Routers used in home networks play a significant role in securing devices from external threats by monitoring network activity and enforcing policies. Almost all security features of these devices rely on some network monitoring solution, where the performance of the monitoring greatly affects the whole system's effectiveness. This drives security providers to explore new, possibly more performant solutions for inspecting network packages.

A possible approach is to offload part of the parsing logic into the kernel and handle only the relevant packets in the userspace. This is typically done by loading classic BPF programs into the kernel, but on up-to-date systems it got replaced by eBPF. This technology is typically utilized in cloud infrastructures and using it to secure home networks is still an unexplored use case. The task is to evaluate the possibility of using eBPF on home routers and explore the costs and benefits of migrating from other solutions.  
  
The task of the student includes:  
- Explore the possible approaches to building eBPF programs for typical home routers.  
- Investigate real-world algorithms that could benefit from sharing state with the kernel via eBPF.  
- Implement network monitoring prototypes with and without eBPF.  
- Compare the performance and complexity trade-offs of the different solutions.  
- Assess the limitations of supporting severely outdated kernels and the potential benefits of targeting newer versions.