Can't connect to Wi-Fi? Tsinghua University has developed the "ultimate solution"

While Wi-Fi is one of the greatest inventions of the last century (can anyone dispute that?), it's also one of the most anxiety-inducing technologies.

[[182460]]

Can't connect to Wi-Fi? You're not alone. After 400 million attempts to connect, a research team from Tsinghua University says they've found the cause of the failure and have a solution.

First, let me introduce the steps of Wi-Fi connection:

• In the first step, your device will scan the available Wi-Fi access nodes around you;
• Once a Wi-Fi node is selected, your device will exchange data packets with the corresponding router;
• Next comes the authentication process, which usually requires entering a password;
• In the next step, according to the Dynamic Host Configuration Protocol, the server will assign an IP address to your device.

The team used an Android software called Wi-Fi Manager, which can record the time taken for each step of connecting to Wi-Fi.

They said that according to experimental results, the quality and performance of current Wi-Fi networks are far from satisfactory: 45% of the time, connections fail.

In the experiment, the time it took to connect to Wi-Fi varied greatly: 15% of the connections took more than 5 seconds, and 5% took more than 10 seconds.

There are many indicators used to judge the quality of Wi-Fi networks, including throughput, latency, etc. However, the probability of successfully connecting to a Wi-Fi node and the time it takes to set up the connection are the important factors affecting the user experience.

The researchers also said that although signal strength has a great impact on the connection process, only by understanding how the router works can the problem be truly solved.

"We found that while signal strength is important, knowing the patterns of Wi-Fi access points and mobile devices can be a good predictor of how long it will take to establish a connection."

Whether the Wi-Fi network was public or private was another important factor. They found that private networks were significantly faster and had a higher success rate in connecting.

In addition, the team proposed a way to use AI to improve the success rate of connection: automatic scanning through AI algorithms to find the best network. "Through comprehensive measurements and detailed analysis, we proposed a Wi-Fi access node selection algorithm based on machine learning."

“This algorithm uses the characteristics of Wi-Fi access points as input to the machine learning model, and then classifies the candidate access points into two categories: SLOW and FAST. Based on the classification results, our algorithm can avoid mobile devices from connecting to access points that are classified as SLOW.”

“Evaluation results show that compared with an algorithm that uses signal strength alone to select access nodes, our algorithm reduces the connection failure rate from 33% to 3.6%, and the connection setup time for 80% of the connections can be reduced by 10 times.”