Last week ARCFIRE researchers were present at the SDN & NFV World Congress 2017, in the context of the ETSI NGP Forum, one of the sessions within the conference. ARCFIRE researchers introduced RINA and discussed a number of research results achieved within the ARCFIRE and PRISTINE projects, in the areas of congestion control, dynamic renumbering, security, network management, routing and forwarding for large-scale datacentres and distributed mobility management.
A number of questions were asked by the audience, such as:
- (Q) What is the address format of RINA networks? (A) Each layer can decide on what is the syntax (length of source/destination address fields in data transfer packets) and the semantics (how the address byte string is to be interpreted, e.g. addresses may be hierarchical, etc.)
- (Q) How fast can handover and routing updates be in the case of mobility? (A) This ultimately depends on the RINA-based products to be deployed in the network. RINA provides a flexible framework in which i) there is no need to setup/tear-down tunnels in mobility events and ii) scope of routing updates (the number of systems that have to be aware of the routing update) can be bounded by clustering mobility regions into layers.
- (Q) How does RINA address end-to-end security? (A) Trust relationships between layers (including applications) are well-defined by the architecture. If an application (including a DIF) doesn’t trust the DIF below, it should protect its traffic (encryption, authentication codes, etc.) before sending it down. Security policies are defined per layer (not per protocol), making the overall security cost smaller compared to current networks.
- (Q) This looks still very academic, are there RINA products in the market already? (A) Not yet, but thanks to the experimental research work carried out by projects like ARCFIRE, the risk of developing and adopting RINA technology commercially is going down, therefore we should not be too far from the initial RINA-based products.