SON Self Optimization (Self Optimizing Network)

- Self Optimisation or optimization aspects of Self Organising Networks, SON, how self-optimizing network techniques operate with cellular telecommunications.

Optimization is required for ensure that once a cell has been installed it operates to its best level of efficiency.

Self optimizing network techniques can analyse the performance and change the operation of the network so that it best meets the needs of the operator and users.

Even though the network may have been automatically configured aspects of the network operation can change.To accommodate these changes, self-optimising network techniques can be incorporated into the network.

Reasons for self-optimizing networks

One of the major elements within SON optimization techniques that can be used. As the environment for the base station, eNB may change after installation and configuration, there is a need to continue to optimise the operation on a regular basis.

Some of the reasons for a change in the environment may be:

  • Change in propagation characteristics:   SON optimization of the network can help take out the effects of any changes to the propagation conditions. These could arise from new buildings going up, or coming down etc. Even changes resulting from leaves falling in autumn can have a significant effect.
  • Change in traffic patterns:   As time progresses usage patterns may change. This could result from increased concentrations of users, from new housing, changes resulting from more people being on holiday, schools being on vacation, or any one of many hundreds of reasons. These can result in further optimisation being required to re-asses the best operational characteristics for the base station, eNB.
  • Change in deployments:   There could be many reasons for the change in deployments in the area. Other base stations, eNBs could have been optimised and changed their characteristics, alternatively new base stations may have been deployed and their operation could affect that of others.

These reasons all mean that to obtain the optimum the performance, it is necessary to optimise the network on a regular basis.

Types of self-optimizing network functionality

There are a number of areas where self-optimisation of the network is undertaken.

  • Mobility robustness optimisation
  • Mobility load balancing and traffic steering
  • Energy saving
  • Coverage and capacity optimisation
  • RACH optimisation

These main areas are addressed below.

Mobility robustness optimisation

The mobility robustness optimisation functionality is included within the self-optimising network routines to enable robust mobility and handovers within the mobile network. There are a number of aims for the mobility robustness optimisation:

  • Minimise dropped calls:   Dropped calls are one of the major causes of customer dissatisfaction and therefore this is a key driver. To improve the perceived quality of the network, reducing the rate of dropped calls is essential.
  • Minimise unnecessary handovers:   Unnecessary handovers lead to inefficient use of network resource and the increased chance of a dropped call. Often many unnecessary handovers take place as a "ping-pong" between two cells as the signal level especially at the cell border varies between the two cells where small changes in position can lead to multiple handovers between the same two cells.
  • Minimise idle mode problems:   When coming out of an idle mode it is necessary for the handset to be able to quickly setup the connection.
  • Minimise radio link failures:   Radio link failures occur at many times. Obviously the first step is to ensure good coverage so that the failures do not occur, but also if they do occur to have in place a capability to quickly re-establish the connection.

There are a great many factors that affect the mobility robustness within the SON self-optimisation. Cell changes are one major issue, and they are initiated dependent upon signal strength indications. As cell handover is the key issue it is necessary to categorise the two main types of handover, and to understand their operation. In this way solutions and their methods can be better understood:

  • Intra-frequency handover:   This form of handover is one that takes place between two cells or sectors using the same carrier frequency. This form of handover needs to take place when the strength of a new cell is greater than that of the existing cell. As they are all on the same frequency it is relatively easy for the handset or UE to monitor the strength of existing and neighbouring cells. This is particularly important because for large cells in particular the overlap between cells can be small, and when moving it is possible to transition from one cell area to another. Fast action is therefore needed if the link is not to be lost, although this can lead to more handovers along the cell borders.
  • Inter-frequency handover   This form of handover takes place when the carrier frequency is changed. It makes monitoring the strength of adjacent channels more difficult because the handset needs to monitor multiple channels, but reduces issue with interference apparent at the edges of the cell when two or more cells use the same frequency. Again, fast action is needed to ensure the handover is initiated in time to retain the link before the handset or UE falls out of range of the current cell. However inter-frequency handovers are less prone to interference and can therefore be required less suddenly.

