Cellular Traffic Map Generation Using Geolocatable Event Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating traffic maps in cellular networks lack accuracy, as they rely on geographical maps and parcel weightings, which limit the precision of traffic distribution and network planning, especially in identifying high-traffic zones.
Innovation Solution
A method that uses geolocatable events to dynamically distribute traffic across pixels, weighting each pixel based on the number and distance of events relative to it, allowing for a more accurate representation of traffic patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If geographical maps with parcel weightings are used to distribute traffic, then the traffic map can be generated using available antenna traffic data, but the accuracy of the traffic map is limited by the coarse granularity of geographical parcel classifications
Solution Approach 1:
The patent segments the coverage area into smaller pixel units and further divides each pixel into Voronoi regions based on antenna coverage probability. This segmentation allows traffic distribution to be calculated at a finer spatial resolution than geographical parcel maps, improving traffic map accuracy while maintaining computational feasibility through systematic decomposition of the problem space.
Solution Approach 2:
The patent introduces geolocatable events as an intermediary data source that bridges the gap between antenna traffic data and spatial distribution. These events serve as observable proxies for actual user locations and traffic generation patterns, enabling more accurate inference of traffic distribution without requiring direct access to proprietary operator data or complex geographical parcel classifications.
2Area of stationary object
If the coverage zone of each antenna is automatically calculated using received power estimation, then the traffic distribution can be performed across the entire coverage area, but the computational complexity increases due to calculating received power at each geographical point
Solution Approach 1:
The patent segments the continuous coverage area into discrete pixels and further divides each pixel into Voronoi regions associated with specific antennas. This segmentation transforms the continuous problem of calculating received power across an entire coverage zone into a discrete problem of assigning pixels to antennas based on probability, significantly reducing computational complexity while maintaining comprehensive coverage.
Solution Approach 2:
The patent calculates received power and coverage probability only for discrete pixel centers rather than every point in the continuous coverage area. This partial action approach provides sufficient coverage information for traffic distribution purposes without the excessive computational burden of evaluating every geographical point, achieving a practical balance between coverage completeness and computational feasibility.
Data Source
AI summary
A method is provided for generating a traffic map of an area covered by a cellular network including a plurality of antennas each managing a traffic and designed to communicate with a plurality of terminals generating events which are at least partially geolocatable, the covered area being discretized by pixels each associated, using probability, with one antenna of the plurality of antennas. The method includes, for each antenna, distributing a quantity of traffic managed by the antenna over the pixels, this distribution being weighted for each pixel as a function of the number and distance of the geolocatable events relative to the pixel.