Service Data Unit Misdelivery Probability

Description

SDU misdelivery probability is a Quality of Service (QoS) metric defined in 3GPP TS 21.905 and referenced in TS 22.060 for GPRS. It quantifies the proportion of Service Data Units that are delivered to an incorrect destination — that is, to a user or entity other than the intended recipient — relative to the total number of SDUs transferred between a specified source and destination user during a measurement period. The definition is derived from ITU-T Recommendation X.140.

Misdelivery is a particularly severe form of data delivery failure because it represents not just a performance degradation but a correctness and potentially a security violation. When an SDU is misdelivered, the intended recipient does not receive the data (which may be counted as a loss from their perspective), and an unintended party receives data that was not meant for them. This can occur due to routing errors, addressing failures, context confusion in mobility management, or software defects in protocol implementations. In cellular networks, misdelivery could theoretically occur if a user's bearer context is incorrectly associated with another user's radio resources, or if handover procedures fail to properly transfer buffered data.

The metric is calculated as the number of misdelivered SDUs divided by the total number of SDUs transferred between the specified source and destination pair. The phrase 'between a specified source and destination user' is significant: it scopes the measurement to a particular communication relationship, making the metric meaningful for evaluating the reliability of a specific connection rather than the network as a whole. This scoping also implies that misdelivery is measured from the perspective of the intended source-destination pair, counting SDUs that were sent by the source but arrived at some other destination instead.

In practice, SDU misdelivery probability is expected to be extremely low in well-designed networks — ideally zero. Unlike error and loss probabilities, which are inherent to wireless transmission and are managed through error correction and retransmission, misdelivery should not occur in a correctly functioning network. Its presence indicates a fault condition: a bug in protocol implementation, a failure in connection management, a security breach, or a misconfiguration. Therefore, the metric serves primarily as a correctness and integrity check rather than as a performance target to be optimized through radio resource management.

3GPP specifications typically do not specify target values for SDU misdelivery probability because the expected value is effectively zero. Instead, the metric is defined to provide a vocabulary for describing network behavior and for use in conformance testing and fault analysis. When misdelivery does occur, it is treated as a defect to be corrected rather than a QoS parameter to be tuned. The term has remained unchanged across all 3GPP releases from Rel-4 through Rel-19, reflecting its role as a foundational correctness metric rather than a performance parameter subject to technological evolution.

Purpose & Motivation

SDU misdelivery probability exists to provide a formal, standardized way to describe and measure the most severe class of data delivery failure: delivering data to the wrong recipient. While error and loss probabilities address performance degradation within an acceptable range, misdelivery represents a fundamental correctness failure that can have security, privacy, and functional implications. The metric was inherited from ITU-T X.140, which recognized that a complete framework for data communication performance must include not only how much data is corrupted or lost but also whether data reaches the correct destination.

The metric solves the problem of having a precise vocabulary for discussing and testing delivery correctness. Without such a definition, engineers and operators would lack a common language for specifying requirements, writing test cases, or reporting faults related to misdelivery. By defining it as a ratio over a specified source-destination pair and sample, the metric enables quantitative assessment: even though the expected value is zero, the definition allows measurement and verification that the actual value meets this expectation.

Historically, the inclusion of misdelivery probability in the ITU-T X.140 framework and its adoption by 3GPP reflects the maturation of packet data networking from a research curiosity to a commercial service requiring rigorous quality specifications. Early data networks might have tolerated occasional misdelivery as a rare anomaly, but as networks became critical infrastructure carrying sensitive data, the need for formal correctness metrics grew. In 3GPP, the metric serves as a reminder that QoS encompasses not only performance dimensions like throughput and delay but also correctness dimensions like delivery accuracy. It remains relevant in 5G and beyond because the consequences of misdelivery — whether privacy breaches or functional failures — are more severe than ever in a connected world.