Adoption barriers of network layer protocols: The case of host identity protocol

摘要

With increasing societal dependence on the Internet and new application areas emerging, the need for securing communications and identifying communication partners is expected to increase. However, the original Internet architecture is lacking these functionalities, and most of the protocols proposed to fix these issues have not been widely deployed. Often one of the reasons for such failure is that protocol designers have insufficient understanding of the potential adopters' economic incentives so one may end up designing protocols based on false or inaccurate assumptions. In this paper, we analyze the Host Identity Protocol (HIP) from this viewpoint Based on 19 expert interviews, we identify six main reasons why HIP has not been widely deployed yet. Most importantly, (1) the demand for the functionalities of HIP has been low. Where demand would have existed, substitute solutions have been favored because (2) they were earlier on the market, (3) they have relative advantage due to some design choices of HIP, (4) HIP lacks early adopter benefits necessitating costly coordination among multiple stakeholders in public deployment scenarios, and (5) people have misconceptions about the deployability of HIP. Additionally, (6) the research-mindedness of HIP developers has lead to strategic mistakes and non-optimal design choices from the perspective of deployment. We also suggest strategies that HIP developers could take to foster the adoption of HIP. Besides providing value to HIP developers, the results propose some new adoption barriers and deployment strategies that could be taken into account when designing new protocols. Finally, the article also provides a template that could be followed when studying the feasibility of other protocols.