasts future occupancy patterns. This allows the DSS scheduler to pre-allocate resources before congestion occurs, minimizing latency and maximizing throughput by avoiding the delays inherent in purely reactive sensing.

A U.S. regulatory framework in the 3.5 GHz band that operationalizes spectrum sharing through a three-tiered access model:

• Incumbent Access: Highest priority for federal radar and satellite users
• Priority Access Licenses (PAL): Licensed by census tract for guaranteed quality of service
• General Authorized Access (GAA): Opportunistic, unlicensed use of any remaining spectrum This framework is managed by an automated Spectrum Access System (SAS).

The automated, cloud-based spectrum coordinator mandated by the FCC for the CBRS band. The SAS dynamically assigns frequencies and enforces interference protection policies across all three tiers of users. It ingests real-time data from environmental sensing capability (ESC) sensors to detect federal incumbent radar and instantly reallocates spectrum, serving as a real-world model for centralized DSS orchestration.

An xApp or rApp hosted on the O-RAN near-real-time or non-real-time RIC that uses AI to guide DSS decisions. This microservice-based application consumes network telemetry and enforces operator-defined policies to dynamically allocate spectrum between LTE and NR. It represents the open, software-defined evolution of DSS, moving away from proprietary vendor implementations toward interoperable, intelligent control loops.