1 Introduction

1.1 Scope

This document contains a high-level description of the Release 99 Features of the 3^rd Generation Mobile System developed within the 3^rd Generation Partnership Project (3GPP). It is part of a series of documents developed by MCC to provide a complete overview of the technical content of each Release.

A Feature is defined as a new or substantially enhanced functionality which represents added value to the existing system (see 3GPP TR 21.900). A feature should normally embody an improved service to the customer and / or increased revenue generation potential to the supplier.

Features are as independent as possible from each other, and relationships between features are clarified here.

The concept and definition of “Feature” was introduced in Release 4, i.e. in the Release following Release 99. When elaborating this document, MCC has applied the “feature” concept to the work done for Release 99, so the “Release 99 Features” are introduced by this document and were not officially defined as such by 3GPP. Also, the use of the 3GPP Work Plan was introduced in Release 4, i.e. there was no official tracking of the work progress in Release 99, so this document was written from a detailed review of all specifications, change requests, meeting contributions and reports, etc.

The features have been grouped into different sections:

- UMTS features,

- features applicable to both GSM and UMTS,

- GSM features,

- charging and testing activities[1],

- GSM features transposed to UMTS and

- “Features” not bringing any additional service.

For each feature (or independent item), references are given to guide the reader on how to deepen his knowledge of the subject: the Work Item Description (WID), when available, as well as the list of impacted specifications are provided in the beginning of the section describing the feature. Only the list of impacted specifications is provided here. The exact impact on a given specification due to a given feature is described in the Change Request (CR) list which can be found at the end of the specification, or in the CR database, which provides the full list of CRs for all 3GPP specifications. All this information is available on the 3GPP web site, as described below.

The second part of this introduction contains global references, and provides links to the 3GPP Specifications, the temporary documents (tdocs), the Work Item Descriptions (WIDs) and the CR database.

The main body of this document is structured according to the 3GPP Release 99 Features: each section corresponds to one Release 99 Feature, grouped as described above.

1.1.1 Change Request database

A specification is originally drafted and maintained by a “rapporteur” (the editor), who compiles the contents from discussions in the WGs and TSGs. When it is considered to be 80% complete, it is brought under "change control". Thereafter, changes to the specification can only be made using Change Requests that are usually agreed by consensus in the Working Group responsible for the specification, and then formally approved by the relevant Technical Specification Group.

The Change Request database contains all available information on Change Requests, including a Work Item code, a Change Request number that is unique within the specification (different versions are possible, but only one version can ever be approved), the status of each Change Request and references to relevant temporary document numbers and meetings.

2.UMTS Features

2.1.Architecture of the GSM-UMTS Platform

UMTS (Universal Mobile Telecommunication System) refers to the interconnection of a new type of Access Network (AN), the UTRAN (UMTS Terrestrial Radio Access Network), to the adapted pre-Release 99 GSM/GPRS Core Network (CN) infrastructure. The UTRAN and its new bearer services are described in the next clause.

A basic requirement for Release 99 UMTS was to minimise the impacts on the Core Network when introducing the UTRAN. This principle was achieved to a great extent. The biggest impacts are the creation of a new type of interface between core and access networks, and the “upgrading” of the CN signalling to take into account the new capabilities offered by the UTRAN.

This section describes the UMTS network, using a top-down approach: the network is logically divided in a number of sets, both from the architectural aspect and from the protocols aspect. From the architectural point of view, the sets are called “domains” (a domain is a group of entities). From the protocols point of view, the sets are called “strata” (a stratum is a group of protocols). These principles, introduced for first time in UMTS, could also apply to GSM (and, indeed, to other types of network).

They do not correspond to any concrete realisation in the network but were established mainly to organise the work as to allow different groups of people to work in parallel, each one being responsible for one (or several) domain(s) and/or stratum(a).

The domains are:

- the User Equipment domain, containing the elements the end-user carries with him, composed of:

o the Mobile Equipment domain (the “phone”), containing the radio transmitting/receiving device (in the Mobile Termination, MT) and the application (in the Terminal Equipment, TE), defined by 3GPP T2 group, and

o the USIM domain, typically embedded in an IC card, defined by 3GPPP T3 group.

