Sketch-based solutions are widely used to estimate item frequencies in infinite data streams.Traditional hand-crafted sketches face the bottleneck of further eliminating errors because they cannot fully utilize the data stream distribution.Although recent neural sketches represented by MetaSketch and LegoSketch have improved generalization capabilities, they face bottlenecks such as high computational overhead and parameter sensitivity.Meanwhile, they ignore load information, fail to fully utilize the local information in hand-crafted sketches, and do not focus on the frequent items that are usually more important in data streams.In this paper, we propose RatioSketch, a novel lightweight neural network correction framework that synergizes the advantages of hand-crafted sketches and neural sketches in a ``micro-correction'' paradigm.The key idea is to retain the efficient underlying data structure of the hand-crafted sketch and to build a neural correction layer in its output space. We select multiple representative hand-crafted sketches as use cases to study the correction performance of RatioSketch on them.Extensive experimental evaluations on several real-world datasets show that RatioSketch-corrected sketches achieve consistently higher estimation accuracy than their uncorrected counterparts, as well as outperforming neural baselines such as MetaSketch and LegoSketch under identical memory budgets.

Ternary Content Addressable Memory (TCAM) is a high-speed, parallel-search memory that enables fast lookups for both exact-match and wildcard-match rules. TCAM serves as the standard hardware implementation of flow tables in Software-Defined Networking (SDN) switches, delivering line-rate packet classification to enforce fine-grained forwarding policies. Due to its high-cost and power-hungry design, TCAM faces a scalability challenge in accommodating large-scale rule sets in modern backbone networks. Inspired by hierarchical cache architectures in modern memory systems, TCAM-based rule-caching systems incorporate Random Access Memory (RAM) as a cost-effective, large-capacity auxiliary memory. Specifically, TCAM caches heavy-hitting rules to capture most packets from hot flows, while RAM maintains the complete rule set to handle cache-miss packets. Meanwhile, an update strategy is employed to manage the replacement of cached rules between TCAM and RAM. However, prior rule-caching systems have either failed to eliminate cross-rule dependencies or have restricted their match patterns to prefix rules only. To address these limitations, we propose AWEsome-Cache, a unified framework designed to fundamentally eliminate cross-rule dependencies for wildcard rules with arbitrary match patterns. The key idea is to construct isolation rules for hot flows by specifying the minimum number of wildcard bits in their best-match rules, thereby eliminating overlaps with all direct dependent rules. AWEsome-Cache further develops efficient replacement algorithms for TCAM updates, enabling adaptation to dynamic traffic locality. Experimental evaluations on both prefix and non-prefix rule sets show that AWEsome-Cache achieves cache-hit rates comparable to state-of-the-art baselines while reducing TCAM occupancy by 75.9%.

With the rapid development of networks, optimizing network utilization has become a crucial task. Consequently, previous research has extensively investigated Traffic Engineering (TE). However, these studies deeply couple TE systems with specific network topologies, resulting in a lack of generalization ability to different network topologies. Therefore, a trained model cannot be deployed to other topologies, causing additional training overhead for network operators. Moreover, these systems struggle to cope with changes in network topology caused by business expansion or network failures. Thus traditional TE systems can bring performance degradation under those scenarios. In this paper, we propose UniTE, a universal TE system. We leverage Graph Attention Network (GAT) and mask matrices to empower the model with the capability to manage traffic in diverse network topologies. To avoid the occurrence of loops along routing paths, we introduce Double Directed Acyclic Graphs (DDAG) to restrict the forwarding nodes for data packets. We train UniTE on various topologies and evaluate its performance. Finally, we assess UniTE's performance under network topology changes induced by network failures, and our method can outperform up to 15.6%.

Offloading machine learning models for network classification on high-throughput programmable switches is a promising technology, enabling line-speed in-network classification. Existing solutions are centralized, deploying a complete but heavy model on a single switch with limited hardware resources, causing unsatisfactory accuracy, network-wide resource wastage, and non-generic single-task classification. Therefore, we propose In-Forest-M, a general distributed multi-task in-network classification framework. Firstly, we develop a Lightweight Ensemble Generic Optional Model (LEGO), which can be transformed into base models with full functionality. Each switch only needs to deploy lightweight base models rather than complete ensemble models. The significant reduction in resource consumption allows the deployment of larger models with higher accuracy and more models that support diverse tasks. We employ a fine-grained enhancement mechanism to enhance the classification performance of base models. As traffic traverses different switches, In-Forest-M aggregates the classification results of multiple enhanced base models to improve accuracy further. Secondly, we introduce a two-phase resource-aware model allocation strategy that assigns different task-specific enhanced base models to switches under resource constraints and task requirements. To respond to dynamic traffic changes, we design an optimization-driven reinforcement learning algorithm. Moreover, we propose a lightweight update mechanism for flexible model scaling. Comprehensive experiments reveal that, compared with state-of-the-art in-network classification solutions in three real network topologies, In-Forest-M achieves increased accuracy and reduced switch rules while exhibiting great generality in multi-task classification.

