Over-smoothing in Graph Neural Networks (GNNs) causes collapse in distinct node features, particularly on heterophilic graphs where adjacent nodes often have dissimilar labels. Although sheaf neural networks partially mitigate this problem, they typically rely on static or heavily parameterized sheaf structures that hinder generalization and scalability. Existing sheaf-based models either predefine restriction maps or introduce excessive complexity, yet fail to provide rigorous stability guarantees. In this paper, we introduce a novel scheme called SGPC (Sheaf GNNs with PAC-Bayes Calibration), a unified architecture that combines cellular-sheaf message passing with several mechanisms, including optimal transport-based lifting, variance-reduced diffusion, and PAC-Bayes spectral regularization for robust semi-supervised node classification. We establish performance bounds theoretically and demonstrate that end-to-end training in linear computational complexity can achieve the resulting bound-aware objective. Experiments on nine homophilic and heterophilic benchmarks show that SGPC outperforms state-of-the-art spectral and sheaf-based GNNs while providing certified confidence intervals on unseen nodes.

Graph Neural Networks (GNNs) often struggle with heterophilic graphs, where neighboring nodes tend to have dissimilar labels-a common scenario in real-world networks. This paper addresses this limitation through a two-phase framework called ConSM (Confidence-based Subgraph Matching). First, we introduce a confidence-aware subgraph matching module that estimates edge coefficients by comparing the structural similarity of 2-hop neighborhoods using optimal transport. This process identifies task-irrelevant or misleading edges based on a tunable confidence ratio. Second, we integrate these edge coefficients into a sign-aware label propagation mechanism that adaptively encourages or discourages message passing based on edge confidence, thereby enhancing GNN robustness under heterophily. Compared to our earlier conference version [1], this manuscript provides (i) a clearer justification of key design choices such as subgraph-based reasoning and the use of 2-hop neighborhoods, (ii) an adaptive strategy to tune the confidence ratio without manual search, and (iii) extensive new experiments covering recent heterophily-oriented baselines and larger, leakage-free datasets. Empirical results show that ConSM improves classification accuracy, mitigates over-smoothing, and remains effective across both homophilic and heterophilic regimes.

Graph Neural Networks (GNNs) thrive on message passing (MP) but are vulnerable when the graph carries many heterophilic or misclassified edges. Prior analyses suggest that signed propagation can mitigate over-smoothing under low edge-error rates, yet they implicitly assume perfect edge labels and the presence of self-loops. We revisit this setting and show that, under high edge uncertainty, propagating any information may harm node separability even with signed weights. Our key insight is to decide not to propagate along uncertain edges adaptively. Concretely, we intentionally omit self-loops to isolate pure neighbor influence for a clearer theoretical analysis, adopt a row-stochastic (asymmetric) operator that matches the Markov-chain view of MP and simplifies spectral-radius proofs, and dynamically estimate the local homophily $b_i$ and edge-classification error $e_t$ during training via an EM procedure. We prove that our selective blocking yields a sub-stochastic propagation matrix whose joint spectral radius exceeds that of signed GNNs under high $e_t$, guaranteeing reduced over-smoothing, and we supply a lemma showing that class-discriminative signals survive even when the operator is rank-deficient. Extensive experiments on seven homophilic and heterophilic benchmarks confirm that the proposed adaptive blocking outperforms strong baselines.

Message-passing neural networks are widely employed in various graph mining applications. However, these methods are susceptible to the scarcity of labeled data, which often leads to overfitting. Our observations suggest that sparse initial vectors further exacerbate this issue by failing to fully represent the range of learnable parameters. This sparsity can hinder the optimization of specific dimensions in the initial projection matrix, as the training samples may not adequately span these parameters. To overcome this challenge, we propose a novel perturbation technique that introduces variability to the initial features and the projection hyperplane. Notably, even without employing grid search, we demonstrate that shifting with a small estimated value mitigates this problem more effectively than other perturbation methods. Experimental results on real-world datasets reveal that our technique significantly enhances node classification accuracy in semi-supervised scenarios.

The issue of data sparsity poses a significant challenge to recommender systems. In response to this, algorithms that leverage side information such as review texts have been proposed. Furthermore, Cross-Domain Recommendation (CDR), which captures domain-shareable knowledge and transfers it from a richer domain (source) to a sparser one (target) has emerged recently. Nevertheless, existing methodologies assume an Euclidean embedding space, encountering difficulties in accurately representing richer text information and managing complex user-item interactions. This paper advocates a hyperbolic CDR approach for modeling review-based user-item relationships. We first emphasize that conventional distance-based domain alignment techniques may cause problems because small modifications in hyperbolic geometry result in magnified perturbations, ultimately leading to the collapse of hierarchical structures. To address this challenge, we propose hierarchy-aware embedding and domain alignment schemes that adjust the scale to extract domain-shareable information without disrupting structural forms. Extensive experiments substantiate the efficiency, robustness, and scalability of the proposed model. The source code is given here https://github.com/ChoiYoonHyuk/HEAD.

