Other Protocols. The predecessor to IKE, Photuris [Xxxx and Xxxxxxx 1999], first introduced the concept of cookies to counter “blind” DoS attacks. The protocol itself is a six-message variation of the StS protocol. It is similar to IKE in the message layout and purpose, except that the SA information has been moved to the third message. For rekeying, a two-message exchange can be used to request a uni- directional SPI (thus, to completely rekey, four messages are needed). Photuris is vulnerable to the same computation-based DoS attack as IKE, mentioned above. Nonetheless, one of the variants of this protocol has four messages and provided DoS protection via stateless cookies. SKEME [Xxxxxxxx 1996] shares many of the requirements for JFK, and many aspects of its design were adopted in IKE. It serves more as a set of protocol building blocks, rather than a specific protocol instance. Depending on the specific requirements for the key-management protocol, these blocks could be combined in several ways. An interesting aspect of SKEME is its avoidance of digital signatures; public-key encryption is used instead, to provide authen- tication assurances. The reason behind this was to allow both parties of the protocol to be able to repudiate the exchange. SKIP [Xxxx and Xxxxxxxxx 1995] was an early proposal for an IPsec key- management mechanism. It uses long-term DH public keys to derive long-term shared keys between parties, which is used to distribute session keys between the two parties. The distribution of the session key occurs in-band, that is, the session key is encrypted with the long-term key and is injected in the encrypted packet header. While this scheme has good synchronization properties in terms of rekeying, the base version lacks any provision for PFS. It was later provided via an extension [Xxxx 1996]. However, as the authors admit, this extension detracts from the original properties of SKIP. Furthermore, there is no identity protection provided, since the certificates used to verify the DH public keys are (by design) publicly available, and the source/destination master identities are contained in each packet (so that a receiver can retrieve the sender’s DH certificate). The latter can be used to mount a DoS attack on a receiver, by forcing them to retrieve and verify a DH certificate, and then compute the DH shared secret. The host identity payload (HIP) [Xxxxxxxxx 2001] uses cryptographic public keys as the host identifiers, and introduces a set of protocols ...
Other Protocols. (Insert additional protocols below.) « Nothing in the foregoing Section 2.4.2 shall relieve the Architect of the full breadth of its professional obligations and duties with respect to the Project and the Model. »
Other Protocols. LNG, SMEDIS (JRC/HSL) The development of the HYMEP was based also on previous experience and projects that were performed for technologies and applications outside of the hydrogen and fuel cell safety sector. A Model Evaluation Group was established by the European Community in the early 1990s (XXXXX 1996). The group was set up to develop methods for the evaluation of models in major industrial hazards areas as it had become apparent that models used in industrial hazard assessment had never been formally assessed for their fidelity. Nevertheless those models were used as the basis for decisions that directly affected public safety and the environment. In 1994 the group published guidance on model evaluation protocols (XXXXX 1996) which provides a framework for the key activities needed to evaluate models. Testing the results of model predictions against experimental data, Kakko et al. (Kakko et al. 1994) highlighted the need for suitable experimentally derived databases for model validation, as these were often difficult to obtain or not presented in a way suitable for model validation. A classic example of a model validation database is the modeller’s data archive (MDA) of Xxxxx et al. (Hanna, Strimaitis, and Xxxxx 1989) who identified the need to collate data in a form that could be accessed by model developers. Since then, some other datasets have been produced, such as the Rediphem database (Xxxxxxx and Xxx 1996). Xxxxx and Xxxxx also developed a model performance evaluation methodology for gas dispersion models and for air quality models (Xxxxx and Xxxxx 0000), (Xxxxx and Xxxxx 2004). The SMEDIS (Scientific Model Evaluation of Dense Gas Dispersion Models) project (Daish et al. 2000), (Xxxxxxxxx et al. 2001) brought together the concept of a model evaluation protocol and specialised database. Its main aim was to provide a methodology not only for validation but also a scientific review of models. As well as evaluating simple scenarios, the project focused on situations in which complex effects such as aerosols, topography and obstacles were important. Ivings et al. (Ivings et al. 2007) set out a Model Evaluation Protocol for models used to predict the dispersion of vapours from Liquefied Natural Gas (LNG) installations. The protocol is based upon the SMEDIS project but is not confined to the modelling of LNG spills as other, simpler cases could also be taken into account in model evaluation. One of the recommendations of the MEP (Ivings et al. 2007),...
