Michael Lawton, Product Marketing Engineer, Electronic Measurement Group, Agilent Technologies UK, discusses the process of validating inter-RAT handover and cell transitions using two base station emulators.
The ability to perform handovers without interruption of service is a key requirement for all cellular networks. Historically the requirement has focused on supporting a voice call during handover for a given cellular technology. This situation has now changed with the need to support cell transitions for multiple services (voice, video and data) and to seamlessly encompass a variety of wireless technologies (GSM/EDGE, WCDMA/HSDPA). The increased scope of handovers has added significantly to the complexity of the conformance test requirements. The designers of User Equipment (UE) are now looking for more cost effective solutions, which provide both broad test coverage and support their needs for validation and regression testing. This article describes the scope of this complexity and proposes a testing strategy that can be used to validate a User Equipment design.
I. Cell Transitions
Cell transitions are a key requirement for all cellular networks. The range of transmission for a battery powered mobile device is limited. Hence, for good coverage it is necessary for the mobile to move from one cell to another as it roams around the network. The network also benefits from the reduced power by working with smaller cells and applying frequency re-use, which provides spectral efficiency benefits.
With the evolution of both networks and their corresponding services there has been a significant increase in the complexity of handovers. This complexity arises due to the following:
1. Network operators are offering a wider range of services – including voice, video and data – each of which will possess different handover needs.
2. Practical roll out of 3G networks dictates the need for handover between 3G and 2G in order to maintain service availability for 3G customers.
II. Testing Cell Transitions
Testing cell transitions is a critical part of the conformance testing process for a UE with both RF and signalling requirements specified for each of the different technologies. These tests are a mandatory part of the UE verification process. The need to support multiple cellular technologies (3G to 2G handover) and services means that the scope of this testing is significant. For signalling conformance alone the mobility management and packet switched mobility managements sections of the 3GPP 34.123 specifications have nearly 150 cell transition tests defined. In order to complete this testing a complex set up of six independent base station simulators is required. This testing is both time consuming and expensive, furthermore designers are increasingly finding that this testing is not sufficient. Whilst the conformance tests do test all the handover mechanisms, they do not perform the testing in conjunction with application servers running real applications. This can lead to performance issues when testing a UE with a real network.
To overcome this limitation designers are choosing to do additional validation testing based around the emulation of a real network.
Validation Testing
Here the testing is focused on functional performance using a “one-box tester” (OBT) as a base station emulator. The OBT needs to emulate the overall functionality of the base station, however, it is not required to offer a protocol scripting interface for running signalling conformance scripts. This simplification makes the testing easier to perform and reduces the costs significantly. A set-up with two OBTs can be used to test multiple different handover types for a variety of application servers. The approach can be used to run automated testing for a wide range of scenarios, for de-bug of known issues or problem areas, or for performing random testing within a limited time frame. Clearly this approach is not a replacement for the rigour of conformance testing. However, it can serve as a powerful tool for increasing the scope of the testing, as well as providing a more economic test strategy when performing validation and regression testing on new platforms.
III. 3G Handover
The term handover is used to describe cell transitions that happen during a voice or video call. These handovers are all network initiated and are categorised as intra-frequency, inter-frequency or inter-RAT (Radio Access Technology). Within a WCDMA network, mobility between cells is managed by handover between different scrambling codes supported by the different cells all working on the same frequency (intra-freq). In this case it is possible for each base station to maintain it’s own link with the mobile, which allows the network to combine two signals and hence achieve a diversity gain at the edge of the cell. This mechanism is referred to as soft handoff diversity gain.
If the multiple links are formed from alternative sectors of the same base station (or node B) then it is referred to as softer handover. Strictly speaking this is a not an actual handover event but more a technique for improving signal quality. Using a One Box Tester such as the Agilent 8960 it is possible to simulate the effects of intra-frequency handover using a single instrument. This is achieved by using the OBT to manage multiple spreading codes.
It is sometimes necessary for a 3G handover to take place between frequencies. This may be done in order to maintain coverage or to balance the load between respective carriers. Inter-frequency handover is more complex than intra-frequency due to the fact that during a call, the receiver of a 3G phone is constantly listening to the serving base station. Therefore it either needs to make a handover without measuring the characteristics of the target cell or the phone needs to use what is termed compressed mode in order to make the necessary measurements. When a phone is instructed to make a handover without having gathered a measurement report for the target cell this is termed a ‘blind’ handover. In compressed mode the downlink transmission is briefly halted in order that the phone can make a measurement.