Graph Neural Networks (GNNs) exhibit satisfactory performance on homophilic networks, where most edges connect two nodes with the same label. However, their effectiveness diminishes as the graphs become heterophilic (low homophily), prompting the exploration of various message-passing schemes. In particular, assigning negative weights to heterophilic edges (signed propagation) for message-passing has gained significant attention, and some studies theoretically confirm its effectiveness. Nevertheless, prior theorems assume binary classification scenarios, which may not hold well for graphs with multiple classes. To solve this limitation, we offer new theoretical insights into GNNs in multi-class environments and identify the drawbacks of employing signed propagation from two perspectives: message-passing and parameter update. We found that signed propagation without considering feature distribution can degrade the separability of dissimilar neighbors, which also increases prediction uncertainty (e.g., conflicting evidence) that can cause instability. To address these limitations, we introduce two novel calibration strategies aiming to improve discrimination power while reducing entropy in predictions. Through theoretical and extensive experimental analysis, we demonstrate that the proposed schemes enhance the performance of both signed and general message-passing neural networks (Choi et al. 2023).

Graph neural networks (GNNs) that collect information from neighbors are commonly utilized in semi-supervised learning contexts. In particular, a significant body of research has been dedicated to developing effective graph filters and aggregation methods to filter the information from adjacent nodes. Despite their efficacy, these approaches may encounter challenges due to the sparsity of training nodes, especially when their features are represented as sparse vectors (e.g., bag-of-words). This condition can lead to the overfitting of certain dimensions within the first projection matrix (hyperplane), as the training samples may not adequately represent the full spectrum of learnable parameters. To solve this limitation, we propose an innovative perturbation technique. Specifically, we introduce additional training variability by modifying both the initial features and the hyperplane, which contributes to the reduction of prediction variance by updating the entire dimensions. To the best of our knowledge, our approach is the first to address the overfitting issue in GNNs precipitated by sparse node features. Comprehensive experiments on real-world datasets and ablation studies affirm that our proposed method significantly enhances node classification performance, with improvements of up to 46.5% in GNN algorithms.

Adaptive bitrate (ABR) algorithms, the de facto standard for video streaming, aim to provide the best user quality of experience (QoE) under fluctuating operating environments. Prior ABR algorithms addressed the QoE maximization problem with a plethora of approximate optimization techniques including model predictive control (MPC), Lyapunov optimization, and deep reinforcement learning (DRL). Even though these algorithms provide adequate performances, most of them primarily focus on choosing optimal bitrates while prohibiting duplicated downloads. We point out that allowing duplicated chunk downloads for replacement can potentially enhance the QoE if they are carefully administered. Moreover, combining ABR algorithms with other techniques may compromise optimization results as each algorithm’s optimization problem does not account for the others. In this paper, we first formulate a novel optimization problem with an expanded decision dimension that encompasses the chunk replacement as well as the bitrate selection in the action space. We then propose COCKTAIL, a DRL-based ABR algorithm that discovers efficient solutions to the new optimization problem by using several learning techniques. Experiments on real-world network traces show that COCKTAIL outperforms state-of-the-art baselines with improvements in average QoE up to 16.4%.

Graph Contrastive Learning (GCL) is an effective method for discovering meaningful patterns in graph data. By evaluating diverse augmentations of the graph, GCL learns discriminative representations and provides a flexible and scalable mechanism for various graph mining tasks. This paper proposes a novel contrastive learning framework by introducing Laplacian perturbation. The proposed framework offers a distinct advantage by employing an indirect perturbation method, which provides a more stable approach while maintaining the perturbation effects. Moreover, it exhibits a wide range of applicability by not being restricted to specific graph types. We demonstrate that a spectral graph convolution based on the Laplacian successfully extracts representations from diverse graph types. Our extensive experiments on a variety of real-world datasets, covering multiple graph types, show that the proposed model outperforms state-of-the-art baselines in both node classification and link sign prediction tasks.

