Achieving Reliable Flow Management In Hybrid SDN By Resolving Link Failures

The centralized architecture of Software-Defined Networking(SDN)reduces networking complexity and improves network manageability by omitting the need for box-by-box troubleshooting and management.The latest advancements in computing,processing,and networking,have a significant impact on the time-critical smart environment applications,such as smart city applications,smart industry,smart health system,and smart home.The traditional networking used in smart environments faces numerous challenges such as real-time data delivery,programmability,path reliability,and scalability to provide communication among the applications of smart objects,collectively called Internet-of-Things(IoTs).To handle the problems mentioned earlier in a smart environments context,integration of two emerging technologies,namely Fog Computing(FC)and SDN,has been proposed and is gaining momentum.However,due to both budget constraints and maturity level of the SDN-capable devices,organizations often are reluctant to adopt SDN in practice.Therefore,instead of migrating to a pure SDN architecture,an incremental SDN deployment strategy is preferred in practice,an incremental SDN deployment strategy known as hybrid SDN-involving simultaneous use of both SDN switches and legacy switches.The links connected to an SDN switch are called SDN links,and the rest are called legacy links.Provision of reliable Quality of Service(QoS)control,operation,and management,including in the case of communication link failure,in such SDN and hybrid/legacy networks is challenging.This doctoral dissertation focuses on the communication link failure problem and proposes three novel solutions as follows.(1)The first challenge is to minimize the impact of a link failure by designing a reliable communication architecture and using emerging technologies,such as SDN-based FC for time-critical applications in a smart environment.Unlike existing traffic engineering approaches,this dissertation proposes a Reliability-Aware Flow Distribution Algorithm(RAFDA)and two associated optimization algorithms called Reactive Reliability-Aware Heuristic Algorithms(RRAHA-1 and RRAHA-2),which distribute the flows proactively on the links based on the links' reliability levels subject to additional constraints like traffic load on the link,bandwidth allocation,link utilization,and end-to-end delay.(2)The second challenge is predicting reliability level of legacy links in hybrid SDN.In this context,this dissertation proposes a novel approach,called Intelligent Solution for Improved Performance of Reliable and Time-sensitive Flows in Hybrid SDNbased Fog Computing IoT Systems(IHSF).The proposed IHSF approach has three solutions:(?)a novel algorithm to deploy SDN switches between legacy switches to improve network observability.(?)A K-Nearest Neighbor Regression algorithm to predict in real-time the reliability of legacy links at the SDN controller based on historic data;this enables the SDN controller to make timely decisions,improving system performance.(?)A Reliable and Time-sensitive Deep Deterministic Policy Gradient algorithm(RT-DDPG),which optimally computes forwarding paths in hybrid SDN-F for time-critical traffic flows generated by IoT applications.(3)The third challenge is predicting link failures before their occurrence and proactively trying to help.In hybrid SDN,the SDN controller can directly poll the status of the SDN links via the connected SDN switches.At the same time,the legacy links' status information passes through SDN switches and reaches the controller,causing delay.As a result,the controller does not have the current status of legacy links in real-time.This delay may lead to undesired outcomes.For example,it causes network reachability problems due to Access Control List(ACL)policies.Therefore,to minimize the impact of network-layer failure in hybrid SDN,this dissertation proposes a Machine Learning(ML)based technique called PrePass-Flow.The main objective of PrePass-Flow is to minimize the ACL policy violations and network reachability problems due to ACL policies in case of link failures.For the link status prediction,PrePass-Flow uses two supervised ML-based models:1)a Logistic Regression(LR)model,and 2)a Support Vector Machine(SVM)model and outperforms existing approaches in terms of Packet Delivery Ratio(PDR)and ACL policy violations.