Improving Investment Effectiveness Based on Geographical Network Data Mining and Analysis


Improving Investment Effectiveness Based on Geographical Network Data Mining and Analysis

精品网 技术专题1

Compared with traditional voice services, the traffic distribution of data services remains off balance. Statistically, 20% of the sites carry 80% of the network traffic and the data traffic on some of the hottest sites is dozens of times greater than the average for other sites. In addition, because the spectrum resources are insufficient and site resources in highly populated urban areas are difficult to obtain, the traditional flattened, single-layered macro network cannot carry dense data traffic. And so, hence, the multilayer stereoscopic Heterogeneous Network (HetNet) is born. How do we effectively find and identify hotspot areas? How do we reduce the interference between macro and micro networks? How do we deploy proper resources in correct locations? How do we accurately ensure VIP users' experience? How do we satisfy the requirements of different service objectives, maximize utilization of the network resources, and ensure the return on investment (ROI)? These are the key questions operators face on there path to network evolution. This also raises the bar for network planning and optimization to support operator networks' smooth evolution: better skills and the ability to generate more accurate geographical data mining. For example, operators represented by Deutsche Telekom have begun to provide geographic visualization for the entire network. They use grid-level correlation analysis to directly detect problems, and use geographical data analysis as part of their routine maintenance on their basic platform.

Location, Simulation and Modeling are Three Core Advantages of Geographical Data Analysis

Generally, mobile operators and third parties evaluate network performance based on NE traffic statistics or on the Drive Test (DT) data from specific terminals. They can optimize or even replan the network based on problems found in the tests. However, traffic performance data only measure cells and does not contain any user information. Therefore, it cannot help locate an accurate user position. The cost of DT on the entire network is very high. In addition, DT only provides outdoor coverage; it excludes indoor areas. Relevant statistics show that 70% of voice services and 80% of data services are consumed indoors. In addition, most DTs can only be performed at night, which represents different network coverage and interference scenarios from those on a busy network during daytime. In other words, data obtained on NEs or DT of specific terminals cannot fully and accurately demonstrate the real online state of the terminal users. To solve these problems, we need to provide a geographical monitoring management and optimization platform based on real network performance that users experience in specific scenarios. Services, geography, and location accuracy are topics that need to be looked at seriously and simply cannot be ignored. Especially in future HetNet scenarios, the improvement of location accuracy will be paramount. Calling positions are identified based on geographic locating engines, and the network performance of every site is quickly and accurately evaluated using grid convergence. The network coverage, traffic hotspots and abnormal events are geographically displayed. In addition, considering that a large number of services are consumed indoors, how do we associate practical building stereoscopic maps and internal environments with the platform geography, and how do we match modeling indoor-outdoor coordination simulation maps with the real maps? Geographical modeling determines whether the service scenarios, simulation capability and location specific solutions can be implemented. The relationship between geographical network data mining, and network planning and optimization may be compared to the relationship between an engine and a car. Without a good engine, the car is obviously underpowered.

6D Model Algorithm Ensures Accurate Site Planning and Effective Investment

Huawei is the first in the industry to put forward and apply a 6D business model. Based on the 50-meter traffic of Assisted Global Positioning System (AGPS) MR locating accuracy, this model performs geographical analysis of data along the following six dimensions: traffic (including data and voice), user, terminal, revenue, complaint, and coverage. By comprehensively analyzing platform data and self defining the weight of operators, this model converts operators' different network planning requirements into differentiated network planning solutions, and shows a geographical display of operators' customized hotspot areas and priority level sequencing of site planning investments. Compared with the traditional coverageand-capacity-oriented planning method, the 6D model algorithm identifies network hotspots more accurately and ensures network coverage quality, which satisfies service requirements and maximizes ROI.

Data along with different dimensions of the network layer is a precious resource for operators and geographical data mining is a core capability for future network planning. The geographical and network data analysis platform helps implement precise values for site selection and accurate site planning. The Huaweideveloped Smart MBB platform helps perform planning and designing, network optimization, test analysis, performance monitoring and analysis, and geographical analysis for HetNet macro-micro network coordination, Multi-RAT coordination, TDD/FDD coordination, indoor-outdoor coordination, and multi-RAT coantenna coordination. These maximize efficiency for sites accuracy; and construct a coordinated network with highly-effective resource utilization. By comprehensively analyzing big data, accurately identifying high value hotspots and analyzing user behavior and revenue streams, geographical network data mining can effectively improve ROI. More than 500 Huawei global experts in the MBB field are dedicated to the research and development along with satisfying customer business needs, accurately distributing resources and to effectively planning solutions. Huawei has obtained more than 50 patents in the field of geography and has provided network planning and designing services for more than 500 operators around the world. Huawei helps operators plan and optimize network effectively, realize smooth network evolution, improve ROI and accelerate business success.