Signed directed graphs contain both sign and direction information on their edges, providing richer information about real-world phenomena compared to unsigned or undirected graphs. However, analyzing such graphs is more challenging due to their complexity, and the limited availability of existing methods. Consequently, despite their potential uses, signed directed graphs have received less research attention. In this paper, we propose a novel spectral graph convolution model that effectively captures the underlying patterns in signed directed graphs. To this end, we introduce a complex Hermitian adjacency matrix that can represent both sign and direction of edges using complex numbers. We then define a magnetic Laplacian matrix based on the adjacency matrix, which we use to perform spectral convolution. We demonstrate that the magnetic Laplacian matrix is positive semi-definite (PSD), which guarantees its applicability to spectral methods. Compared to traditional Laplacians, the magnetic Laplacian captures additional edge information, which makes it a more informative tool for graph analysis. By leveraging the information of signed directed edges, our method generates embeddings that are more representative of the underlying graph structure. Furthermore, we showed that the proposed method has wide applicability for various graph types and is the most generalized Laplacian form. We evaluate the effectiveness of the proposed model through extensive experiments on several real-world datasets. The results demonstrate that our method outperforms state-of-the-art techniques in signed directed graph embedding.

Graph Neural Networks (GNNs) have proven to be powerful in many graph-based applications. However, they fail to generalize well under heterophilic setups, where neighbor nodes have different labels. To address this challenge, we employ a confidence ratio as a hyper-parameter, assuming that some of the edges are disassortative (heterophilic). Here, we propose a two-phased algorithm. Firstly, we determine edge coefficients through subgraph matching using a supplementary module. Then, we apply GNNs with a modified label propagation mechanism to utilize the edge coefficients effectively. Specifically, our supplementary module identifies a certain proportion of task-irrelevant edges based on a given confidence ratio. Using the remaining edges, we employ the widely used optimal transport to measure the similarity between two nodes with their subgraphs. Finally, using the coefficients as supplementary information on GNNs, we improve the label propagation mechanism which can prevent two nodes with smaller weights from being closer. The experiments on benchmark datasets show that our model alleviates over-smoothing and improves performance.

A cross-domain recommendation has shown promising results in solving data-sparsity and cold-start problems. Despite such progress, existing methods focus on domain-shareable information (overlapped users or same contexts) for a knowledge transfer, and they fail to generalize well without such requirements. To deal with these problems, we suggest utilizing review texts that are general to most e-commerce systems. Our model (named SER) uses three text analysis modules, guided by a single domain discriminator for disentangled representation learning. Here, we suggest a novel optimization strategy that can enhance the quality of domain disentanglement, and also debilitates detrimental information of a source domain. Also, we extend the encoding network from a single to multiple domains, which has proven to be powerful for review-based recommender systems. Extensive experiments and ablation studies demonstrate that our method is efficient, robust, and scalable compared to the state-of-the-art single and cross-domain recommendation methods.

Time Slotted Channel Hopping (TSCH), defined in the IEEE 802.15.4e standard requires complicated slot scheduling to enjoy its collision-free multihop communication capabilities fully. Recently, several autonomous scheduling algorithms that allocate slots without central control and additional control message exchange have been devised. However, autonomous algorithms are traffic oblivious, assigning the same number of slots regardless of traffic demands. Undifferentiated resource allotment is identified as the root cause that impairs the performance of TSCH networks. To address the issue, we propose A3, an autonomous and adaptive slot allocation scheme that adjusts the number of slots per slotframe according to varying traffic loads. A key component of A3 is receiver-side load estimation in real-time without any explicit control message exchange. A3 can be combined with any autonomous scheduling protocols, such as Orchestra and ALICE. With extensive evaluation on a large public testbed comprising 62 nodes, we verify that A3 significantly enhances autonomous scheduling algorithms' performance. It improves throughput by more than twice, PDR (Packet Delivery Ratio) by more than six times, and reduces the latency by ten times.

As low-power wide-area (LPWA) networks emerge as a cost-effective choice of technologies for city-wide Internet-of-Things (IoT) applications, LoRaWAN, one of the most promising unlicensed band techniques, has received much attention from academia. LoRaWAN presents a set of tunable transmission parameters, along with an adaptive data rate (ADR) mechanism, to promote the best performance under the variable link state. But the performance of ADR, whose design neglects the complex correlation between such parameters, is yet to be practical in terms of both efficiency and scalability. In this article, we derive theoretical performance models of class A unconfirmed-mode LoRaWAN, focusing on the impact of coding rate (CR), a parameter that has not been explored in prior researches. Then, we present EARN, an enhanced greedy ADR mechanism with CR adaptation, to optimize the tradeoff between delivery ratio and energy consumption. In EARN design, we leverage the capture effect to increase the survival rate of colliding signals and introduce a concept of adaptive SNR margin to endure noisy link states. We validate our models and the feasibility of the CR adaptation with an empirical study, and large-scale simulations reveal that our method outperforms the conventional schemes.

