Search

One of the critical targets of the fifth generation (5G) mobile network is supporting low latency and high reliable services at the network edge servers. Multiple service instances of the same service can be deployed on different geographically distributed edge sites for high availability purposes. However, the dynamic changes over time in edge servers’ computing capabilities and underlay infrastructure network quality might affect service performance. To avoid service performance degradation, an efficient computing-aware traffic steering (CATS) solution between different service instance locations is required. Multiple advanced traffic steering algorithms have been proposed in different existing 5G traffic steering works. However, no available works addressed the possible 5G architecture and procedure changes to convert these CATS algorithms’ decisions into 5G mobile user plane routing paths. This work aims to cover this research gap. We propose enhancing the current 5G architecture by applying the Mobile User Plane Controller (MUP-C) concept and using anycast address for multi-instance services. We discuss three implementation options to apply these enhancement features to 5G architecture: CATS Application Function (AF) traffic influence method, CATS MUP-C dummy User Plane Function (UPF) method, and CATS MUP-C non-UPF method. We implemented and evaluated the performance of our solutions against a 3GPP standard dynamic edge discovery method, which can also be used to support CATS. The results showed that our solutions had better Protocol Data Unit (PDU) session setup procedure efficiency for CATS than the standard method by reducing the setup latency from 30% to 50%.

This paper proposes a new load-balancing method that supports Per-Connection Consistency (PCC) by leveraging ECMP with standardized routing protocols for operational simplicity. Load balancers must consistently deliver packets of a flow to one of the backend servers, even when the backend server pool changes. Achieving this property—PCC—is not a straightforward task and usually introduces additional complexity into networks, such as special packet forwarding mechanisms, synchronizing states across load balancers, and often complicated orchestration. Our approach, VRF-shadowing, achieves PCC on ECMP of routers without additional control plane systems by exploiting Segment Routing over IPv6 (SRv6). VRF-shadowing employs the daisy-chaining concept, achieving PCC in a stateless manner with a standardized SRv6-based Layer-3 VPN (L3VPN). VRF-shadowing is composed of two components: a minimal new feature for software of SRv6 routers and a new SRv6 End behavior at servers. We implemented the former in a commercial router and the latter in Linux and evaluated VRF-shadowing on an SRv6 L3VPN network built with only standardized routing protocols. The evaluation results demonstrate that VRF-shadowing preserves connections from being disrupted when the ECMP next-hops change. In the most severe scenario, VRF-shadowing reduced the failed HTTP requests rate from 77% to 0.7% compared to traditional ECMP.

Nowadays, ocean networks gradually become increasingly important for communication among network entities such as maritime and sea-crossing users. However, ocean networks are highly dynamic because the communication links are composed of satellite and microwave links, which could be easily influenced by the environment such as local climate. Thus, network transmission in ocean networks faces great challenges, including low reliability and low efficiency. In this paper, we propose a smart ocean network architecture, where we use Software Defined Network (SDN) to perform unified management of the network, and Segment Routing (SR) to control data forwarding paths. In this way, we can control network flows and optimize network routing among diverse network entities in an ocean network. However, many Quality of Service (QoS) guaranteed applications in ocean networks, such as remote control, require lower delay. To guarantee the performance for such applications, we further propose QoS routing algorithms based on Fuzzy-Lagrange for the smart ocean networks architecture, where the optimization objective is to ensure service quality provided to users. According to experimental results, it is proved that, in comparison with the benchmark algorithms, the Fuzzy-Lagrange (FuzLag) algorithm proposed based on link fuzzification and Lagrangian Relaxation can improve the performance by 23% at most.