Stateful multicast protocols manage multicast group memberships by maintaining state information about active groups and their members. They have seen limited adoption in the modern internet due to lack of scalability, simplicity, and flexibility. Although stateless multicast protocols, like BIER, eliminate extensive state management, they still face complex tree computation and limited scalability for concurrent requests. In this paper, we propose Hawkeye, a stateless multicast mechanism with deep reinforcement learning (DRL) for real-time responses to dynamic multicast requests with near-optimal multicast TE performance. This mechanism is suited for Software-Defined Networking (SDN) environment where the controller has a global view of the network and supports flexible configuration of network resources for traffic engineering. For real-time responses to multicast requests, we leverage DRL enhanced by a temporal convolutional network (TCN) to model the sequential feature of dynamic group membership, and thus are able to build multicast trees proactively for upcoming requests. We develop a novel source aggregation mechanism to facilitate the convergence of the DRL agent under high volume of multicast requests. Moreover, to improve the practicality and robustness of Hawkeye, we design incremental deployment and single failure handling mechanisms, which take advantages of source aggregation and fit well with multicast routing. Evaluation with real-world topologies and multicast requests demonstrates that Hawkeye responds effectively to dynamic multicast requests. Itoffers rapid routing decisions, e.g., making routing decisions in under 5ms on a tested topology, and reduces path latency variation by up to 89.5%, with less than a 10% increase in bandwidth consumption compared to the offline theoretical minimum.

The emergence of metaverse applications and services heralds a new era of immersive driving experiences in future vehicular ad hoc networks (VANETs). This unprecedented metaverse-infused driving experience, however, requires high-throughput transmissions for real-time high-quality 360° video streaming, which challenges today’s limited bandwidth resources provisioned by wireless communication infrastructure. To this end, this article introduces virtual road, real connection (VRRC), a novel framework for enhancing the metaverse-infused driving experience in VANET through the Internet of Robotic Things (IoRT). In the face of limited bandwidth resources, VRRC effectively addresses the challenge of high-throughput demand from two key perspectives. First, VRRC reduces redundant transmission by implementing a graph neural network (GNN)-based vehicle clustering method for dynamic multicast group formation, taking into account both the geographical status of vehicles and their communication patterns. Second, VRRC aggregates bandwidth resources across various channels by employing a multiagent reinforcement learning (MARL)–based multipath packet scheduling policy to adapt to heterogeneous channel conditions and dynamic vehicular mobility. Extensive experiments with real-world vehicular traces validate the effectiveness of VRRC and demonstrate its outperformance in reducing redundant traffic by 54% and improving overall throughput by 28%. VRRC represents a substantial leap forward in the integration of the metaverse experience into VANET.

Ternary Content Addressable Memory (TCAM) is a specialized high-speed memory that enables fast parallel lookups for both exact-match rules and wildcard-match rules. TCAM has become a standard hardware component in Software-Defined Networking (SDN) switches to implement flow tables for packet classification. However, limited TCAM storage capacity poses a significant scalability challenge for SDN to enforce fine-grained policy-based forwarding. To this end, TCAM-based rule-caching systems are proposed by combining TCAM with Random Access Memory (RAM). Specifically, TCAM caches heavy-hitting rules to capture packets from hot flows, while RAM accommodates the complete ruleset for other cache-miss packets. However, previous rule-caching systems either failed to eliminate cross-rule dependencies or restricted their applications to prefix rules only. In this work, we propose AWEsome-Cache, a unifying framework to fundamentally eliminate cross-rule dependencies for wildcard rules with arbitrary matching patterns. The rationale behind AWEsome-Cache is to concretize a minimum number of wildcard bits in the best-match rule, thereby pruning its overlapping match fields with all direct dependent rules. AWEsome-Cache also develops replacement algorithms during TCAM updates to adapt to dynamic traffic locality. Experiments with prefix and non-prefix rules show that AWEsome-Cache outperforms baselines in achieving a comparable cache-hit rate but requiring 75.9% less TCAM occupancy.