- the Infrastructure domain, i.e. the set of all the network entities, composed of:

o the Access Network domain, comprising all the entities closely related to the radio technology, defined by 3GPP RAN1 to RAN4 groups and

o the Core Network domain, defined by 3GPP CN1 to CN4 groups, composed of:

§ the Serving Network domain, composed of

· the Circuit Switched (CS) domain

· the Packet Switched (PS) domain

§ the Transit Network domain (potentially composed of CS and PS also), and

§ the Home Network domain, containing permanently all the user specific data and responsible for management of subscription information.

The domains are shown in the following figure:

The strata are:

· the Transport stratum, supporting the transport of user data and network control signalling from other strata through UMTS. It encompasses the Access Stratum, which is the part of the transport stratum located between the edge node of the serving core network domain and the MT;

· the Home stratum, which contains the protocols and functions related to the handling and storage of subscription data and possibly home network specific services;

· the Serving stratum, which consists of protocols and functions to route and transmit data/information, user or network generated, from source to destination; and

the Application stratum, which represents the application process itself, provided to the end-user. It includes end-to-end protocols and functions which make use of services provided by the home, serving and transport strata and infrastructure to support services and/or value added services.

The connection between domains is a network interface or a reference point, the connection between strata is a service primitive (which may use a standardised Service Access Point Identifier or may not be subject to standardisation) , which is internal to a network node.

Further definition of the domains and strata as well as their relationship is given in TS 23.101.

The next step in the network description is the division between “entities” and “protocols”: a domain is a group of (potentially just one) entities, a stratum is a group of (potentially just one) protocols.

UMTS introduces new entities in the AN ‑ all the UTRAN entities are new, as described in the corresponding section ‑ but not in the CN. The connection between entities is network interface or reference point. The following figure, extracted from TS 23.002, shows the UMTS and GSM Network Architecture.

UMTS and GSM Network Architecture

Legend:

Bold lines: interfaces supporting application traffic (also called “user data”);

Dashed/thin lines: interfaces supporting signalling.

Red lines and boxes: interfaces and entities specific to UMTS

Turning to protocols, again new protocols are introduced for the UTRAN (see corresponding section) but not for the CN, where the impacts are limited to modifications to the existing protocols, in particular the Mobile Application Part (MAP) defined in TS 29.002.

The functions performed by the UTRAN are different from those of the GSM BSS, in particular for the PS domain, so the nature of the interface between the CN and the AN is also different. The split of functions between CN and UTRAN, and a description of the transport services expected to be provided by the UTRAN to the rest of the network, are subject to a dedicated specification, TS 23.110. In the CS domain, the differences between GSM and UMTS are not particularly relevant (GSM’s A interface is quite similar to the UMTS’ Iu_CS interface) whereas in the PS domain, the UMTS’ Iu_PS offers “connections” (called “Iu Bearers”) contrarily to GSM’s Gb interface. This is an important milestone for enabling future support of end-to-end Quality of Service in PS domain, although this is not supported in this Release.

2.2.The UMTS Terrestrial Radio Access Network

While looking into data rates, the first phase of GPRS (Releases 97 and 98) allowed a maximum of 171,2 kbit/s (see TS 05.01). That was achievable by using all eight available timeslots, and in the best radio-traffic conditions. The radio interface used for UTRAN, a Wideband Code Division Multiple Access (W-CDMA), was originally designed to allow for the Release’99 a maximum (theoretical) peak rate of around 2 Mbits/s.

For this, a Direct-Sequence Code Division Multiple Access scheme was chosen. The figure below shows the process applied to the user information.

From user information to W-CDMA signal

The channel coding and multiplexing chain is variable and allows the system to "fit" the selected rate into the physical pipe, offering the flexibility to select a data rate versus the level of interference created. In other words, this gives the possibility of achieving a trade-off between network capacity, coverage and data/speech rate. The spreading multiplies in time the data sequence with a variety of CDMA spreading codes, creating the "wideband" dimension of the signal. The choice of the scrambling code ensures that the information relative to one user can be decoded while minimising interferences towards other users (orthogonality).

A simple example is provided here. Let’s defined the three channels i, j, and k. Upon transmission, channel i transforms the user information’s binary signal into:

and .

Channel j is defined by

and channel k by .