Although low-power lossy network (LLN), at its early stage, commonly used asynchronous link layer protocols for simple operation on resource-constrained nodes, development of embedded hardware and time synchronization technologies made Time-Slotted Channel Hopping (TSCH) viable in LLN (now part of IEEE 802.15.4e standard). TSCH has the potential to be a link layer solution for LLN due to its resilience to wireless interference (e.g., WiFi) and multi-path fading. However, its slotted operation incurs non-trivial cell scheduling overhead: two nodes should wake up at a time-frequency cell together to exchange a packet. Efficient cell scheduling in dynamic multihop topology in wireless environments has been an open issue, preventing TSCH's wide adoption in practice. This work introduces ALICE, a novel autonomous link-based cell scheduling scheme which allocates a unique cell for each directional link (a pair of nodes and traffic direction) by closely interacting with the routing layer and using only local information, without any additional communication overhead. We implement ALICE on Contiki and evaluate its effectiveness on the IoT-LAB public testbed with 68 nodes. ALICE generally outperforms Orchestra (the state-of-the-art method) and even more so under heavy traffic and high node density, increasing throughput by 2 times with 98.3% reliability and reducing latency by 70%, route changes by 95%, and radio duty cycle by 35%. ALICE can serve as an autonomous scheduling framework, which paves the way for TSCH-based LLN to go on.

Considering location data for friendship prediction has become prevalent due to the huge success of online social networks. However, few studies have focused on investigating the possibility of using mobility information in a dense place such as a campus. The research direction for those mobility networks should be treated differently. We propose a CNN-based noble framework for friendship prediction, which starts from collecting location data to a classification model to learn their relation between friendships and mobility. From the experiment, we show that our system outperforms empirical supervised learning techniques and also can be useful for friendship prediction of the future.

Adaptive video streaming can provide adequate Quality of Experience by dynamically adjusting video rates in responding to fluctuating service conditions. However, adaptive video streaming shows dismal performances in wireless networks where several players share a wireless bottleneck link. This paper proposes a novel video rate selection scheme called XMAS for efficient video streaming in wireless networks. XMAS consists of two components: an available bandwidth estimation part and a video rate selection part. To redress the problem of inaccurate bandwidth estimation due to peculiar on-off transmission patterns of mobile video streaming, we devised a novel client-based traffic shaping scheme that effectively throttles server's packet transmission. Equipped with accurate bandwidth estimates, XMAS determines target transmission rates considering playback buffer levels. We implemented XMAS on Linux and performed rigorous experiments to analyze its behavior and performance. Our performance results showed that XMAS achieves up to 20% increase in average video rates while reducing rebuffer rates significantly.

Today, online social media is being used as a primary means of interaction among people rather than any other communication medium. People use social media for socializing and collecting and sharing information, which presumes online social media is always reliable. However, there are a lot of advertisements or cyberattacks that take advantage of the impact of online social media. For example, spam, phishing, and advertising through influence manipulation. They are not only directly damaging users, but they are also giving indirect damages that will cause damage to the credibility of online social media to decline. We (1) propose measurements for online social status to identify the attackers in online social media and (2) analyze the difference in the manner in which a normal user and an attacker make a neighbor relationship.

Wi-Fi channel scanning-the task of searching for available channels at a given location-is a fundamental feature to maintain always-available and high-quality wireless connectivity in today's mobile devices. However, it is a challenging task to design an intelligent scanning algorithm that can discover available access points (APs) in a short time period, because the scanning station has no prior knowledge on the APs in its vicinity. Therefore, traditional scanning algorithms seek to discover available APs by scanning the full set of channels, including channels where no APs exist. This often induces unnecessary scanning latency and/or energy consumption. In this paper, we present a novel scheme, called C-SCAN, that exploits a low-power wireless personal area network interface, such as Bluetooth or ZigBee integrated into the device to offload Wi-Fi scanning overhead by suppressing unnecessary scanning of AP-free Wi-Fi channels. To this end, C-SCAN inspects channel information with a lowpower Bluetooth radio and identifies which Wi-Fi channels are in use, prior to the actual channel scanning with a Wi-Fi interface. By excluding the channels determined to be empty, the Wi-Fi scanning manager can perform scanning only on available Wi-Fi channels. Thereby, a significant performance gain in terms of delay and energy is obtained. We implement a prototype of C-SCAN using a Bluetooth-compliant wireless transceiver and demonstrate its efficiency. Experimental results show that CSCAN achieves high detection accuracy with low latency, even in dense Wi-Fi environments.

