Abstract

Aiming to efficiently support the Locator/Identifier Separation Protocol (LISP), in this paper, we present an enhanced pointer-based DHT mapping system: LISP-PCHORD. The system creates a pointer space to build on top of standard DHTs. Mappings within the pointer space are (Endpoint Identifiers (EID), pointers) where the pointer is the address of the root node (the physical node that stores the mappings) of the corresponding (EID, Routing Locators (RLOCs)) mappings. In addition to enabling architectural qualities such as scalability and reliability, the proposed LISP-PCHORD can copy with flat EIDs such as self-certifying EIDs. The performance of the mapping system plays a key role in LISP; however, DHT-based approaches for LISP seldom consider the mismatch problem that heavily damages the system performance in terms of lookup latency. In order to mitigate the mismatch problem and achieve optimal performance, we propose an optimization design method that seeks an optimal matching relationship between P-nodes (nodes within the pointer space) and the physical nodes on the basis of the given lookup traffic matrix. In order to find the optimal matching relationship, we provide two solutions: a linear programming method and a genetic algorithm. Finally, we evaluate the performance of the proposed scheme and compare it with that of LISP-DHT.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call