Upon reception, the decoder for channel i is specified by S_i = x₆ -x₅ -x₄ +x₃ -x₂+x₁(i.e. using the coefficients of C_i). This decoder applied to a bit sequence received on channel i provides:

S_i ( C_i) = (1x1) + (-1x-1) + (-1x-1) + (1x1) + (-1x-1) + (1x1) = 6 and S_i (

) = - 6

Whereas this decoder applied to a bit sequence received on another channel provides:

S_i ( C_j) = S_i (

) = S_i ( C_k) = S_i (

) = 0.

The speech codec chosen for UTRAN is suited to the data rate flexibility and can use different (up to eight) source rates, as mentioned in the clause on speech codecs.

The chip rate is 3,84 Mcps. This translates to an occupied bandwidth (99% of the total integrated power) of less than 5 MHz (see e.g. TS 25.101 or TS 25.102). Hence, the "Carrier spacing" is 5 MHz (compared to the 200 kHz of GSM/GPRS).

The larger occupied bandwidth of 5 MHz allows the system to benefit from the multipath nature of the radio propagation. At 3,84 Mcps, a receiver can separate the multipath components and combine them in an constructive way if the time difference between the two multipaths is at least of 0,26μs (a chip duration), i.e. 78 cm (as a reminder, the slot duration is 577 μs in GSM). This permits optimisation of the receivers to make the most of the diversity in the multipath propagation.

Two W-CDMA modes co-exist in UTRAN: the Frequency division duplex (FDD) mode and the Time division duplex (TDD) mode. In the FDD mode, two different frequency bands are used for the uplink and downlink directions. The frequency separation between uplink and downlink, or duplex distance, is 190 MHz or 80 MHz in ITU-R’s Regions 1 or 2 (the use of other duplex distances is not precluded). In TDD mode, the same frequency is used for both the uplink and downlink directions: intended to operate in an unpaired spectrum, the direction of the transmission is alternated in time, which allows asymmetric traffic in uplink and downlink depending on the number of timeslots that are configured for each link.

At higher layers, the definitions of the two modes converge.

The demodulation is coherent, in other words an internal time reference is used. Either the Common Pilot Channel (CPICH) (for FDD) or the Dedicated Physical Control Channel (DPCCH) can be used (as a result, the Base Station and User Equipment (UE) do not need to be synchronised to a third party system).

Alternatively, W-CDMA is sometimes used to refer only to the FDD mode. In this case, TDD is said to use the TD-CDMA technology (Time Division – Code Division Multiple Access). In this document, W-CDMA is said to be the technology both for FDD and TDD.

The frequency of the carrier is shown in the table below (the exact spectrum available remains country-specific):

While the Releases 97 and 98 GSM specifications only allowed up to two simultaneous Packet-Switched connections (one in the uplink and one in the downlink direction), the Radio Resource Management (RRM) of UTRAN offers the possibility to multiplex services with different quality requirements on a single connection, e.g. video, packet data and speech.

Selection of the properties of a radio bearer is possible, with its associated throughput, transfer delay (from real-time to best-effort) and data error rate (from 10% on frame error rate to 10^-6 bit error rate). This is aimed at fulfilling different applications, having different Quality of Service (QoS) requirements. Furthermore, bearer reselection is possible when e.g. the system becomes overloaded (on a 10 ms frame duration - basis). Bearer reselection is one side of the load controls, in effect a load-based packet scheduling correlated with interferences, given by the nature of the UTRAN interface.

Due to the intrinsic correlation between load and interference within UTRAN, output powers and their variations are of prime essence for controlling/allowing the load/services within the cells. Hence, UTRAN has defined a power control, both in the uplink and downlink directions (e.g. in FDD it is controlled on a 1500 Hz basis; it was of a maximum of 2 Hz in GSM). The Network instructs the UE to go up/down in output power. The Base Station uses a target Signal-to-Interference Ratio (SIR) to adjust its output power. One of the prime goals of the power control is to compensate the "near-far" effect in the uplink direction: if a UE was not able to rapidly adjust its transmission it could cause, for example, an undesirable noise rise at the base station receiver.

The overall architecture of the radio access network is shown in the red elements of the figure on UMTS and GSM Network Architecture (in clause “Architecture of the GSM-UMTS Platform”).

