The Link State Content Routing (LSCR) protocol is presented, which supports routing over multiple paths to named content using link-state information. LSCR uses two types of link-state advertisements (LSAs): a Router LSA that contains information about links connected to each router, and an Anchor LSA that carries information regarding a name prefix and the router that advertises being attached to that name prefix, also called an anchor of the prefix. Anchor LSAs are propagated selectively based on a diffusing mechanism. In contrast to prior content routing solutions based on link-state information, LSCR allows routers to establish multiple routes to name prefixes, without requiring each router to know about all the instantiations of each prefix. LSCR is shown to avoid permanent routing loops and to have better performance compared to traditional link-state routing protocols when a name prefix is replicated at multiple sites in the network.

We introduce CCN-RAMP (Routing to Anchors Matching Prefixes), a new approach to content-centric networking. CCN-RAMP offers all the advantages of the Named Data Networking (NDN) and Content-Centric Networking (CCNx) but eliminates the need to either use Pending Interest Tables (PIT) or lookup large Forwarding Information Bases (FIB) listing name prefixes in order to forward Interests. CCN-RAMP uses small forwarding tables listing anonymous sources of Interests and the locations of name prefixes. Such tables are immune to Interest-flooding attacks and are smaller than the FIBs used to list IP address ranges in the Internet. We show that no forwarding loops can occur with CCN-RAMP, and that Interests flow over the same routes that NDN and CCNx would maintain using large FIBs. The results of simulation experiments comparing NDN with CCN-RAMP based on ndnSIM show that CCN-RAMP requires forwarding state that is orders of magnitude smaller than what NDN requires, and attains even better performance.

No abstract

No abstract

Addressing performance degradations in end-to-end congestion control has been one of the most active research areas in the last decade. Active queue management (AQM) aims to improve the overall network throughput, while providing lower delay and reduce packet loss and improving network. The basic idea is to actively trigger packet dropping (or marking provided by explicit congestion notification (