Search

In cellular networks, users that make frequent handovers and have a low time-of-stay in a cell (i.e., highly mobile users) might have a negative impact on the network performance. Furthermore the Quality of Service (QoS) experienced by these users might be low. This paper introduces a Self-Organising Network (SON) function, called the High Mobility SON function, that aims at reducing the amount of short stays in a multi-layer Long-Term Evolution (LTE) network. It does this by predicting the mobility behaviour of currently active users based on measurements, which were collected by users that were active in the past. Based on these predictions, the SON function aims at refraining from handovers to cells in which the user is likely to stay for a small amount of time, and at steering the user more appropriately. To assess the ability of the SON function to achieve its goals, simulations were performed in a scenario in which both macro and micro cells are deployed. Results show that the developed SON function is able to reduce the number of handovers by 17-23% and the number of short stays by as much as 43-49% at the cost of reducing the spectral efficiency by 13-15%.

A Self-Organising Network (SON) represents an approach where the optimisation of a mobile radio network is automated through a set of independently operating SON functions. These SON functions, however, require to be configured in order to allow for an optimised network performance with respect to technical objectives defined by the network operator. The SEMAFOUR demonstrator shows a concept for SON management based on operator objectives, where the SON function configuration is performed in an automated way. The demonstrator illustrates the different aspects and complexity of the management of SON in a heterogeneous network. © 2015 IEEE.

The demonstrator shows a self-management system for heterogeneous mobile wireless networks that uses context-specific and weighted Key Performance Indicator (KPI) target values defined by the operator to automatically and autonomously configure and control the operation of Self-Organising Network (SON) functions such that they contribute to achieving these KPI targets by appropriately optimising the network configuration. Changing KPI targets, context or weights leads to an automatic re-configuration of the SON functions by using a policy system, and the impact of the changes to the policy and the network configuration can be seen and traced in the demonstrator's realistic network scenario and KPI charts.

With an increasing number of SON functions deployed in cellular radio networks, conflicts between the actions proposed by independently-designed and distributed SON functions may arise. The process of minimizing the occurrence, and the consequences, of such conflicts is referred to as SON coordination. SON coordination can be achieved either by a separate entity, a SON coordinator, controlling the actions of each SON function during operation, or in the design of the SON functions as such, through a conflict free SON design. In both cases, the SON functions should be designed in a way compliant with the coordination method used. This paper proposes a number of SON design principles to apply in order to achieve this. In the case of an operational SON coordinator, SON functions should be able to deal with the possible actions that are taken by the SON coordination function in order not to worsen the problems experienced in the network. In the case of conflict free SON design the aim is at removing potential conflicts already in the design phase. The ambition of this paper is not to provide a complete set of design guidelines that address all aspects. Instead, the target is to open up the discussion on the distributed SON design principles in the community.

In mobile cellular networks the handover (HO) algorithm is responsible for determining when calls of users that are moving from one cell to another are handed over from the former to the latter. The admission control (AC) algorithm, which is the algorithm that decides whether new (fresh or HO) calls that enter a cell are allowed to the cell or not, often tries to facilitate HO by prioritising HO calls in favour of fresh calls. In this way, a good quality of service (QoS) for calls that are already admitted to the network is pursued. In this paper, the effect of self-optimisation of AC parameters on the HO performance in a long term evolution (LTE) network is studied, both with and without the self-optimisation of HO parameters. Simulation results show that the AC parameter optimisation algorithm considerably improves the HO performance by reducing the amount of calls that are dropped prior to or during HO. © 2011 IEEE.