These forms of cell handover give an insight into some of the problems that can occur with handover.

To minimise the number of problems that occur, SON self-optimization processes provide a number of capabilities that can reduce the handover issues:

  • The SON self-optimization solution should better optimise the cell boundaries to gain a better idea of where they occur, and to try to limit the pin-pong effect.
  • Carry out optimisation of cell boundaries more often to accommodate any changes that may arise from new cells being added, changes in propagation characteristics resulting from a variety of occurrences from the construction or demolition of buildings to the propagation changes cause by seasonal issues such as leaves on trees, etc.
  • Improve measurements capability and statistics analysis of the self-optimisation solution relating to handover so that more accurate decisions can be made.
  • Decentralise the handover decision making to enable swifter and more accurate decisions to be made.

Load balancing

The aim of the load balancing elements of the SON optimization are to try to level out the data hotspots as much as possible. Some cells are likely to be more heavily loaded than others and methods are used to try to even the load out, proving the most effective service for users while maintaining overall capacity while keeping investment to reasonable levels.

With data usage rising exponentially, management of the data load is a key element of SON self-optimisation. However load balancing and traffic steering require sophisticated routines within the SON self-optimization elements to be able to make complex decisions regarding the optimum solutions for any given case.

Some of the key processes within the load balancing and traffic steering software include the following:

  • To make the data traffic more even over the network and thereby reduce and data hotspots the SON optimisation should enable traffic to be moved from highly loaded cells to less loaded neighbours within the limitations of coverage and interference.
  • To offload traffic from the macro cells to the smaller low power cells such as HeNB or Wi-Fi. In this way better use is made of the macro cells.
  • To ensure optimum performance, any handsets or UEs that are moving should not be handed down to smaller cells as they will soon move out of range requiring a large number of handovers. The self-optimisation software should be able to detect signs of movement.

Energy saving

With general green issues and costs both acting as drivers, energy saving is becoming an increasingly important feature for SON self-optimization network functionality. Energy savings are motived by the need to reduce carbon dioxide emissions as well as cost savings from the reduced power consumption.

Energy savings are achieved with both the UE, handset and within the network. Obviously different strategies are used for both, although many of the fundamental concepts are the same.

As far as SON optimization is concerned the major energy savings can be made within the network, and in particular within the eNBs.

Energy savings can be made in many ways within a self-optimizing network. Traditionally not a lot of functionality has been added to networks to provide energy savings, but as traffic drops dramatically over night there are significant opportunities for energy savings.

There are a number of options that can be implemented to provide an energy self-optimizing network:

  • Reduce active carries for off-peak times:   Many base stations transmit a number of carriers to enable the data capacity requirements to be met. During off peak periods, the number of carriers that are active could be reduced with the resultant reduction in power required.
  • Sleep mode:   In some areas it may be possible to put some base-stations into a sleep mode and increase the coverage of others. In business areas where there is a very high level of usage during the day, at night usage may be next to zero. Similarly at weekends the same effect would be noted.

    Accordingly it may be possible to put some base-stations, eNBs to sleep over night or at other times, and increase the coverage or more fully utilise an umbrella cell. Care would need to be taken with this strategy not to reduce the coverage and leave holes that may provide an annoyance to any users that may be around.

    Typically sleep mode should be a mode from which the base station, eNB could quickly be awoken.
  • Local generation:   While most base stations utilise mains / grid for their power source, more options are now available for local generation. This has been particularly applicable to developing countries where power supplies may not be as easy to provide. However, this is now more of an option for many especially with renewable energy sources such as wind and solar. While the costs of renewable energy are not cheap, the carbon footprint of the network may become a standard which could be adopted.

No doubt with more development further energy saving solutions can be added to the network.

Coverage and capacity optimisation

The concept behind this element of the SON self-optimisation is to adapt parameters such as antenna tilts, transmitter power levels and the like to maximise coverage while optimising the capacity by ensuring the inter-cell interference levels are minimised.

The coverage and capacity optimization, CCO, aspect of the self-optimizing network can create some significant advantages, although it is very time consuming and expensive to manage manually.