In order to support the period of no transmission on the downlink the signal just prior to and just after the break is sent with either a change to the spreading factor (and appropriate transmit power), or following a change to the higher layer scheduling. These changes are made so that the gap can be accommodated without significant loss of data.
One final handover type supported within 3G networks is the inter-RAT handover from a serving 3G to a target 2G cell. 3G networks use this to provide both 3G service hot spots and full 2G coverage for voice and low data services. For single receiver mobiles the handover will either be blind or will use compressed mode in order to allow the mobile to measure a neighbouring 2G cell.
IV. 3G Packet Switched Data Cell Changes
3G cell changes arise either when the phone is idle, has a PDP (Packet Data Protocol) context, or is actively transferring packet switched data.
All mobiles that are either out of service or in an idle state need to be able to initiate a cell selection or idle mode cell re-selection procedure. This allows mobiles both to join the network and also to maintain network availability as they roam between cells in standby mode. As a 3G mobile moves around the network it monitors each of the intra-frequency cells it can see within its ‘active set’. If the signal quality from the attached cell goes below a certain threshold, or an alternative cell becomes significantly better, then the mobile will initiate an idle mode cell reselection and attach to a new cell.
With 3G, once the phone has an active PDP context then all cell changes are network initiated by a packet cell change order command.
If there is an Inter-RAT packet cell change order during active data transfer then clearly there is a requirement for the physical radio bearers to be re-configured as the cell change takes place. If the capability of the two cells is significantly different then the re-configuration is likely to result in a change to the data rate. Using a base station emulator – in conjunction with real application servers – is very useful as it allows the user to test the actual experience of using this application in such challenging and dynamic conditions.
V. HSDPA handovers
In addition to the handover and cell transitions described in section III and IV there are also some special cases relating to new physical layer technologies such as HSDPA (High Speed Downlink Packet Access).
HSDPA is a new technology for 3G which seeks to maximize downlink data throughput by using complex modulation schemes to exploit periods of favourable RF conditions. Mobility for HSDPA is different from W-CDMA packet switched data because HSDPA does not use soft handover. Instead the mobile monitors all base stations within its active set and informs the network when there is a change in the ‘best’ cell. The network then performs a cell transition and re-configures the HS-DSCH (High Speed Downlink Shared Channel).
VI. Two Cell Test System Configuration
A LAN connection is used to facilitate message exchange between the two boxes. Prior to initiating cell transition commands the following set-up steps are required:
1. If remote control is required then establish GPIB connectivity to both units
2. Connect the OBT’s using either a single cross-over cable or a small Ethernet network (the network allows the use of both logging and application servers).
3. Configure each OBT to represent the two cells, ensure neighbour cells are configured correctly
This set-up is now ready for two-cell testing. Network initiated cell transitions are implemented either through the manual user interface or by sending the instrument a SCPI (Standard Commands for Programmable Instrumentation) command. The OBT receives measurement reports but does not use the information to perform any automatic handover choices, although the data is there to support this approach if required. The characteristics of a given cell change are determined by both the state of the serving cell (technology and active call type) and the settings for the target cell. For example, if on a WCDMA voice call with the second cell configured as a GERAN cell then an inter-RAT handover will be executed. In addition if the OBT has compressed mode enabled, then it will be instructing the mobile to take measurement reports. The instrument will receive and display the measurement report data but only perform cell changes as instructed by the user.
VII. Summary
This article has shown how two one box testers can be used to perform validation testing for the handover performance of a given UE. This kind of testing is not a replacement to the rigours of conformance testing. It is however a useful addition and offers the following benefits:
1. Cost advantages for OBT vs. scripting based testers
2. Ease of use through the programming of a command based base station simulator rather than writing C++ or TTCN scripts to drive a protocol signalling tester
3. Perform real world testing that incorporates application servers to test handover performance during a ‘live’ application, for example ftp or video streaming
4. Use of regression testing to ‘stress’ mobile and search for fault conditions.