While mobile networks are getting more complex with several technologies operating and interacting simultaneously to meet new customer demands, market saturation continuously challenges mobile operators' revenue streams. The resulting demand for CAPEX rationalization makes optimization crucial, since it maximizes the value of the deployed assets.
The area of optimization can be divided into two main parts: initial and continuous optimization. Initial optimization is performed as part of a network implementation to get the network fully prepared for the launch. The aim is to ensure that the agreed objectives for coverage, quality, and service performance will be met. These objectives are often defined as target values for sets of key performance indicators (KPIs) for network performance and quality of end-to-end services (QoS).
Continuous optimization is an integral part of network operations due to the fact that networks are in constant change:
new network elements are regularly taken into operation;
new services are introduced and coverage is extended;
and the traffic load from users shows both short and long term variations.
Continuous optimization can be seen as a cyclic process where at first the network and end-to-end services performance is assessed and necessary data is collected. Next, the data is processed and analyzed; finally, corrections and network changes are decided and implemented. The input data sets are collected from various sources ranging from statistical counters, generated in the network elements, to measurement data, provided by drive tests as well as walk tests conducted with handheld measurement systems. Due to the importance of keeping OPEX under control, these solutions must include measurement systems that are:
efficient and productive;
are easy to use with low maintenance efforts;
deliver rich sets of data;
and provide automatic data processing and accurate analysis.
Even when RF planning is done carefully, the resulting coverage after a network rollout is affected by so many factors that one planning tool alone cannot take them all into account. In order to have the coverage objectives verified with a high degree of accuracy, the coverage must be measured. This is best done with a dedicated scanner (such as R&S®TSME and R&S®TSMW) that can measure coverage parameters such as signal strength; signal quality; and interference per RF channel and cell on multiple technologies simultaneously. Another useful feature of a scanner is the ability to detect broadband interferers and to measure uplink activity.
Services, such as voice, data, and messaging depend on a multitude of network parameters that control the various stages of connection, e.g. cell selection and reselection; channel access and setup; and radio resource allocation and mobility procedure including handover. These measurements are performed with commercial user equipment such as smartphones and data sticks, preferably the actual models that are used by customers. Part of the initial optimization might be to trial and tune different parameter sets to achieve a best possible balance between trade-offs that are part of the overall objectives such as service coverage vs. capacity, and setup reliability vs. latency.It is common to start with voice service, since it is simple and easy to understand; then move on to packet switched (PS) data and messaging. For PS data, the modulation and coding scheme allocation, channel switching, power control, and mobility are key aspects to achieve the optimal balance between performance and capacity.For continuous optimization, it is crucial that the solution can process large sets of data and automatically find and categorize any experienced network problem as well as provide a root cause analysis, including:
data service including application layer, IP and PS data bearers (LTE, HSPxA, EDGE/GPRS);
The quality of end-to-end services, such as voice, data, video, and messaging, is directly affecting customer satisfaction. Because of that, mobile operators must regularly verify quality of service (QoS) and quality of experience (QoE) and initiate optimization tasks when it goes below target. Key performance indicators (KPIs) are defined in technical standards produced by organizations such as ITU-T and ETSI. Standards include requirements regarding the technical functionality of the measurement system, i.e. describing the testing procedures and defining the algorithms and KPI calculations.
Since quality perceived by end users is influenced by many factors – such as the actual equipment used (mobile device manufacturer and model), the user’s location, behavior, and the network performance – measurement solutions must address them adequately. This includes:
connecting and using multiple devices (preferably the actual models that are used by customers);
being efficient and easy-to-use in various environments during prolonged measurement campaigns;
and allowing flexibility in automatically executed service tests with preconfigured rules for timing, sequencing, and device control.
One key element affecting end user satisfaction is speech quality, which can be degraded by factors such as compression, transmission artefacts, bandwidth limitations, or imperfect restoring of voice in case of losses. All of these stem from different sources and occur for different technical reasons.
Applications and video services
Another element is the video quality where a wide range of popular applications and video services with different functionalities need to be tested and analyzed from a functionality, performance, and video content quality perspective.
The introduction of new techniques, such as adaptive bitrate streaming, disrupt the existing testing methodologies because visual quality is not fixed anymore but adapts to the conditions of the bearer and the phone. Such techniques have been introduced by video on-demand service providers such as YouTube to enhance efficiency over large distributed networks and improves customer experience.
For end-to-end services optimization, it is important that the solution can process large sets of data coming from various measurement devices. The solution must be able to automatically find and categorize the devices to determine and classify the causes behind quality degradation, including the quality analysis of voice, video, and data.
Indoor environments, such as offices, shopping malls, airports, subways, and event venues, have become very important for network operators, since nowadays more than half of the voice and data traffic is generated in such locations. It is therefore essential to pay special attention to RF conditions and to perform a dedicated network optimization.
The best way to collect relevant data is to perform walk tests.
There are particular requirements for test tools to be used indoors, among others:
portability is important as are lightweight hardware and good operational autonomy since the equipment is battery powered;
navigating and documenting the positions must be possible without GPS using local floor plans, downloaded in electronic format or captured on spot with a camera;
forcing functions features that allow the device to use the specific RF channels and/or cells of an indoor system.