Computer science researchers have found a new way to make wireless networks faster and more reliable.
This method uses a technique called a “digital twin” to predict the data needs of wireless users before they even know they need it.
The focus here is on something called edge caching. Edge caching involves storing data on a nearby server that a network thinks users will likely use soon.
This way, the system can quickly meet user demands without having to retrieve data from the original source.
Edge caching usually happens on servers that are closest to the user, like those built into network routers.
The challenge with edge caching is figuring out which data to store and how much to store at any given time.
According to Yuchen Liu, an assistant professor of computer science at North Carolina State University, “Systems can’t store everything in edge caches, and storing too much can slow down the server if it uses too many resources. So, systems must constantly decide which data to keep and which to remove.”
The key to better performance is accurately predicting what data users will want and how much the edge servers should store.
Liu and his team developed a new method called D-REC to improve these predictions. D-REC uses a digital twin, a virtual model of the real wireless network, to help with this.
“A digital twin is a virtual copy of a real object. In the case of D-REC, it’s a virtual model of a wireless network, whether it’s a cellular network or a Wi-Fi network,” Liu explains.
The digital twin takes real-time data from the wireless network and uses it to run simulations that predict which data users will likely request. These predictions then help inform the network’s edge caching decisions. Because these simulations happen outside the network, they don’t slow it down.
Researchers tested D-REC using open-source datasets to see if it made wireless networks more efficient. They ran many experiments considering different variables like network size and the number of users. “D-REC outperformed conventional approaches,” Liu says.
“Our technique improved the network’s ability to accurately predict which data to cache and helped balance data storage across the network.”
Additionally, D-REC’s digital twin can foresee potential problems. For example, if it predicts that a specific server will be overloaded, the network can be alerted to redistribute data and maintain performance and reliability.
“We’re open to working with network operators to explore how D-REC can improve network performance in real-world situations,” Liu adds.
The paper, “Digital Twin-Assisted Data-Driven Optimization for Reliable Edge Caching in Wireless Networks,” is published in the IEEE Journal on Selected Areas in Communications.