The architecture of this radio interface consists of a set of radio network subsystems (RNSs) connected to the CN through the Iu interface. An RNS consists of a radio network controller (RNC) and one or more entities called Node B. Node B is connected to the RNC through the Iub interface. Each Node B can handle one or more cells. The RNC is responsible for the handover decisions that require signalling to the user equipment (UE). The RNCs of the RNS can be interconnected through the Iur interface. Iu and Iur are logical interfaces, i.e. the Iur interface can be conveyed over a direct physical connection between RNCs or via any suitable transport network.

The figure “Radio Interface Protocol Architecture of the RRC Sublayer, L2 and L1” below shows the radio interface protocol architecture for the radio access network. On a general level, the protocol architecture is similar to the ITU-R protocol architecture as described in Rec. ITU-R M.1035. Layer 2 (L2) is split into the following sub-layers:

- Radio Link Control (RLC),

- Medium Access Control (MAC),

- Packet Data Convergence Protocol (PDCP), and

- Broadcast/Multicast Control (BMC).

Layer 3 (L3) and RLC are divided into control (C-plane) and user (U-plane) planes. In the C-plane, L3 is partitioned into sub-layers where the lowest sub-layer, denoted as radio resource control (RRC), interfaces with L2. The higher-layer signalling such as mobility management (MM) and call control (CC) are assumed to belong to the CN. There are no L3 elements in this radio interface for the U-plane.

Each block in this figure represents an instance of the respective protocol. Service access points (SAPs) for peer-to-peer communication are marked with circles at the interface between sub-layers. The SAP between MAC and the physical layer provides the transport channels. A transport channel is characterized by how the information is transferred over the radio interface.

The general classification of transport channels is into two groups:

- Common transport channels where there is a need for explicit UE identification when a particular UE or a particular group of UEs is addressed.

- Dedicated transport channels where a UE is implicitly identified by the physical channel, i.e. code and frequency.

The SAPs between RLC and the MAC sub-layer provide the logical channels. A logical channel is characterized by the type of information that is transferred over the radio interface. The logical channels are divided into control channels and traffic channels.

In the C-plane, the interface between RRC and higher L3 sub-layers (CC, MM) is defined by the general control (GC), notification (Nt) and dedicated control (DC) SAPs. These SAPs are not further discussed in this overview.

Also shown in the figure below are connections between RRC and MAC as well as RRC and L1 providing local inter-layer control services (including measurement results). An equivalent control interface exists between RRC and the RLC sub-layer. These interfaces allow the RRC to control the configuration of the lower layers. For this purpose separate control SAPs are defined between RRC and each lower layer (RLC, MAC, and L1).

Within the Release 99 standards, a number of schemes are available to extract maximal functionality from the system:

Transmission diversity:

The purpose is mainly in order to improve the reception quality in the downlink direction (exploit diversity gain to reduce power consumption for radio links in the cell i.e. increase the downlink capacity). Two antennas are used at the Base Station, the UE combines the received signals. This is a form of spatial/antenna diversity (performance requirements on the final recombination are defined for the UE).

Soft handover used in the FDD mode:

In effect, the soft handover scheme is a form of macro diversity. Two sectors from two different base stations communicate simultaneously with the UE (i.e. two radio interface links are used, using two power control loops). Both signals are received and used by the UE. In the uplink direction, the Radio network controller (RNC) within the network selects the best frame, at each interleaving period (every 10-80 ms).

Softer handover used in the FDD mode:

In this case, the two sectors belong to the same base station. The two signals (using the same power control loop) can be combined within the receivers of the UE and the base station. Two separate codes are used in the downlink direction, so that the UE can separate the signals. The difference is that in the uplink direction, combining is performed within the same base station. This combining can be performed in the baseband receiver of the base station, in comparison with the more drastic selection in the RNC performed by the soft handover.

Compressed mode used in the FDD mode:

When parallel measurements to another UTRAN frequency or a GSM frequency has to be performed (e.g. for UE reporting, in order to allow handovers), a pure parallel measurement would require a dual receiver. The Compressed Mode avoids this complexity, by operating the receiver in a slotted mode leaving some time for the UE to perform measurement on another frequency.