Satellite networks, which have wide coverage and high throughput, are more and more important for the Internet. Low earth orbit (LEO) satellites have been more widely used than other orbiting satellites because they have lower unidirectional links (UDL) transmission delay. However, due to the variability and instability of satellite links, leading to relatively higher UDL transmission delay and more link failures for satellite networks. Thus, fast reroute (FRR) schemes, which can bypass the failure links and improve network performance, are badly needed in satellite networks. However, FRR need relatively large computing resources while satellites have limited resources. Thus, traditional routing technologies cannot be directly applied to satellite network due to frequent topology changes and poor computing performance. To meet the challenges, we propose a hybrid FRR schemes for satellite networks that combines the centralized computing with distributed segment routing (SR). With the scheme, satellites with relatively more computing resources can pre-compute the reroute paths, and distribute the routing rules to satellites that have more forwarding resources. We formulate the problem, propose a classification algorithm that can distinguish satellites according to their computing and forwarding resources. Additionally, we also propose a real-time backup path maintenance algorithm in the hybrid rerouting scheme. Finally, we conduct comprehensive simulations to evaluate the performance of the proposed algorithms, and the results show that under extreme conditions, the proposed LFR algorithm is 30.9% better than traditional algorithms in storage resource utilization; and the proposed LFA+ algorithm is 74.3% better than traditional algorithms in update time comparison.

It is still a challenging task to provide satisfactory quality of experience (QoE) to users by traditional routing mechanism in current public Internet for real-time multimedia services with high bandwidth and strict end-to-end latency constraint. Multipath transmission is a promising way to improve the performance of data delivery, due to its advantage in bandwidth aggregation capacity over multiple paths to speed up high volume transfers between endpoints. However, existing multipath transmission has many disadvantages such as the requirements for end-host support (i.e. the necessary multi-homed network setting, and operating system or upper applications updates of user devices), and blindingly pursuing maximum of transmission throughput. Path allocation and end-to-end transmission control are two important factors affecting the quality of multipath transmission. As an emerging network architecture, software-defined networking (SDN) is flexible, manageable, and responsive to rapid changes in traffic requirements. In this paper, we propose a multipath transport scheme based on SDN with segment routing (SR) which can meet the bandwidth and low end-to-end transmission latency requirements of real-time interactive multimedia services. The centralized SDN controller is extended to allocate multiple disjoint paths meeting bandwidth requirement meanwhile balancing the network load. The proprietary multipath media server (MMS) dynamically controls the assignment between subflows and routing paths to reduce the end-to-end transmission latency and thus improve QoE of end-users. To show the feasibility of our approach, we construct an SDN platform using Mininet. Simulation results show that our approach performs better compared to the conventional approaches in terms of traffic balancing and end-to-end transmission delay.

The unprecedented growth of network traffic has brought excessive challenges to network operators. To prevent network congestion, network operators conduct traffic engineering (TE) for their routing optimization. In recent years, segment routing traffic engineering (SRTE) has emerged as one of the promising approaches for its high scalability and low control overheads. However, conventional SRTE approaches in large-scale networks are computationally prohibitive, which may lead to delayed system operations and unsatisfactory service qualities. In this paper, we formulate a bi-objective mixed-integer nonlinear program (BOMINLP) to investigate the trade-off between link utilization and computation time in SRTE. Due to the difficulty in solving the original problem directly, we decompose it into two sequential sub-problems. The first sub-problem is to minimize computation time through node selection, and the second one is to minimize maximum link utilization via flow assignment. To this end, we first employ randomized sampling based on stretch bounding to obtain a reduced solution space and then solve a linear program (LP) using existing software tools for the sub-problems. To evaluate our proposed solution, we employ network topologies and traffic matrices from publicly available datasets. Our simulation results show that our proposed solution can effectively reduce computation time while retaining comparable maximum link utilization as compared with several comparison approaches.

With the ubiquitous use of networking technologies in the military domain, network-centric warfare, which aims at improving the system of systems (SoS) capability, is a no-way-back trend of military development. Recently, it has become increasingly important to construct an airborne backbone network (ABN) to maintain coverage and provide reach-back to military units because the terrestrial networks cannot be flexibly deployed and have poor survivability. Given the advantages of software-defined networking (SDN), a software-defined airborne backbone network architecture (SD-ABN) is proposed for the ABN to ensure network flexibility, openness, interoperability, and evolvability. To meet the challenges of traffic management in SD-ABN, segment routing (SR) is applied with some practical modifications for the SD-ABN. Moreover, a network traffic scheduling algorithm termed MRP-TS is designed based on SD-ABN to improve the transmission reliability and bandwidth utilization by balancing network traffic to multiple reliable transmission paths. Relevant theories have been proposed and proved to illustrate the correctness of the algorithm. The network simulation experiments have been conducted using the network simulator EXata 5.1, and the simulation results validate the effectiveness of our proposals.

