Application for radio signal coverage calculation
The PieSoft software architects and developers contributed significantly to the Customer's project - an application for a radio signal coverage calculation - by switching it to the distributed system.
Background
The major Customer’s project was a desktop application development for a radio signal coverage calculation for a building/area/room (5G, Wi-Fi).
To work with the application, the user had to build the room plan indicating wall material and place antennas, indicating the model for a particular room. The application calculated the radio signal, considering the room for each particular antenna. Then it built a common prediction for the radio signal room coverage. It allowed choosing the modes for one antenna coverage or all of them in total. All indications were fixed by graphical visualization.
Challenge
The Сustomer took an order for enormous areas and buildings coverage in Japan. The calculations took a long period of time (a few hours). The scope of area coverage reached several kilometers.
The PieSoft team’s main task was to minimize the calculation time by switching to the distributed system, which resulted in two software development phases:
- Implementing the distributed system for a single antenna signal coverage calculation and for the signal coverage prediction of the total number of antennas.
- Developing a web application for its interaction with a distributed computing system.
Solution
On the Customer’s side, micro applications for GPU Computing with OpenCL were developed. They also implemented models of buildings and streets based on photos. PieSoft contribution:
01
Migrated the calculation on Amazon Web Services (Microsoft Azure was potentially considered)
02
Created a web application using Spring Framework for the major application and distributed system interaction
03
Implemented the distributed system using Spark Framework
Technologies used
Languages & frameworks
- Java
- Spring WebFlux
Databases
- Hadoop HDFS
- Amazon S3
Architecture
- Containers
Tools
- Docker
- Azure DevOps
Result
PieSoft team contributed significantly to
the Customer’s project, preparing the existing application for the distributed system. We built software architecture, broke down the user’s primary task into subtasks, which can be calculated on computing units, described
the distributed system’s principles, and optimized big data environments with Hadoop and Spark Frameworks so that the solution could be quickly delivered and run smoothly.
Do You Have a Similar Project Idea?
No matter how complex your IT project goals are, our professional business analysts will increase the product quality while reducing its costs.
OUR HEADQUARTERS
We are open to new challenging tasks and we'd love to learn more about your project.