The compressed mode is available in the uplink and downlink directions. Several transmission time reduction techniques are available to allow this creation of gaps: "spreading factor reduction by 2", "higher layer scheduling" or "puncturing". The goal is to create "holes" in time, so that they can be used for measuring other frequencies. When this scheme is used, receiver and transmitter clearly need to be synchronised in time, so that they know exactly when the "holes" become available. That is why compressed mode patterns are precisely defined. Those compressed mode patterns define e.g. Transmission Gap Lengths (TGL).

Handovers between the two different modes, FDD and TDD, are possible.

As hinted at in the previous paragraphs, handover to/from GSM radio access networks are also possible. The aim was to have the same requirements available as in the intra-GSM case. All handovers (in dedicated mode), network-controlled cell-reselection from GPRS also applies to UTRAN as a target system. The user can expect a continuity of service between the two different systems (with the GSM limitations on e.g. the number of simultaneous connections). Dedicated messages have been introduced for the network to request handover / cell-reselection between the different systems. Thresholds indicating values to take into account for autonomous UE inter-system cell-reselection have also been introduced (with the possibility of using different thresholds within each source system, to avoid ping-pong effects).

A scheme allowing a quicker implementation and fulfilling e.g. agreements between operators of different systems (GSM and UTRAN) was also introduced for Release 99: the "equivalent PLMN" scheme. This allows an autonomous cell-reselection in (packet) idle modes for the UE, between different systems. In effect, a set of PLMN Identities is indicated by the networks, instead of one PLMN Identity. The UE can reselect between the systems, using the thresholds required for the decision, in effect as if it were roaming within the same PLMN.

This was once mentioned as a stand-alone item called “GSM/UMTS service continuity and equivalent PLMN”.

2.3. Mandatory Speech Codec for Narrowband Telephony Service

The scope of the feature “Mandatory Speech Codec for Narrowband Telephony Service” is to define the default speech codec for UMTS (both for FDD and TDD). This definition was in fact limited to a selection of one codec among several already existing ones: the proposed codecs were GSM AMR, IS127 EVRC, ITU G.729 and MPEG-4 speech codec.

A set of subjective tests was developed to compare the performance of the proposed candidates in different conditions: with and without background noise, with channel errors (using error patterns specifically developed by ARIB for this project), in tandeming and with music-on-hold. A number of organisations performed the required subjective tests with the proposed candidate speech codecs.

The codec selection was completed by April 99 and the codec characterisation was completed at a later date, mainly in TR 26.975 but not completed until Release 6 for the PS domain.

As a result of the selection, 3GPP adopted the GSM AMR (narrowband) speech codec as the mandatory default 3G speech codec, for the following reasons:

· The GSM AMR includes multiple (eight) codec modes providing the required flexibility to offer a toll quality speech service without compromising the system capacity;

· It includes the GSM EFR (at 12,2 kbit/s) and the IS136 EFR (at 7,4 kbit/s) offering a high level of compatibility with key 2G systems;

No other candidate codec provides better performances than the GSM EFR (highest mode of GSM AMR). The GSM EFR was found to provide the best performance with respect to the requirements set by ARIB for the mandatory speech codec, often exceeding the required performance level;

“tandeming” is the use of two codecs in the transmission path, e.g. in GSM, the voice is AMR-encoded in the source terminal, then AMR-decoded in the source BTS, then transcoded to be transported in the core network, and is again AMR-encoded in the destination BTS and finally decoded in the destination terminal.

This was once mentioned as a stand-alone item called “GSM/UMTS service continuity and equivalent PLMN”.

[1] “tandeming” is the use of two codecs in the transmission path, e.g. in GSM, the voice is AMR-encoded in the source terminal, then AMR-decoded in the source BTS, then transcoded to be transported in the core network, and is again AMR-encoded in the destination BTS and finally decoded in the destination terminal.

· At equivalent source rate, the internal codec modes of AMR always provide equivalent or better performance than the other candidate speech codecs. For example the AMR codec modes at 7,95 kbit/s (and 7,4 kbit/s) were found to be equivalent or better than the IS127 EVRC (8,55 kbit/s mode) or the G.729 (8 kbit/s);

· The AMR speech codec specifications were already approved by ETSI TC SMG. The corresponding C-code was released as part of the specifications. The completion of the 3GPP mandatory speech codec specifications in the time frame presented above would not have been achievable if the selected codec specifications and C-code had not already been publicly available.

