OD Matrix Generation Using Cellular Network Data

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for estimating origin-destination (OD) matrices in transportation planning, such as surveys and link traffic measurements, are expensive, inaccurate, or limited by the availability of infrastructure, while existing cellular network-based approaches provide rough estimates and are not scalable for sparse data.

Innovation Solution

A computerized system that generates OD matrices using network data from cellular telecommunications, including a receiving unit for user IDs, cell IDs, and timestamps, a station identification unit to determine geographical positions, and an OD matrix generating unit to create and merge sub-matrices, utilizing Hadoop map/reduce for scalable processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If surveys are used for OD estimation, then information quality is improved, but cost increases significantly

Engineering Contradiction:
Improveinformation qualityVSAvoidcost
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent uses cellular network data as an intermediary to estimate OD matrices without conducting expensive surveys. The network data serves as a mediator that provides sufficient information for OD estimation while avoiding direct interaction with travelers, thus reducing cost while maintaining reasonable accuracy.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent creates a virtual copy of survey data by using cellular network measurements to infer OD matrices. Instead of collecting actual survey responses from travelers, the system copies the necessary information from network data, achieving similar analytical goals at lower cost.

Inventive Principle:
Principle #26Copying

2Quantity of substance

If link traffic measurements are used for OD estimation, then cost is reduced, but measurement precision deteriorates

Engineering Contradiction:
ImprovecostVSAvoidOD estimation accuracy
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent transitions from one-dimensional link traffic counts to two-dimensional OD matrices by using cellular network data that captures origin and destination information separately. This dimensional transformation enables more precise OD estimation while maintaining cost efficiency.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent changes the parameters being measured from simple link traffic volumes to detailed OD pairs with origin, destination, and temporal information. By changing measurement parameters from aggregate counts to disaggregated trip data, the system achieves higher precision without proportionally increasing cost.

Inventive Principle:
Principle #35Parameter changes

3Quantity of substance

If measurement devices are installed along roads, then link traffic data is obtained, but device complexity and maintenance requirements increase

Engineering Contradiction:
Improvelink traffic dataVSAvoidmeasurement device infrastructure
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent leverages the cellular network infrastructure that already serves travelers for communication purposes. The same network towers and cellular devices that provide telecommunication services also collect location data for OD estimation, eliminating the need for separate measurement devices and reducing overall system complexity.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The cellular network infrastructure performs multiple functions: providing telecommunication services and simultaneously collecting location data for transportation analysis. This multi-functionality eliminates the need for dedicated measurement devices, reducing device complexity while maintaining data collection capabilities.

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Quantity of substance

If existing cellular network approaches are used, then cost is reduced, but measurement precision and scalability deteriorate due to sparse data

Engineering Contradiction:
ImprovecostVSAvoidOD estimation accuracy
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent merges data from multiple cellular network sources and combines it with OD estimation algorithms to overcome data sparsity. By integrating various network data types and combining them systematically, the system achieves better precision than individual approaches while maintaining cost efficiency.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent performs preliminary data processing and filtering of cellular network data before OD estimation to improve data quality. By preprocessing the data to remove noise and fill gaps, the system enhances measurement precision while working with the cost-effective cellular network data source.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2740115B1Generating an OD matrix
Publication Date: 2023.05.24 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)
  • EP2740115B1 patent drawingFigure 1~8
  • EP2740115B1 patent drawingFigure 2
  • EP2740115B1 patent drawingFigure 3

AI summary

Computers, a user equipment, a computer system, computer program products, computer programs and a method for a computer system are disclosed. The method comprises the steps of: - receiving user IDs, associated cell IDs and time stamps associated with the cell IDs and the user IDs, - requesting and receiving, from a cell ID database, geographical coordinates corresponding to the cell IDs, - finding stations and times for each user ID based on the geographical coordinates and time stamps associated with each user ID; - exchanging the stations with a place for each one of the stations, - generating at least one OD sub-matrix associated with each user ID based on places and times of arrival and departure associated with these places, and - merging the OD matrices into an OD matrix.