Printable character Error Rate

Description

Printable character Error Rate (PER) is a key performance indicator (KPI) defined in 3GPP specifications to quantify the reliability of character-oriented telecommunication services. It measures the accuracy of information transfer for services where the payload consists of printable characters from defined alphabets (e.g., GSM 7-bit default alphabet, UCS2). PER is defined as the ratio between the number of printable characters that are delivered erroneously to the destination and the total number of printable characters sent by the source, over a given observation period. An erroneous delivery includes characters that are substituted, deleted, or inserted compared to the original message.

The architecture for measuring PER involves the end-to-end service layer. The source application (e.g., an SMS Center) and the destination application (e.g., a receiving UE) are the logical points of measurement. The calculation considers the entire path, including core network transport, any interworking functions, and the radio access. Key components in the evaluation are the defined character sets and the rules for character mapping and counting. For instance, in SMS, a message containing national language shift tables or locking shift characters requires careful interpretation to count printable characters correctly. The measurement is typically conducted in a test environment or via network probes that can compare sent and received payloads.

How PER works as a metric is tied to service testing and benchmarking. Test equipment generates messages with known sequences of printable characters and sends them through the network under test. The receiving equipment compares the delivered characters to the original. Discrepancies are logged, and the PER is calculated. This process tests the robustness of the underlying transport protocols (e.g., MAP for SMS), encoding/decoding algorithms, and any transcoding that may occur between different network domains. A low PER indicates high fidelity in the text service, which is paramount for services where data integrity is critical, such as authentication codes (e.g., one-time passwords), financial alerts, or emergency notifications. PER is often measured under various network conditions, including marginal radio coverage, to ensure service robustness.

Purpose & Motivation

The Printable character Error Rate metric exists to provide a standardized, quantifiable measure of quality for non-voice, character-based services. As mobile networks evolved beyond voice to offer messaging services like SMS, it became necessary to move beyond simple success/failure metrics (like delivery reports) to understand the quality of the delivered content. A message could be delivered but corrupted, which is unacceptable for many applications.

PER solves the problem of undefined service quality for text. Prior to its standardization, operators and vendors lacked a common way to benchmark and compare the performance of messaging platforms. It was motivated by the need to ensure user trust in messaging services, especially as they began to be used for critical communications. A high PER would mean users receive garbled messages, leading to confusion, failed transactions, and poor customer satisfaction.

Its creation addresses the limitations of simpler metrics like Frame Error Rate (FER) or Bit Error Rate (BER), which measure lower-layer performance but do not directly translate to user-perceived quality at the application layer. PER is service-specific, focusing on the information meaningful to the end-user: the characters. It is particularly important for the evolution of messaging towards rich communication services (RCS), machine-to-machine (M2M) communication where commands must be exact, and interoperability between different networks and character sets. By defining PER, 3GPP provided a tool for network engineers to objectively assess, troubleshoot, and improve the reliability of a fundamental class of mobile services.

Key Features

• Measures accuracy at the service/application layer for character-based services
• Defined for various character sets including GSM 7-bit default and UCS2
• Considers character substitution, deletion, and insertion as errors
• Used for end-to-end performance testing and benchmarking
• Critical KPI for SMS, USSD, and instant messaging services
• Supports testing under controlled and live network conditions

Evolution Across Releases

Defining Specifications