Domain Selection Function

Description

The Domain Selection Function (DSF) is a network function defined in 3GPP specifications, particularly in TS 23.206 and TS 24.206, which cover voice call continuity and related protocols. Its primary role is to decide which domain—circuit-switched (CS) or packet-switched (PS)—should be used for establishing a communication session, such as a voice call. This decision is critical in networks that support both CS and PS domains, like 2G/3G/4G converged networks, where services like Voice over LTE (VoLTE) might fall back to CS for coverage reasons. The DSF operates by evaluating various inputs, including user equipment capabilities, network conditions, operator policies, and service requirements, to make an optimal domain selection that ensures service continuity and quality.

Architecturally, the DSF is typically implemented within the core network, often as part of the IMS (IP Multimedia Subsystem) or integrated with mobility management entities. In a VoLTE scenario, for example, when a user initiates a call, the DSF assesses whether the UE and network support PS voice (e.g., via IMS) or if a CS fallback is necessary due to limited LTE coverage. Key components involved include policy servers that store domain selection rules, the MME for mobility context, and the MSC (Mobile Switching Center) for CS domain coordination. The DSF interacts with signaling protocols like SIP or MAP (Mobile Application Part) to communicate its decision to other network elements, ensuring that the session is routed appropriately. Its role extends beyond voice to include other services where domain selection impacts performance, such as video calling or messaging.

In practice, the DSF works through a sequence of checks and balances. Upon session initiation, it retrieves subscriber profiles from the HSS (Home Subscriber Server) to determine preferences—for instance, a subscriber might be configured to prefer PS voice for cost savings. It also considers real-time factors like network load, available bandwidth, and UE location. If PS resources are insufficient, the DSF may trigger a handover to CS, coordinating with elements like the eNodeB and MSC to ensure a seamless transition. This functionality is essential for maintaining call reliability, especially in areas with patchy LTE coverage. By intelligently selecting domains, the DSF helps optimize network resource usage and enhances user experience by minimizing dropped calls or poor audio quality.

Purpose & Motivation

DSF was introduced to address the challenge of domain selection in hybrid networks that support both CS and PS services. As networks evolved from 2G/3G (primarily CS for voice) to 4G (PS-based for all services, including voice via IMS), there was a transitional period where both domains coexisted. Without a standardized selection mechanism, devices and networks might make suboptimal choices, leading to issues like unnecessary CS fallbacks, increased signaling overhead, or service interruptions. DSF, introduced in Release 7, provided a centralized function to make informed domain decisions, solving problems related to service continuity and efficient resource utilization.

Historically, the motivation came from the rollout of IMS and VoLTE, which promised high-quality voice over IP but required fallback to CS in areas without LTE coverage. Early implementations relied on device-centric decisions, which could be inconsistent and cause interoperability problems. DSF standardized the process, allowing operators to enforce policies based on network conditions and subscriber data. This addressed limitations like erratic fallback behavior that could degrade user experience or waste network capacity. By centralizing domain selection, DSF enabled smoother transitions between domains, supporting features like SRVCC (Single Radio Voice Call Continuity) for handovers from LTE to 2G/3G CS.

Moreover, DSF supports operational flexibility and service innovation. Operators can use it to steer traffic based on business rules, such as preferring PS domain to offload CS infrastructure or prioritizing CS for emergency calls. It also facilitates the introduction of new services that might require specific domain characteristics, like low-latency PS for real-time applications. By ensuring that domain selection is policy-driven and network-controlled, DSF enhances reliability and paves the way for full IP-based networks in the 5G era, where CS domains are phased out but legacy support remains crucial.

Key Features

• Selects between circuit-switched and packet-switched domains for sessions
• Integrates with IMS and core network elements for policy enforcement
• Supports service continuity features like SRVCC
• Evaluates UE capabilities, network conditions, and subscriber profiles
• Facilitates seamless handovers between CS and PS domains
• Enables operator-defined policies for domain preference and fallback

Evolution Across Releases

Defining Specifications