Achieving Near-Optimal Traffic Engineering Using a Distributed Algorithm in Hybrid SDN

Abstract

To empower advanced traffic engineering (TE) mechanism, while considering the infeasibility of one-step migration to software-defined networking (SDN), SDN nodes are incrementally deployed into legacy network, which gives rise to hybrid SDN. In hybrid SDN, redirecting flow of every source-destination pair through at least one SDN node, can enhance TE performance and obtain flow manageability, while on the other hand leading to increasing demands of TCAM resources in SDN nodes. In this paper, we make minimization of maximum link utilization as the TE objective, and comply with SDN waypoint enforcement and TCAM resource limitation. We first formulate the TE problem as an integer linear programming (ILP) model and solve it in a centralized manner, where SDN waypoint selection and splitting fractions for each flow are jointly determined. Then, based on a fact that the logically centralized control plane in hybrid SDN is composed of multiple physically decentralized controllers, each of which manages part of SDN nodes, as well as considering a real situation that a centralized solution is infeasible or too fragile for large-scale network, we develop a distributed algorithm deriving from Lagrangian decomposition theory to effectively solve the TE problem. The simulation results indicate that, when 30% of the SDN nodes are deployed, the proposed traffic engineering-aware distributed routing (TEDR) algorithm obtains maximum link utilization comparable to that of full SDN, and has a limited influence on the routing efficiency. INDEX TERMS Software-defined networking, hybrid SDN, traffic engineering, resource allocation, distributed algorithm.

Flow-Aware Forwarding in SDN Datacenters Using a Knapsack-PSO-Based Solution

“…Ren et al [1] propose a TE method through integer linear programming formulation. This method redirects each flow to the proper node(s) and then splits them into multiple paths.…”

“…Therefore, due to the necessity of acquiring a global view of the network, Software Defined Networking (SDN) has emerged as a widely adopted paradigm in DCNs. Decoupling the control plane and data plane is a fundamental principle in SDN architecture design [1]- [3]. A multitude of researchers and industries intend to achieve better performance parameters through SDN's specific characteristics, such as network programmability, open interfaces, centralized administration, and abstraction, all of which, in turn, lead to agility and flexibility [4]- [7].…”

Comparison of Software Defined Networking with Traditional Networking

“…The study [83] presented a secure and systemic SDN framework capable of avoiding spoofing attacks and DoS with a common SDN router configuration overhead. The reference [84] proposed TEDR algorithms that can achieve optimum connection use if the SDN nodes are deployed as 30 percent as complete SDN and have a small effect on routing efficiency. The research [85] explored that using SD Guardian can reduce table-miss special attacks.…”

Hybrid SDN evolution: A comprehensive survey of the state-of-the-art

“…The switches can be configured using a centralized controller to monitor and push configurations. The model is able to accomplish conflict-free separation of centralized [186] Identify flows that reduce Flow spread based algorithm estimation error most Cheng et al [185] Compressive traffic monitoring Collect load of small subset of links, estimate the link load of the rest Zhang et al [187] Selection of critical flows Linear programming approach for traffic matrix using ML Tian et al [188] Link latency measurement Minimize number of monitors Traffic management Casado et al [189] Vendor-neutral network model Use MPLS for hSDN Tu et al [190] Mitigate congestion Tunnel splicing between MPLS and SDN Sugam et al [191] Improve network utilization Fully Polynomial time approximation schemes He et al [192] Optimize TE performance Provable approximation guarantees Ren et al [193] Distributed algorithm for TE Integer Linear Programming formulation Guo et al [194] Minimize maximum link utilization Change weights and flow-splitting ratio Nakahado et al [195] Decrease network congestion Edge nodes replaced with SDN switches Sharma et. al.…”

“…Ren et al [193] propose a distributed algorithm for near optimal TE. In hSDN, redirecting flows from every source-destination pair through at least one SDN node, can improve TE performance and obtain flow manageability, while on the other hand leading to increasing demands of TCAM resources in SDN nodes.…”