· Speech quality is equivalent to wireline speech codec (ADPCM - G.726) in “No Errors” conditions

· The degradation is limited under normal operational conditions (with channel errors, in tandeming)

· It offers a good trade-off between complexity and performances for low cost implementation in 3G systems.

After the selection of the speech codec, the complete operation of the codec when used on top of FDD and TDD channels was defined, including the discontinuous transmission operation and/or variable rate operation. The definition of the best channel coding (based on existing bearers versus dedicated bearer with unequal protection) was defined by WG SA4 in cooperation with WGs RAN1 and RAN2.

Finally, the operation of the mandatory speech codec was fully characterized in multiple 3G operational environments (except for PS domain, left to Rel-6).

2.4.Codec for Low Bitrate Multimedia Telephony Service

The scope of the feature is to specify the default codec for multimedia telephony service for UMTS. In this release, multimedia telephony service is limited to low bitrate, circuit switched connections.

The specification of a default multimedia telephony codec enables terminals capable of low-cost, high-quality, real-time, two-way multimedia communications. It also allows interoperability of different manufacturers’ equipment, thus broadening the potential market for such devices.

Here again, the specification was in fact just a selection. The results on the tests were included in TR 26.912 on the quantitative evaluation of circuit switched H.324 based multimedia codecs over 3G.

ITU-T H.324/ANNEX C (Multimedia Telephone Terminals Over Error Prone Channels) was chosen as the core of the protocol. It makes efficient use of the radio resources and takes into account the error prone nature of radio based networks. 3GPP believed that it was essential to complete this set of mandatory requirements with a number of "recommendations" to help in the implementation of 3G terminals in order to guarantee enough error resilience and favour efficient terminal interworking. Where H.324/ANNEX C falls short, other relevant standards are used as follows:

AMR speech codec is adopted as the only mandatory speech codec for CS multimedia telephony services to offer the same level of speech quality as the basic speech service. Support of G.723.1 is defined in 3GPP as an additional option. Note that the ITU-T H.324 mandates the support of the G.723.1 speech codec, which is considered by the experts as providing a lower quality level than the higher modes of AMR.
H.263 was adopted as the only mandatory video codec. Support of MPEG-4 is also possible as an additional option. Note that H.324 also mandates terminals to support the less advanced H.321 video codec.
H.223 Annex B (which includes Annex A) is specified as the minimum multiplex error detection and protection level. This level was considered to provide an acceptable performance/complexity trade-off.

Additionally, call setup and termination are not defined in H.324/ANNEX C. 3GPP described it in TS 24.008 (and not in TS 26.112, withdrawn before its completion and replaced by 24.008).

The interoperability with other or existing systems was a low priority because of the low penetration of fixed access multimedia terminals and services. A consequence of this choice is that transcoding or gateway functions will be required for interoperability with existing multimedia terminals not supporting H.324 Annex C.

The related codec requirements were specified assuming that 3G systems will carry the multimedia data as a single data flow at the output of the H.223 multiplex, and not separate the different media flows before the H.223 multiplex to send them over separate radio access bearers. This decision was essentially guided by time constraints for the completion of the corresponding specification and the well established performance of H.324 in this configuration.

2.5. 3G audio-visual terminal characteristics

The scope of the feature “3G audio-visual terminal characteristics” is to specify the acoustic and visual performance of terminals.

The provision of speech, multimedia (e.g. video telephony) and wideband audio services in 3G terminals requires the specification of certain terminal characteristics, notably acoustic and visual (display/camera) characteristics. This feature develops the acoustic and visual requirements and the test methods needed to support these requirements for 3G speech and multi-media terminals in support of the mandatory speech service, the H.324 and H.323 narrowband video telephony service and wideband speech service work items. The set of requirements and test specifications were passed to TSG T for inclusion in its terminal specification work.

The specifications TS 26.131 and TS 26.132 detail the requirements for acoustic parameters, such as SLR (Send Loudness Rating), and the test methods to assess terminal conformance and performance.

Click here : for UMTS and GSM Features