There are a number of ways in which it can be achieved:

  • Adjustment of antenna parameters:   In order to provide CCO using an adjustment of the antenna parameters, a Remote Electrical Tilt, RET antenna is required. Previous generations of base stations only enabled manual adjustment of the antennas. Now it has become viable for them to be electrically steered.

    The adjustment is generally the angle if tilt. Moving it upwards increases the boundaries of the cell, although care has to be taken to ensure that coverage is maintained close to the antenna tower.

    The adjustment can be made either mechanically or electrically. However when a system is installed with electrical tilt, sometimes mechanical tilt is also needed to give a wider range because the electrical tilt is limited.

    The antenna tilt needs to be carefully adjusted. If it is lowered too much, then the cell boundaries will be brought inwards and coverage holes may appear causing issues with handover. If it is adjusted too high, then the coverage will be extended and it may result in interference levels rising at the cell borders where signals from the adjacent cell will also be received if they use the same channel. Accordingly the self-optimization and adjustment needs to take in these requirements.
  • Adjustment of power level parameters:   While antenna adjustments may be the most obvious solution, self-optimization and adjustment may also be applied to the power levels.

    In many respects, base station transmitter power optimization is more challenging than using antenna tilt and control. There are issues with amplifier behaviour and also issues with reciprocity with the handsets. It is possible to increase the transmitted power so that the handset receiver can receive the base station further away, but it may not be possible for the handset to increase its power sufficiently to match any improvements specially at the cell edge where it may already be operating close to its maximum level.

RACH optimisation

The Random Access Channel, RACH uses valuable resources. It is a necessary element of the access scheme, and therefore a balance needs to be obtained between devoting sufficient resources to enable access and sacrificing performance and valuable resources.

The RACH needs to be accurately configured so that it provides a sufficient number of random access opportunities for any handsets or UEs operating within the cell. Obviously the more UEs likely to access the base station or eNB, the greater the number of opportunities needed.

The automatic management of RACH requires the network to be continuously optimised to meet the changing conditions. To achieve this two separate self-optimization mechanisms are used:

  • Handset reporting:   This part of the feedback routine for the RACH network self optimization. The UE or handset reports the number of MAC preambles that are issued before the RACH access was successfully completed. Additionally any collisions that are detected are also reported.

    These SON self optimization routines are defined as separate routines so that they are not activated on any networks or base stations that do not support SON.
  • Inter base station data exchange:   The exchange of information between base stations as part of the self-optimizing network element for the RACH is needed because RACH configuration may affect other cells.

    To provide neighbouring cells with information, data is provided over the X2 interface on LTE networks.

SON self-optimization networks employ many techniques to provide their overall functionality. Although self-optimizing networks provide significant gains in performance the routines they run are not simple and therefore the implementation of SON self-optimisation required a significant investment by the operator.

By Ian Poole

<< Previous   |   Next >>

Share this page

Want more like this? Register for our newsletter

Too good to be true - the cost of counterfeit electronics and how to avoid them Miguel Fernandez | Avnet EMEA
Too good to be true - the cost of counterfeit electronics and how to avoid them
The issue of counterfeit electronic components is one that has troubles the electronics industry - using them can have some major issues, everything from being removed from a preferred suppliers list to a reduction in quality.
Online - Fundamentals of Modern RF and Wireless Communications Engineering
This on-line course enables you to quickly get up-to-speed & understand key concepts of modern radio frequency, RF & wireless communications systems

More training courses

Low Loss Dynamic Compression of CPRI Baseband Data
Read this paper from Altera that describes a method of using Mu-Law compression for Gaussian-like waveforms providing an efficient methodology.

More whitepapers
 is operated and owned by Adrio Communications Ltd and edited by Ian Poole. All information is © Adrio Communications Ltd and may not be copied except for individual personal use. This includes copying material in whatever form into website pages. While every effort is made to ensure the accuracy of the information on, no liability is accepted for any consequences of using it. This site uses cookies. By using this site, these terms including the use of cookies are accepted. More explanation can be found in our Privacy Policy