To tackle the network congestion problem caused by ground gateway stations arranged within a limited area in low earth orbit (LEO) satellite networks, a routing algorithm based on segment routing for traffic return is proposed. Light and heavy load zones are dynamically divided according to the relative position relationship between gateways and the reverse slot. The pre-balancing shortest path algorithm is used in the light load zone, and the minimum weight path defined by congestion index is the routing rule in the heavy load zone. Then, the consistent forwarding is performed referring to segment routing in all zones. Simulation conditions are different sizes of heavy load zone, different traffic density distributions, and different traffic demands. Simulation results confirm that the load-balancing performance is improved significantly with the extension of the heavy load zone size in terms of the average rejection ratio, the average relative throughput, the maximum link utilization, and the average delay. The proposed algorithm is an alternative solution and guidance for routing strategy in LEO satellite networks.

Segment routing (SR) is a new network paradigm to optimize network performance. Through leveraging source routing, SR is able to achieve more fine-grained control of data flow in the SR domain. However, it is difficult to introduce large-scale full SR network into existing traditional network due to the economic constraints and immature operation technology. Thus, incrementally deploying SR nodes into an existing network is preferred, which forms the hybrid IP/SR scenario. In the hybrid network, with SR enabled devices, the routing path of flow can be dynamically adjusted. Network utility is an important factor to reflect the user's satisfaction with the allocated bandwidth. In this paper, we propose a bandwidth allocation algorithm to maximize network utility in hybrid IP/SR network and thus improve customer's satisfaction. The simulation results show that the bandwidth allocation algorithm proposed is able to improve the utility of network significantly, and the utility of network will increase with the number of SR nodes deployed in the network.

Segment routing is an emerging routing technology that was initially driven by commercial vendors to achieve scalable, flexible, and controllable routing. In segment routing, multiple multi-protocol label switch labels are stacked in the packet header to complete end-to-end transmission, which may lead to a large label stack and a long packet header. Thus, scalability issues may occur when segment routing is applied to large-scale networks. To address this issue, multiple mechanisms and algorithms have been proposed for minimizing the label stack size. However, we argue that these methods ignore the constraint on the maximum segment list depth (SLD), since the typical network equipment can currently only support three to five layers of labels. In this paper, we study segment routing with the maximum SLD constraint and demonstrate that issues, such as explosive increases in the size of the label space and the management overheads will arise when the maximum SLD constraint is imposed. To address these issues, we make contributions from two main aspects. First, based on the network programmability that is provided by openflow, a novel segment routing architecture with improved data plane is proposed that reduces the overhead of additional flow entries and label space. Second, a new path encoding scheme is designed to minimize the SLD under the given maximum constraint, while taking multiple types of overhead into consideration. Moreover, we also perform simulations under different scenarios to evaluate the performances of the proposed algorithms. The simulation results demonstrate that the proposed mechanisms and algorithms can address the issues of segment routing when there is a constraint on the maximum SLD.

Large-scale data centers are major infrastructures in the big data era. Therefore, a stable and optimized architecture is required for data center networks (DCNs) to provide services to the applications. Many studies use software-defined network (SDN)-based multipath TCP (MPTCP) implementation to utilize the entire DCN's performance and achieve good results. However, the deployment cost is high. In SDN-based MPTCP solutions, the flow allocation mechanism leads to a large number of forwarding rules, which may lead to storage consumption. Considering the advantages and limitations of the SDN-based MPTCP solution, we aim to reduce the deployment cost due to the use of an extremely expensive storage resource-ternary content addressable memory (TCAM). We combine MPTCP and segment routing (SR) for traffic management to limit the storage requirements. And to the best of our knowledge, we are among the first to use the collaboration of MPTCP and SR in multi-rooted DCN topologies. To explain how MPTCP and SR work together, we use four-layer DCN architecture for better description, which contains physical topology, SR over the topology, multiple path selection supplied by MPTCP, and traffic scheduling on the selected paths. Finally, we implement the proposed design in a simulated SDN-based DCN environment. The simulation results reveal the great benefits of such a collaborative approach.