Multicast traffic is growing rapidly due to the development of multimedia streaming. Lately, stateless multicast protocols, such as BIER, have been proposed to solve the excessive routing states problem of traditional multicast protocols. However, the high complexity of multicast tree computation and the limited scalability for concurrent requests still pose daunting challenges, especially under dynamic group membership. In this paper, we propose Hawkeye, a dynamic and stateless multicast mechanism with deep reinforcement learning (DRL) approach. For real-time responses to multicast requests, we leverage DRL enhanced by a temporal convolutional network (TCN) to model the sequential feature of dynamic group membership and thus is able to build multicast trees proactively for upcoming requests. Moreover, an innovative source aggregation mechanism is designed to help the DRL agent converge when faced with a large amount of multicast requests, and relieve ingress routers from excessive routing states. Evaluation with real-world topologies and multicast requests demonstrates that Hawkeye adapts well to dynamic multicast: it reduces the variation of path latency by up to 89.5% with less than 12% additional bandwidth consumption compared with the theoretical optimum.

Ternary Content Addressable Memory (TCAM) is an essential hardware component in SDN-enabled switches, which supports fast lookup speed and flexible matching patterns. However, TCAM’s limited storage capacity has long been a scalability challenge to enforce fine-grained forwarding policies in SDN. Based on the observation of traffic locality, the rule-caching mechanism employs a combination of TCAM and Random Access Memory (RAM) to maintain the forwarding rules of large and small flows, respectively. However, previous works cannot identify large flows timely and accurately, and suffer from high computational complexity when addressing rule dependencies in TCAM. Worse still, TCAM only caches the forwarding rules of large flows but ignores the latency requirements of small flows. Small flows encounter cache-miss in TCAM and then will be diverted to RAM, where they have to experience slow lookup processes. To jointly optimize the performance of both high-throughput large flows and latency-sensitive small flows, we propose a hybrid rule-caching framework, H-Cache, to scale traffic-aware forwarding policies in SDN. H-Cache identifies large flows through a collaboration of learning-based and threshold-based methods to achieve early detection and high accuracy, and proposes a time-efficient greedy heuristic to address rule dependencies. For small flows, H-Cache establishes default paths in TCAM to speed up their lookup processes, and also reduces their TCAM occupancy through label switching and region partitioning. Experiments with both real-world and synthetic datasets demonstrate that H-Cache increases TCAM utilization by an average of 11% and reduces the average completion time of small flows by almost 70%.

Segment Routing (SR) provides Traffic Engineering (TE) with Explicit Path Control (EPC) by steering data flows passing through a list of SR routers along a desired path. However, large-scale migration from a pure IP network to a full SR one requires prohibitive hardware replacement and software update. Therefore, network operators prefer to upgrade a subset of IP routers into SR routers during a transitional period. This paper proposes EPC-TE to optimize TE performance in hybrid IP/SR networks where partially deployed SR routers coexist with legacy IP routers. We propose a concept of key nodes to achieve EPC over desired paths and a criterion to select which IP routers to upgrade first under a pre-defined upgrading ratio. EPC-TE leverages Deep Reinforcement Learning (DRL) to inference the optimal traffic splitting ratio across multiple controllable paths between source-destination pairs. EPC-TE can achieve comparable TE performance as a full SR network with an upgrading ratio less than 30%. Extensive experimental results with real-world topologies show that EPC-TE significantly outperforms other baseline TE solutions in minimizing maximum link utilization.

Segment Routing (SR) provides Traffic Engineering (TE) with the ability of explicit path control by steering traffic passing through specific SR routers along a desired path. However, large-scale migration from a legacy IP network to a full SR-enabled one requires prohibitive hardware replacement and software update. Therefore, network operators prefer to upgrade a subset of IP routers into SR routers during a transitional period. This paper proposes SmartTE to optimize TE performance in hybrid IP/SR networks where partially deployed SR routers coexist with legacy IP routers. We use two centrality criteria in graph theory to decide which IP routers should be upgraded into SR routers under a given upgrading ratio. SmartTE leverages Deep Reinforcement Learning (DRL) to infer the optimal traffic splitting ratio across multiple pre-defined paths between source-destination pairs. Extensive experimental results with real-world topologies show that SmartTE outperforms other baseline TE solutions in minimizing the maximum link utilization and achieves comparable performance as a full SR network by upgrading only 30% IP routers.