LPWAN (Low‐Power Wide Area Network) technologies such as LoRa and SigFox are emerging as a technology of choice for the Internet of Things (IoT) applications where tens of thousands of untethered devices are deployed over a wide area. In such operating environments, energy conservation is one of the most crucial concerns and network protocols adopt various power saving schemes to lengthen device lifetimes. For example, to avoid idle listening, LoRaWAN restricts downlink communications. However, the confined design philosophy impedes the deployment of IoT applications that require asynchronous downlink communications. In this paper, we design and implement an energy efficient downlink communication mechanism, named TRILO, for LoRaWAN. We aim to make TRILO be energy efficient while obeying an unavoidable trade‐off that balances between latency and energy consumption. TRILO adopts a beacon mechanism that periodically alerts end‐devices which have pending downlink frames. We implement the proposed protocol on top of commercially available LoRaWAN components and confirm that the protocol operates properly in real‐world experiments. Experimental results show that TRILO successfully transmits downlink frames without losses while uplink traffic suffers from a slight increase in latency because uplink transmissions should halt during beacons and downlink transmissions. Computer simulation results also show that the proposed scheme is more energy efficient than the legacy LoRaWAN downlink protocol.

This paper deals with the problem of severe wireless performance degradation when multiple wireless technologies are concurrently utilized in a same user device. This type of usage is already frequent in most smartphones and laptops, such as streaming Bluetooth audio while using a Wi-Fi download, and is more intensifying with IoT device deployment which triggers the coexistence of heterogeneous wireless technologies. To lower the form factor and the cost, chip vendors package multiple wireless interfaces into a single combo-chip where a common antenna is shared by multiple network technologies in a time division multiplexing manner. We issue that the careless operations of combo-chip design incur indeed performance degradation for in-device wireless coexistence and show the experimental results via TCP performance measurements in several smartphones and laptops. Our analysis reveals that the behavior negatively affects not only on the transmit power management of wireless access point, but also on the congestion control of TCP sender. We propose a cooperative switching scheme which incorporates TCP control behaviors for better coexistence and implement it on Android and Linux devices. Under the simultaneous use of in-device network interfaces, our approach led a WLAN throughput increment up to eight times without the mentioned issues. Further, this does not require any modification of TCP sender and wireless access point. Thus, the approach is directly applicable to existing mobile devices and also easily extendable to the combination of other in-device wireless technologies.

The concurrent deployment of heterogeneous wireless networks such as Wi-Fi, Bluetooth, and ZigBee has led to the severe interference problems in the 2.4GHz ISM band. In particular, ZigBee networks are susceptible to the interferences from other wireless technologies; For example, strong Wi-Fi signals trigger false alarms to ZigBee device that is performing low power idle listening and cause appreciable energy waste. In this paper, we propose a novel ZigBee signal detection scheme, called CoSense that accurately identifies ZigBee signals in the presence of the cross-technology interferences. CoSense, which is a highly reliable signal correlation technique, enjoys the following three advantages: First, CoSense reduces false wake-ups, which typically consume energy unnecessarily. Secondly, CoSense is robust against heterogeneous interference scenarios because its signal correlation feature has been shown to work well in bad channel conditions. Third, CoSense is backward-compatible and does not require to change the traditional ZigBee networks. We have implemented CoSense on the USRP/GNURadio platform in order to prove its feasibility. The results show that, under typical setting, CoSense indeed reduces the false alarm rate and its overhead is tolerable. We can conclude that CoSense saves energy by up to 63% in heterogeneous network environments where weak ZigBee signals are overwhelmed by strong signals such as Wi-Fi.

Recently, we have received a growing number of reports that complain about poor and unstable internet connections at bus stops in metro Seoul. Careful analyses led us to conclude that Wi-Fi APs equipped on buses instigate the trouble. According to the ambitious free Wi-Fi expansion plan by the city of Seoul, public buses started to equip Wi-Fi APs. As buses with APs stop and go, they actualize intermittent connection opportunities to riders waiting at the bus stops. However, the connection durations are too short such that bus APs are a nuisance rather than a convenience. We collected the basic statistics such as AP inter-arrival and sojourn times and measured link level performance metrics. We observed the effect of frequent frame losses on the TCP congestion control and eventually on the TCP throughput. We also measured the performance of applications such as PLT (Page Load Time). The measurement results showed that passing APs are useful only for some applications in very limited situations while they are virtually useless and just irritations in many cases. We also discovered that poor Wi-Fi connections pervert MPTCP; MPTCP performs worse than the generic single path TCP over the LTE network. We expect that our results will be used as the reference data in redesigning Wi-Fi offloading mechanisms as well as in planning and deploying urban Wi-Fi networks.

