Our current work is focused on designing and implementing the detailed protocols required by TRIAD if it is to be deployed in the Internet, and performing further evaluation and study to support this direction. Specifically, we expect to perform much more comprehensive studies of the large-scale behavior of name-based TCP and routing through simulation. The ability of name rebinding to handle routing topology changes efficiently and the effects of route aggregation need to be clearly demonstrated before TRIAD can achieve wide-scale deployment.
The performance of relay nodes also merits further investigation, especially demonstrating hardware that performs WRAP relaying at the necessary speeds to deploy at ISP boundaries. The effects of opaque tokens on relaying performance and their additional costs (such as potential loss of cachability) are two other performance issues.
However, we have confidence in TRIAD's scalability, since the dynamics of naming and routing are similar to what already exists in the IPv4 Internet. Futher, the incremental deployment of TRIAD gives us the opportunity to tune and refine protocols futher as they are being implemented and deployed in the network.