# Introduction Wireless traffic is becoming increasingly important for future communication networks, such as 6G. However, predicting wireless traffic with precision is a challenging task that requires integrating spatial and temporal data. # The TIDES project A research team has developed a new framework called TIDES (Traffic Intelligence with DeepSeek-Enhanced Spatial-temporal prediction) that uses large language models to predict wireless traffic with greater accuracy. The TIDES project integrates the ability to analyze spatial and temporal data to improve prediction accuracy. This is possible thanks to a technique called prompt engineering, which embeds statistical traffic features as structured inputs. In addition, the TIDES project introduces a module called DeepSeek, which allows the LLM to use information from spatially correlated regions. This enables the model to predict wireless traffic with greater accuracy across different areas. # Experiments and results Experiments conducted on real-world cellular traffic datasets have shown that TIDES outperforms baseline models in both prediction accuracy and robustness. Furthermore, the results indicate that integrating spatial awareness into LLM-based predictors is key to unlocking scalable and intelligent network management in future 6G systems. # Conclusion The TIDES project represents an important step towards creating more intelligent and adaptable wireless networks. Its ability to analyze spatial and temporal data enables predicting wireless traffic with greater accuracy, improving network management.