With the increasing move towards wide band operation in recent Wi-Fi networks, the frequency diversity awareness has become critical for throughput optimization. To exploit frequency diversity in Wi-Fi channels, the access point should measure the channel quality and coordinate the channel contention for all the stations. However, there is a tradeoff between achieving the frequency diversity gain and sustaining protocol efficiency because the channel estimation and coordination consume time and frequency resource that ideally should be used for data transfer. In this paper, we present Diversity-aware Wi-Fi (D-Fi), a novel PHY/MAC protocol, that efficiently exploits frequency diversity. In particular, D-Fi leverages an OFDM-based Bloom filter that synergistically integrates two operations: (i) the channel quality estimation and (ii) the contention based channel allocation. D-Fi also employs a machine learning (ML) method to resolve the false-positive ambiguity caused by the Bloom filter. Furthermore, we develop a decentralized algorithm, called Kε-greedy, based on the Multi-Armed Bandit (MAB) framework, so that it achieves sub-optimal performance by studying the gain for exploring new channel quality information. We implement the prototype of D-Fi on the USRP/GNURadio to validate the feasibility of our work. The experiments and trace-driven simulations show that D-Fi provides up to 3 x throughput improvement compared to the existing solutions.

With the growth of Internet usage and online social networks, the online Recommender Systems are becoming popular among system users. Although the influence of the recommender systems is expanding, the possibility of residing fake identities (Sybils) from nefarious users increase due to various reasons. To mitigate the impact of such users, several approaches are proposed. However, the need for robust algorithms is still necessary regarding recommender systems since the small portion of Sybils can distort the accuracy of predictions extremely. We propose a novel robust recommendation algorithm (RobuRec) using information level and admission control. The performance of RobuRec is experimented on various recommendation datasets with all possible Sybil attacks. The evaluation result shows that RobuRec can improve prediction error by 21% and 49% compared to two comparable schemes (LTSMF [23] and PCA [24], respectively). On all datasets and against various attack strategies, in turn, our RobuRec scheme shows the best peformance in terms of prediction shift.

To exploit frequency diversity in Wi-Fi channels, instantaneous channel quality must be estimated. However, there is a trade-off between acquiring channel quality information and improving protocol efficiency because channel estimation consumes time and frequency resource that ideally should be used for data transfer. In this paper, we present D-Fi (Diversity-aware Wi-Fi), a novel Wi-Fi PHY/MAC protocol, that capitalizes on frequency diversity gains while sustaining protocol efficiency. The D-Fi design allows to estimate channel quality while D-Fi is performing channel contention using an OFDM-based Bloom filter. To resolve the ambiguity caused by the Bloom filter, we adopt two methods: (i) An analysis-based multi channel backoff method enables to explore/exploit frequency diversity while reducing the occurrence of the ambiguity. (ii) Applying machine learning (ML) methods to the D-Fi PHY/MAC protocol corrects the ambiguity taken place already and makes our protocol reliable. We have shown the feasibility of D-Fi by implementing it on the USRP/GNURadio platform. Experiments and trace-driven simulations show that D-Fi successfully achieves frequency diversity gains without losing improved protocol efficiency.

"Real money trading" or "Gold farming" refers to a set of illicit practices for gathering and distributing virtual goods in online games for real money. Unlike previous work, we use network-wide economic interactions among in-game characters as a lens to monitor, detect and identify gold farming networks. Our work is based on a set of real in-game trade activity logs collected for one month in year 2010 from the world's second largest MMORPG called AION (with 3.4 million subscribers). This is the first work that empirically (i) shows that "free money network" is a promising measure/approximation for detecting and characterizing gold farming networks, and (ii) measures the size of the free money net and in-game virtual economy in a large-scale MMORPG in terms of the cash flow.

With the advantage of practical way to experiment with new network protocols in realistic settings, NOX and OpenFlow switch networks are becoming extremely popular. However, because of basic characteristics of NOX and OpenFlow switch based on the separation between control and data plane, every OpenFlow switch faces a long transmission and retransmission delay when it fails to transmit its data. Until now, the virtualized programmable networks only consider how to achieve the throughput for the direct link between OpenFlow switches. Since wireless channel experiences different conditions and NOX and OpenFlow switch networks supports the maximum flow size threshold, the aggregated flow size of a neighbor OpenFlow switch may be delivered faster than through the direct link if the neighbor link has higher RSS (Received Signal Strength). In this paper, we propose a flow-based hybrid mechanism to improve performance in NOX and wireless OpenFlow switch networks. The main idea of this scheme is that when the transmission of a OpenFlow switch fails, one of neighbor OpenFlow switches with better channel condition transmits the lost frame as well as the own data using flow aggregation scheme. To do so, every OpenFlow switch should manage overhear table to buffer the transmitted packets that is not yet acknowledged. We also present algorithms to retransmit lost packets, to maintain the overhear table and to compensate for the retransmission of packets of other OpenFlow switches. Simulation results show that the proposed flow-based hybrid mechanism can significantly improve the system throughput and the throughput gain.

