PERFORMANCE-RELATED DATA IN UTRAN
1.1.4 PERFORMANCE MEASUREMENT DEFINITIONS OF 3GPP
Anyone searching for KPI definitions in 3GPP specifications will be disappointed. 3GPP has not defined any key performance indicators. However, there are definitions of performance- related data that is either transmitted between network elements using measurement report messages or data measured in network elements and provided for analysis in higher level network management systems. Basically those definitions cover two major groups of performance-related data: radio-related measurements and protocol event counters.
Also only a fairly general framework document exists that describes performance measurement tasks from a high level perspective. This document is 3GPP 32.401 ‘Concepts and Requirements for Performance Measurement’. It defines all tasks for measurement job administration, e.g. the start and stop of measurement procedures and how measurement jobs are initialised, modified and terminated. The second part deals with measurement type definitions and measurement result generation. It contains e.g. descriptions of how gauges are measured (e.g. max/min/mean values of response time measurements and when cumulative counters need to be reset). In addition, there are some interesting statements about granularity and accuracy, e.g. a success event will always belong to the same sampling
Figure 1.7 The soft handover advantage
period as the attempt event to which it is bound. All in all, this standard document gives a good overview on how to store and transport performance data that is already available, but is says nothing about how to get such data or about how to compute and analyse it.
Measurement definition can be found in 3GPP 32.403 ‘Performance Measurement – UMTS and Combined UMTS/GSM’. The title sounds like a perfect compendium, but actually it is just a list of protocol event counter definitions. It is highly recommended to look for the latest available document version, because in some older versions (e.g. v4.5.0) even the protocol events are not always well defined, as the following example shows. It defines for instance a single counter that counts NBAP and RNSAP radio link addition events, but in the two protocols – despite the procedures having the same name – they have totally different functions. Newer versions like v5.8.0 have corrected this error and define separate counters for NBAP and RNSAP radio link addition.
All in all the main limitations of 3GPP 32.403 remain the same, no matter which specific document version is used:
Only counters for higher layer protocol layers are defined – no events are defined that indicate that problems might appear on transport layers, for instance if AAL2 SVCs cannot be set up or are dropped.
From an RNC perspective only the RRC, NBAP and RNSAP dealing with connection set up and handover are taken into account. There is no definition of ALCAP events, neither are counters for other interesting events defined (e.g. NBAP radio link failure indication).
There is no multi-interface perspective, e.g. it is not possible to distinguish between NBAP radio link setup for intra-RNC soft handover and inter-RNC soft handover (triggered by the RNSAP radio link setup or radio link addition procedures).
There is no multi-layer perspective, e.g. it is not possible to have combined soft handover analysis based on NBAP, ALCAP and RRC events that interact to perform this mobility management procedure.
And finally there is no relation between defined protocol events and network services as they can be found in service level agreements between network operator and subscriber.
Also there is another problem with precise counter definitions. Usually protocol events are defined on a very general (which means not specific) level. It is one thing to mention that hard handover procedures can be executed using RRC physical channel reconfiguration, RRC transport channel reconfiguration, RRC radio bearer reconfiguration, RRC radio bearer establishment or RRC radio bearer deletion, however, there is no information about which message is used in which case and which information element would distinguish between intra-frequency and inter-frequency hard handover (as shown in Figure 1.8). Such specific information is also missing in 3GPP 25.331Radio Resource Control.
Some of those limits might be covered by definitions found in 3GPP 52.402Telecommu- nication Management; Performance Management (PM); Performance Measurements – GSM. This standard document is well structured and covers more aspects than 32.403, but it does not contain any UTRAN-relevant definitions. It could be seen as a combination of contents of 32.401 and 32.403 with emphasis on GERAN. Obviously it would be correct to call 3GPP 52.402 a sister document of 3GPP 32.403.
A much better situation is found when looking for definitions of radio-related measure- ments. Here the specifications are written in a detailed and complex way although it does
take a lot of knowledge to understand what is measured, how measurement results are computed, transmitted and finally presented.
It starts with 3GPP 25.215 Physical Layer – Measurements (FDD). This standard document gives a comprehensive overview of measurement abilities on UE and UTRAN.
It is very useful to understand which parameters are measured in and reported by UE (using RRC measurement reports) and cell/Node B (using NBAP measurement reports). For each measurement parameter a clear definition is given including exceptions and limitations.
3GPP 25.216 is a comparable document of 3GPP 25.215 for UTRAN TDD mode but details of radio quality measurement in TDD mode are beyond the scope of this book.
The most important radio measurement abilities defined in 3GPP 25.215 are:
From Node B (NBAP measurement reports):
Common measurements – related to a cell, not to a single connection:
– Received total wideband power(RTWP)ẳtotal UTRA uplink frequency noise on cell antenna
– Transmitted carrier powerẳtotal downlink Tx power of cell antenna
– Preambles of PRACHẳnumber of connection request attempts on radio interface necessary until network reacts by sending either positive or negative confirmation. In the case of positive confirmation UE is allowed to send an RRC Connection Request as the first higher layer message.
Dedicated measurements – related to a single connection or single radio link, but not related to a single cell:
– SIR (signal-to-interference ratio)ẳratio between measured uplink RSCP of a single UE’s signal and the interference code power (ISCP) after spreading of received signals
Figure 1.8 Same protocol message used for different handovers procedures
– SIR error– measured difference between uplink SIR target set by SRNC and measured SIR
– Transmitted code powerẳtransmitted power for one downlink dedicated physical channel (DPCH) sent by a single cell/antenna
– Round trip timeon Uu interfaceẳestimation based on time difference between RLC frames and appropriate RLC Acknowledgement (RLC AM)
From UE (RRC measurement reports). All those measurements are related to a single radio link (ẳsingle cell) used for a single connection, because they all depend on the location of the UE related to a single cell at the time of measurement:
Chip energy over noise (Ec/N0)ẳdownlink equivalent of SIR, but based on measure- ment of common pilot channel (CPICH)
Received signal code power (RSCP)– Rx level of downlink dedicated physical channel (DPCH) on UE antenna
UTRA received signal strength indicator (RSSI) – total UTRA downlink frequency noise on UE antenna, downlink equivalent to RTWP.
Event-IDs– used to report predefined measurement events measured downlink frequency.
There are other measurement abilities defined in the same 3GPP standard document, but they have not been implemented yet in UTRAN, which can be understood as a sign of lower priority.
All the above-mentioned radio-related measurements will be explained in more detail in Section 2.2.
While 3GPP 25.215 defines the measurement parameters there is another interesting specification named 3GPP 25.133Requirements for Support of Radio Resource Management (FDD). It explains how the measurement results of 3GPP 25.215 parameters are reported, which reporting ranges are defined and how these measurement results are encoded in signalling messages (measurement reports). There are also some formula definitions that describe how radio-specific measurement values are computed in UE or Node B, but these formulas are not KPI definitions. Additionally, measurement requirements for all UTRAN mobility management procedures such as cell (re-)selection and handovers and RRC connection control are described including timing and signalling characteristics related to these procedures.
These few standard documents contain what is defined for UTRAN performance measurement in 3GPP standards and it is always a good idea to remember these standard documents when talking about KPI definitions.