Online Multicast Traffic Engineering for Software-Defined Networks

TL;DR

This work tackles online multicast traffic engineering for SDNs under dynamic group membership, formulating the Online Branch-aware Steiner Tree (OBST) problem to jointly minimize tree cost, branching, and rerouting across time. It introduces the Online Branch-aware Steiner Tree Algorithm (OBSTA), which uses budget, deposit, and a Reference Tree to construct a sequence of correlated multicast trees with provable $|D_{max}|$-competitive guarantees. The paper provides hardness results showing OBST is challenging and cannot be approximated beyond certain bounds, then presents OBSTA with a four-phase design (RT construction, DT generation, DT patching, and DT selection) and theoretical analysis supporting its competitiveness. Empirical evaluations on real and large synthetic networks, including a live OpenDaylight implementation with YouTube traffic, demonstrate that OBSTA reduces total cost by about 25% relative to SPT and ST while keeping rerouting costs low and enabling scalable multicast in SDN environments.

Abstract

Previous research on SDN traffic engineering mostly focuses on static traffic, whereas dynamic traffic, though more practical, has drawn much less attention. Especially, online SDN multicast that supports IETF dynamic group membership (i.e., any user can join or leave at any time) has not been explored. Different from traditional shortest-path trees (SPT) and graph theoretical Steiner trees (ST), which concentrate on routing one tree at any instant, online SDN multicast traffic engineering is more challenging because it needs to support dynamic group membership and optimize a sequence of correlated trees without the knowledge of future join and leave, whereas the scalability of SDN due to limited TCAM is also crucial. In this paper, therefore, we formulate a new optimization problem, named Online Branch-aware Steiner Tree (OBST), to jointly consider the bandwidth consumption, SDN multicast scalability, and rerouting overhead. We prove that OBST is NP-hard and does not have a $|D_{max}|^{1-ε}$-competitive algorithm for any $ε>0$, where $|D_{max}|$ is the largest group size at any time. We design a $|D_{max}|$-competitive algorithm equipped with the notion of the budget, the deposit, and Reference Tree to achieve the tightest bound. The simulations and implementation on real SDNs with YouTube traffic manifest that the total cost can be reduced by at least 25% compared with SPT and ST, and the computation time is small for massive SDN.

Figures (9)

• Figure 1: (a) $T_{i-1}$ for $d_1$ and $d_2$. (b) $T_{i-1}\ominus \{d_1\}$. (c) $T_i\ominus \{d_3\}$. (d) $T_{i}$ for $d_2$ and $d_3$. (e) Symmetric difference of $T_{i-1}\ominus \{d_1\}$ and $T_i\ominus \{d_3\}$.
• Figure 2: (a) Input network $G$. (b) $T_1$ of SPT, ST, and OBST. (c) $T_2$ of SPT. (d) $T_2$ of ST. (e) $T_2$ of OBST. (f) Alternative $T_2$ of OBST.
• Figure 3: Example of sprouting $A\otimes (a_{1},a_{2},a_{3})$.
• Figure 4: Example of Candidate DT Generation. (a) $T_{i-1}$. (b) $T_{i,leave}$. (c) $T_{i}^{3}$.
• Figure 5: Example of grafting $A\oplus B$.

Theorems & Definitions (23)

• Definition 1
• Definition 2
• Definition 3
• Example 1
• Definition 4
• Definition 5
• Theorem 1
• proof
• Corollary 1
• Definition 6