Network coding is a promising technology that increases the system throughput via reducing the number of transmissions for the packets delivered from the source node to the destination node in the saturated traffic scenario. Nevertheless, some packets can suffer from the metric of end-to-end delay. Since it takes the queuing delay in the intermediate node to wait for other packets to be encoded with (XOR). Therefore, in this paper, we analyze the delay according to the packet arrival rate and propose a new network coding scheme, iXOR (Intelligent XOR). It reduces the average delay even unsaturated traffic load through the Holding-χ strategy. Through an analysis and extensive simulations, we show that iXOR is better than the general forwarding scheme (FWD) without XOR and XOR without the holding-χ strategy, χ=0, in aspect of the average delay as well as the delivery ratio.

Recent research on Internet traffic classification has yield a number of data mining techniques for distinguishing types of traffic, but no systematic analysis on "Why" some algorithms achieve high accuracies. In pursuit of empirically grounded answers to the "Why" question, which is critical in understanding and establishing a scientific ground for traffic classification research, this paper reveals the three sources of the discriminative power in classifying the Internet application traffic: (i) ports, (ii) the sizes of the first one-two (for UDP flows) or four-five (for TCP flows) packets, and (iii) discretization of those features. We find that C4.5 performs the best under any circumstances, as well as the reason why; because the algorithm discretizes input features during classification operations. We also find that the entropy-based Minimum Description Length discretization on ports and packet size features substantially improve the classification accuracy of every machine learning algorithm tested (by as much as 59.8%!) and make all of them achieve >93% accuracy on average without any algorithm-specific tuning processes. Our results indicate that dealing with the ports and packet size features as discrete nominal intervals, not as continuous numbers, is the essential basis for accurate traffic classification (i.e., the features should be discretized first), regardless of classification algorithms to use.

With the rapidly increasing demand of traffic applications, the need to support seamless multimedia services in the Vehicular Wireless Networks and Vehicular Intelligent Transportation Systems (V-WINET/V-ITS) is growing. Several mobility support protocols such as the Mobile IPv6 (MIPv6) and the fast handover for the MIPv6 (FMIPv6) have been developed to support seamless handover. However, MIPv6 depreciates Quality-of-Service (QoS) especially for multimedia service applications due to the long handover latency and the packet loss problem. FMIPv6 tries to solve these problems of MIPv6 through handover prediction but the high speed and sudden direction change of vehicles make predictions inaccurate. In this paper, we propose a seamless and robust handover scheme that supports multimedia services in V-WINET/V-ITS. Unlike MIPv6 or FMIPv6 where a new Care-of-Address (nCoA) has to be configured every time when a vehicle meets a new AR (nAR), the proposed scheme continuously maintains the original CoA (oCoA) configured at original Access Router (oAR) and reduces the handover delay caused by the Duplicate Address Detection (DAD). While a vehicle maintains its oCoA, the data packet destined to the vehicle is forwarded from the oAR to the nAR, and finally to the vehicle. At the intersection, the vehicle creates a nCoA to limit the packet forwarding hops between the oAR and the nAR. However, our background DAD scheme reduces the DAD delay at the intersection and also reduces the number of Home Agent (HA) binding updates. Through extensive simulations, we show that the proposed scheme significantly reduces the average handover latency by up to 40%.

An auto-mated network management has been not only critical but also difficult in the network research area. Among the artificial intelligent techniques, traditional supervised learning techniques are not appropriate for an auto-mated network management and specially to detect temporal changes in network intrusion patterns and characteristics. The reason is that supervised learning needs the manager. Therefore, unsupervised learning techniques such as SOM (self-organizing map) are more appropriate for an auto-mated network management such as configuration, performance and anomaly detection. In this paper, we propose an auto-mated network management based on hierarchical SOM that groups similar data and visualize their clusters. Our system labels the map produced by SOM using correlations between features for an auto-mated network management. We experiments our system with KDD Cup 1999 data set. Our system yields the reasonable misclassification rates and takes 0.5 seconds to decide whether a behavior is normal or attack.

