PDU Set Integrated Handling Information

Description

PDU Set Integrated Handling Information (PSIHI) is a protocol feature introduced in 3GPP Release 18 as part of 5G-Advanced enhancements, designed to improve the handling of data units in the user plane. It refers to control information that is integrated with a set of Protocol Data Units (PDUs) to provide instructions on how these PDUs should be processed, transmitted, or managed collectively by the network and User Equipment (UE). PSIHI is carried within the data flow, typically in packet headers or associated metadata, and conveys semantics related to the application data, such as dependencies between PDUs, processing deadlines, or aggregation rules. This allows the 5G system to apply optimized handling strategies—like scheduling, duplication, or discard—based on the application's needs, rather than treating each PDU independently.

Architecturally, PSIHI operates within the Packet Data Convergence Protocol (PDCP) layer and potentially interacts with higher layers like the Service Data Adaptation Protocol (SDAP) and application layer in the 5G protocol stack. It is generated by the application or a network function (e.g., an edge server) and embedded into the PDU set before transmission over the air interface. The gNodeB (gNB) and UE use this information to make intelligent decisions in the radio access network (RAN) and core network. For instance, PSIHI might indicate that a group of PDUs belongs to a single video frame, allowing the RAN to prioritize or discard them as a unit to maintain quality of service. Key components include the PSIHI field itself, which contains flags or parameters defining the handling rules, and the associated protocols that parse and act on this information, as specified in technical specifications like TS 38.300 and TS 38.835.

In practice, PSIHI enhances user plane efficiency by enabling context-aware data handling. When a PDU set with PSIHI arrives at the gNB, the scheduler can consider the integrated information to optimize resource allocation—for example, by ensuring that all PDUs in a set are transmitted contiguously to reduce latency for time-sensitive applications. On the UE side, the receiver uses PSIHI to reassemble or process PDUs correctly, potentially reducing buffer requirements and improving application performance. This mechanism is particularly valuable for advanced use cases like extended reality (XR), industrial IoT, and ultra-reliable low-latency communication (URLLC), where data semantics are critical for meeting stringent requirements. By bridging the gap between application intent and network behavior, PSIHI supports more dynamic and efficient 5G systems.

Purpose & Motivation

PSIHI was created to address the limitations of traditional PDU handling in 5G networks, where each data unit is processed independently without awareness of application-level context. Previous approaches, while efficient for generic data traffic, struggled to meet the complex requirements of emerging applications like XR, autonomous systems, and tactile internet, which involve structured data sets with interdependencies and strict timing constraints. Without integrated handling information, the network might inadvertently discard or delay critical PDUs, degrading user experience and reliability.

The primary problem PSIHI solves is the inefficiency in resource utilization and quality of service management for application-aware services. By providing explicit handling instructions within the data flow, it enables the RAN and core network to optimize transmission strategies based on semantic knowledge. This reduces overhead compared to out-of-band signaling and allows for real-time adaptations, such as grouping PDUs for joint scheduling or applying specific error recovery mechanisms. The motivation for PSIHI stems from the 5G-Advanced vision of supporting more intelligent and flexible networks, where user plane enhancements are key to achieving higher performance and energy efficiency.

Historically, earlier 3GPP releases relied on QoS flows and differentiated services code points (DSCP) for traffic differentiation, but these mechanisms lacked granularity for PDU-level coordination within a flow. PSIHI builds on concepts like PDCP duplication and data burst handling, extending them with integrated control to support the evolving needs of vertical industries. Its introduction in Release 18 was driven by industry demands for better support of XR and industrial applications, as documented in specs like TS 26.804 and TS 23.501, where handling of media frames and sensor data sets requires tight integration between application and network layers.