Service Delay

Description

Service Delay is a quality of service (QoS) metric defined in 3GPP TS 21.905 and TS 22.060. It measures the time taken from when a service request is invoked (e.g., by an application or user) to when the corresponding service request indication is received at the Service Receiver, signaling the arrival of application data. This metric is crucial for real-time and interactive services, such as voice over IP (VoIP), video telephony, online gaming, and mission-critical communications. Service Delay encompasses various delay components, including processing delay at the sender, transmission delay over the radio and core network, queuing delay, and processing delay at the receiver. It is closely related to end-to-end delay and is often used to specify QoS requirements in service level agreements (SLAs). In 3GPP specifications, Service Delay is used to define performance targets for different service classes, such as conversational, streaming, interactive, and background. For example, TS 22.060 (General Packet Radio Service) specifies delay requirements for GPRS services. The metric helps network operators and service providers to design, dimension, and optimize their networks to meet user expectations. It is also used in QoS negotiation and admission control to ensure that sufficient resources are available to meet the delay budget. With the evolution of mobile networks from 2G to 5G, Service Delay requirements have become more stringent, especially for ultra-reliable low-latency communication (URLLC) in 5G, where the target is 1 ms or less for user plane latency. The definition of Service Delay has remained stable across releases, indicating its fundamental nature in characterizing service performance.

Purpose & Motivation

Service Delay was introduced to provide a standardized metric for quantifying the responsiveness of packet-switched services in mobile networks. In the early days of GPRS (2.5G), there was a need to define performance parameters that could be used to specify and measure the quality of service experienced by users. Service Delay captures the total time from when a user or application initiates a service request to when the first data packet of the response arrives at the receiver. This is critical for services like web browsing, email, and later VoIP, where excessive delay degrades user experience. The metric allows operators to set QoS targets and to verify that the network meets them. It also enables application developers to design services that are tolerant to the expected delay. Before its definition, there was no common understanding of how to measure service responsiveness, leading to inconsistent performance assessments. Service Delay provides a clear, measurable quantity that can be used in SLAs and network planning. Over time, as networks evolved to support real-time services, the importance of low Service Delay grew, leading to the development of techniques like edge computing and 5G URLLC to minimize it. The definition has remained unchanged, underscoring its enduring relevance.