Browsing by Author "Teng, Hung-Yi"

Now showing 1 - 2 of 2

Fast Failover and Switchover for Link Failures and Congestion in Software Defined Networks
(2016)
Lin, Ying-Dar
;
Teng, Hung-Yi
;
Hsu, Chia-Rong
;
Liao, Chun-Chieh
;
Lai, Yuan-Cheng
In this paper, we present a fast failover mechanism and a fast switchover mechanism to deal with link failure and congestion problems. In the fast failover mechanism, the controller pre-establishes multiple paths for each source-destination pair in the related OpenFlow-enabled (OF) switches. When a link becomes faulty, OF switches are able to failover the affected flows to another path. Based on the pre-established paths, in the fast switchover mechanism, the controller periodically monitors the status of each port of each OF switch. When the average transmission rate of a port consistently exceeds the rate threshold, the controller would decrease the transmission rate of the port by iteratively switching the flow with the minimum rate to another path. The emulation on Ryu controller and Mininet emulator shows the average recovery time of the fast failover mechanism is less than 40 ms, compared to hundreds of ms in the fast restoration mechanism. And, the fast switchover mechanism can reduce 47.5%-72.5% sustained time of link congestion depending on the parameter setting.
Scalable multicasting with multiple shared trees in software defined networking
(2017-01-15)
Lin, Ying-Dar
;
Lai, Yuan-Cheng
;
Teng, Hung-Yi
;
Liao, Chun-Chieh
;
Kao, Yi-Chih
With Software Defined Networking (SDN), IP multicast becomes promising again. For IPTV applications over SDN, existing works would not scale well since they are based on per-source trees. As control-plane in SDN is logically centralized, constructing multiple shared trees is more feasible than that in traditional IP networks. Thus, in this work, we present a locality-aware multicast approach (LAMA) to construct multi-group shared trees in SDN, where each shared tree covers multiple multicast groups. In LAMA, the controller first clusters the multicast sources located in the vicinity into the same multicast cluster. For each multicast cluster, the controller selects the center switch which has the minimum distance to all multicast sources as its rendezvous point (RP) and then constructs a shortest-path multicast tree from the RP to its hosts. Finally, based on the multi-group shared trees, the controller can establish coarse-grained flow entries into on-tree switches to reduce the number of installed flow entries. Emulations on the Ryu controller and the Mininet emulator show that only 2-5 shared trees would suffice. The computation time in the controller using LAMA is around 70 ms, much less than hundreds ms required for per-source trees. Moreover, LAMA only establishes 2300 flow entries, 4% of that with per-source trees in a large topology.