Jul 2026· Applied Networking Research Workshop· pp. 75-81· 0 citations· 33 references
Computer Science
TL;DR
This paper surveys multicast forwarding-state capacity across ISP core routers and enterprise routers, and proposes a capacity-aware path-selection algorithm for PIM-SSM based on Shortest Widest Path (SWP), allowing to load-balance multicast joins over paths with remaining multicast hardware entries.
Abstract
IP multicast was designed in an era of software-based routers, when forwarding state lived in expandable RAM. Today, multicast relies on hardware for efficient replication of packets across interfaces, which has hard, fixed limits on the number of multicast forwarding entries it may hold. We argue that the limited available multicast state calls for a redefinition of the multicast protocol stack. In this paper, we survey multicast forwarding-state capacity across ISP core routers and enterprise routers to highlight their hardware limits. We then propose to extend existing routing protocols (OSPF and IS-IS) to expose the hardware replication capacities of nodes to the control plane. Building on these exposed capacities, we propose a capacity-aware path-selection algorithm for PIM-SSM based on Shortest Widest Path (SWP), allowing to load-balance multicast joins over paths with remaining multicast hardware entries. Our results show that, by exposing multicast capacities, we can not only improve the join acceptance rate by up to 22%, but our solution also allows to reach a two times higher join rate without rejecting any join. By increasing the join acceptance rate, more concurrent multicast flows can coexist in networks.
Network routing scalability becomes difficult when forwarding state scales with an external identifier space rather than network topology. Named Data Networking (NDN) faces this challenge acutely because routing directly on application name prefixes ties forwarding state to an unbounded namespace. This paper presents SRAN, an intra-domain NDN routing architecture that applies the routing-scalability principle underlying Map-and-Encap through Map-and-Attach: application prefixes are mapped to egress routers, and the resulting mapping information is attached to the original Interest. This enables network routing and forwarding to operate on topological identifiers while keeping NDN Interests intact and preserving native NDN communication semantics, including Interest/Data exchange, in-network caching, and data-centric security. SRAN further uses the same prefix-to-egress mapping to realize Bit Index Explicit Replication (BIER) for scalable NDN Interest multicast. SRAN leverages existing NDN mechanisms to securely maintain prefix-to-router mappings among user-facing routers without introducing new protocols. Evaluation on Rocketfuel topologies shows that routers'forwarding state scales with network topology rather than application-prefix count, while prefix updates are disseminated in real time with low communication overhead.
Multiprotocol Label Switching (MPLS) plays a critical role in the backbone networks of Internet Service Providers (ISPs), ensuring robust and scalable network operation. By employing label-based switching instead of destination-based IP forwarding, MPLS significantly reduces forwarding complexity and supports advanced traffic engineering mechanisms. Within MPLS, the Label Distribution Protocol (LDP) is typically used for label distribution, relying on routing information from the Interior Gateway Protocol (IGP). When a link or node failure occurs, LDP is forced to wait for the IGP to re-calculate and update the new paths before it can assign labels for Forwarding Equivalence Classes (FECs). This dependency significantly increases service disruption time due to the combined IGP convergence time and the subsequent label replacement mechanism (withdrawing old labels and installing new ones). Therefore, this paper proposes a novel scheme to mitigate LDP’s dependence on the IGP, thereby minimizing network restoration time during network topology changes. The method utilizes an event-driven signaling mechanism to instantly announce link and node outages across the network. Furthermore, it establishes and installs both primary and backup Label Switched Paths (LSPs) during the initial setup phase of the router. Experimental results demonstrate that the proposed method significantly outperforms LDP, achieving a convergence time up to four times faster in large-scale network scenarios and proving superior scalability.
Trung Van Vu, La Van Thien, Quyet Hoang Dinh et al.· International Conference on...· 0 citations
Ensuring security without compromising low-latency determinism is critical for modern time-sensitive (TS) applications. However, existing authentication mechanisms like Timed Efficient Stream Loss-Tolerant Authentication (TESLA) suitable for local-area real-time networks lack scalability for wide-area environments. We propose Det-TESLA, which integrates deterministic IP (DIP) networking with the TESLA protocol to ensure stable key disclosure latency. By decoupling multicast TS flows from key distribution (KD) flows, the framework provides superior scheduling flexibility for complex wide-area networks. We further develop a joint TS-KD scheduling model and an efficient online algorithm based on combinatorial auctions. Simulations reveal that Det-TESLA supports 2.9 times more low-latency flows than standard TESLA. Furthermore, our algorithm achieves an 11% performance improvement with only 20% of the execution time required by baseline algorithms.
Wei-Peng Tan, Binwei Wu, Shuo Wang et al.· International Conference on...· 0 citations
Collective communications (CC) and distributed artificial intelligence (AI) training workloads involve substantial one-to-many chunked data transfers. Enabling optical multicast in next-generation optical circuit switching (OCS)-based data centers could streamline such traffic and improve CC efficiency. This paper presents our investigation of when and how to exploit optical multicast for CC and AI training. We discuss the topological property of OCS with multicast capability, showing reduced port-count overhead for the same connectivity degree as its point-to-point (P2P) counterpart. Further, by formulating and addressing the grouped-multicast CC schedule optimization (GMS), we study and compare the completion time of two key types of CC operations in GMS against the P2P schedule under identical port use conditions. Simulations with representative topologies reveal the potential of multicast OCS to enhance CC via a co-design of multicast-based topology and scheduling.
Lihua Ruan, Linbojie Huang, Longquan Dai et al.· International Conference on...· 0 citations
Periplus, an in-band SDN control plane designed to address four challenges that this approach presents in such contexts: automatic bootstrapping, source-based routing, fast failure recovery, and multi-controller coordination is presented.
E. M. C. Barbero, P. Quirós, F. J. S. Reigadas· 1 citation
This paper presents detailed algorithm for calculating L-LSR coefficient, and shows that L-LSR algorithm not only performs better than OSPF, but also has verySignificant performance improvement over the other LSR family of algorithms.