Page 236 - Big data - Concept and application for telecommunications
P. 236
5 Big data - Concept and application for telecommunications
6.3.1 Big-data-driven interference coordination
Within a heterogeneous network (HetNet) that has small cells, interference among macro and small cells
requires coordination in the time domain in lieu of the frequency domain, such as the enhanced inter-cell
interference coordination (eICIC) scheme in long term evolution (LTE)-Advanced.
The determination of an appropriate almost blank subframe (ABS) ratio of the macro cell to the small cells
depends on many factors, including the service types and traffic load in a given area. As is well known, service
behaviour in small cells vary with time. Moreover, the traffic patterns of individual services also change. Thus,
inter-cell interference does not remain constant. Therefore, the optimal ABS ratio essentially changes
dynamically with time. In a bDDN system, network analytics can be used to optimize the allocation of radio
resources. Resources can be allocated to adapt to both environmental and traffic changes based on
information gained from data analytics. To enable a quick response, some bDDN optimization functions can
be deployed at the micro cell eNB (MeNB), so that they can collect and analyse evolved node B (eNB)
originated raw big data in real time (e.g., service and traffic feature characteristics). Consequently, the
performance of each cell and users can be optimized by periodically processing raw data to obtain statistics
and automatically detecting traffic variations, targeting the prediction of eICIC-optimized parameters, such
as the ABS ratio. Moreover, a global optimization process can jointly optimize the location and traffic
demands of multiple eNB users. For instance, a certain small-cell eNB (SeNB) can be deactivated in order to
avoid interference with its nearby SeNB, which might have a larger throughput due to a higher
signal-to-interference-plus-noise ratio (SINR). Additionally, a reduction in energy consumption may be
another optimization objective to be considered.
6.4 Big-data-driven network operation (marketing)
By analysing user behaviour and preferences, as well as network status data, elaborate network operation
can be achieved.
Analysing customer network service use behaviour is crucial to understanding traffic. It is important to
understand customer objectives and their interactions in performing operations when network operations
are executed.
6.4.1 Big-data-driven prediction of customer upgrade in mobile network
With the evolution of the mobile network, faster speed improves customer experience. Mobile network
operators (MNOs) invite more customers to use a higher rather than lower generation mobile network.
MNOs need effective methods to predict customer preferences. MNOs can then guide customers to
complete a mobile network upgrade, from third generation (3G) to fourth generation (4G) or from second
generation (2G) to 3G. By analysing big data of customer service use behaviour, combined with terminal and
billing data, customers preferring an upgrade can be identified, and marketing operation strategy
personalized. Figure 6-2 shows the architecture of big-data-driven prediction of customer upgrade in a
mobile network.
228 Network and infrastructure