XDP-Based SmartNIC Hardware Performance Acceleration for Next-Generation Networks

Abstract

Next-generation networks are expected to combine advanced physical and digital technologies in super-high-speed connected system infrastructures, gaining critical operation competitiveness of improved efficiency, productivity and quality of services. Towards a fully digital and connected world, these platforms will enable infrastructure virtualization and support of edge processing, making emerging sectors, such as Industry 4.0, ready to exploit its full potentials. Nevertheless, the fast growth of data-centric and automated systems may exceed the capabilities of the overall infrastructure beyond the radio access networks, becoming unable to fulfil the demands of vertical sectors and representing a bottleneck. To minimize the negative effects that could affect critical services in a heavily loaded network, it is essential for network providers to deploy highly scalable and prioritisable in-network optimisation schemes to meet industry expectations in next-generation networks. To this end, this work presents a novel framework that leverages extended Berkeley Packet Filter (eBPF) and eXpress Data Path (XDP) to offload network functions to reduce unnecessary overhead in the backbone infrastructure. The proposed solution is envisioned to be implemented as a Network Application (NetApp) service, which will greatly benefit the compatibility with next-generation networking ecosystem empowered by Artificial Intelligence (AI), advanced automation, multi-domain network slicing, and other related technologies. The achieved results demonstrate key performance improvements in terms of packet processing capacity as high as about 18 million packets per second (Mpps), system throughput up to 6.1 Mpps with 0% of packet loss, and illustrate the flexibility of the framework to adapt to multiple network policy rules dynamically on demand.

Next Generation AI-Based Firewalls: A Comparative Study

“…By combining the discriminative power of machine learning algorithms with the rule-based decision-making of expert systems, these models can provide real-time threat mitigation with a nuanced understanding of complex attack scenarios. The research by authors in [28] exemplifies the efficacy of such hybrid frameworks, demonstrating their potential to offer a more holistic defense against a broad spectrum of cyber threats [23]. As cyber adversaries continually evolve their tactics, the versatility of hybrid models stands out, offering a promising avenue for fortifying NGFWs and adapting to the dynamic nature of modern cybersecurity challenges [32].…”

“…Behavioral analysis, in the context of NGFWs, involves the continuous monitoring and learning of normal network behavior, enabling the identification of anomalies indicative of malicious activities [23]. By adopting reinforcement learning, the authors in [28] harness the power of iterative decision-making processes, allowing NGFWs to adapt and learn from changing network dynamics over time [12]. This approach not only enhances the accuracy of threat detection but also facilitates a more adaptive and responsive defense mechanism against novel and emerging threats [12].…”

Empowering Cloud Computing With Network Acceleration: A Survey

“…In this context, the need to integrate network acceleration technologies is perhaps even more compelling than in cloud data centers. In the telco ecosystem, the literature on NFV (see Section I-A) already explored the adoption of software and hardware acceleration technologies to reduce the virtualization overhead [10], [52]. However, as we previously discussed, this integration is not considered an option available as a service to general-purpose applications.…”

“…In the networking domain, these devices can effectively execute compute-intensive operations, such as protocol processing, much faster than general-purpose processors, by removing the bottleneck of software-based protocol implementations and by allowing application developers to leverage the full potential of modern network links. Also in this case, different approaches can either retain the flexibility and isolation of in-kernel networking [52] or adopt a kernelbypassing solution for the sake of performance (Fig. 4d).…”

NetLabeller: Architecture with data extraction and labelling framework for beyond 5G networks

“…This decision will depend on multiple factors, in particular, the state of the network at any given time, the type of rule to be enforced and the number of rules currently enforced in every datapath technology. For example, in a use case were a drop rule is desired to be optimised, we will prioritise speed of execution, choosing datapath technologies located at the lowest point in the software stack, such as XDP [12]. Another example in the same scenario, taking into account other criteria such as having a lot of rules enforced in XDP, we could optimise the insertion of a new rule using a faster one such as Open vSwitch (OVS) [13].…”