Recent works in empirical 802.11 wireless LAN performance evaluation have shown that cross-layer interactions in WLANs can be subtle, sometimes leading to unexpected results. Two such instances are: (i) significant throughput degradation resulting from automatic rate fallback (ARF) having difficulty distinguishing collision from channel noise, and (ii) scalable TCP over DCF performance that is able to mitigate the negative performance effect of ARF by curbing multiple access contention even when the number of stations is large. In this paper, we present a framework for analyzing complex cross-layer interactions in 802.11 WLANs, with the aim of providing effective tools for understanding and improving WLAN performance. We focus on cross-layer interactions between ARF, DCF, and TCP, where ARF adjusts coding at the physical layer, DCF mediates link layer multiple access control, and TCP performs end-to-end transport. We advance station-centric Markov chain models of ARF, ARF-DCF with and without RTS/CTS, and TCP over DCF that may be viewed as multi-protocol extensions of Bianchi's IEEE 802.11 model. We show that despite significant increase in complexity the analysis framework leads to tractable and accurate performance predictions. Our results complement empirical and simulation-based findings, demonstrating the versatility and efficacy of station-centric Markov chain analysis for capturing cross-layer WLAN dynamics.

In this paper, we consider how to exploit multiuser diversity in the uplink of IEEE 802.11 WLAN when its uplink and downlink are asymmetric. In the uplink, there is no central node that arranges transmission schedules of all stations. To overcome this limitation, we devise a novel MAC protocol, called WLAN Opportunistic Waiver (WOW), where an access point (AP) indirectly controls transmission instants of stations by not sending a CTS frame if the received signal strength (RSS) of an RTS frame is below a certain threshold that is station- dependent. We develop an analytic model for WOW with a three- dimensional Markov chain. We can find the optimal threshold to maximize the system throughput. To avoid optimization burdens, we propose a simple algorithm which can determine a near- optimal threshold. Both analysis and ns-2 simulation results show that the throughput of WOW increases with the number of stations and the improvement compared with Receiver Based Auto Rate (RBAR) is up to 43 % while maintaining access fairness as achieved by IEEE 802.11.

One of the key challenges in designing a rate adaptation scheme for IEEE 802.11 wireless LANs (WLANs) is to differentiate bit errors from link-layer collisions. Many recent rate adaptation schemes adopt the RTS/CTS mechanism to prevent collision losses from triggering unnecessary rate decrease. However, the RTS/CTS handshake incurs significant overhead and is rarely activated in today's infrastructure WLANs. In this paper we propose a new rate adaptation scheme that mitigates the collision effect on the operation of rate adaptation. In contrast to previous approaches adopting fixed rate-increasing and decreasing thresholds, our scheme varies threshold values based on the measured network status. Using the "retry" information in 802.11 MAC headers as feedback, we enable the transmitter to gauge current network state. The proposed rate adaptation scheme does not require additional probing overhead incurred by RTS/CTS exchanges and can be easily deployed without changes in firmware. We demonstrate the effectiveness of our solution by comparing with existing approaches through extensive simulations.

Abstract — Rate adaptation is one of the basic functionalities in today’s 802.11 wireless LANs (WLANs). Although it is primarily designed to cope with the variability of wireless channels and achieve higher system spectral efficiency, its design needs careful consideration of cross-layer dependencies, in particular, linklayer collisions. Most practical rate adaptations focus on the time-varying characteristics of wireless channels, ignoring the impact of link-layer collisions. As a result, they may lose their effectiveness due to unnecessary rate downshift wrongly triggered by the collisions. Some recently proposed rate adaptations use RTS/CTS to suppress the collision effect by differentiating collisions from channel errors. The RTS/CTS handshake, however, incurs significant overhead and is rarely activated in infrastructure WLANs. In this paper, we introduce a new approach for optimizing the operation of rate adaptations by adjusting the rate-increasing and decreasing parameters based on link-layer measurement. To construct the algorithm, we study the impact of rate-increasing and decreasing thresholds on performance and show that dynamic adjustment of thresholds is an effective way to mitigate the collision effect in multi-user environments. Our method does not require additional probing overhead incurred by RTS/CTS exchanges and may be practically deployed without change in firmware. We demonstrate the effectiveness of our solution, comparing with existing approaches through extensive simulations. I.

Wireless Internet access is facilitated by IEEE 802.11 WLANs that, in addition to realizing a specific form of CSMA/CA-distributed coordination function (DCF)- implement a range of performance enhancement features such as multi-rate adaptation that induce cross-layer protocol coupling. Recent works in empirical WLAN performance evaluation have shown that cross-layer interactions can be subtle, sometimes leading to unexpected outcomes. Two such instances are: significant throughput degradation (a bell-shaped throughput curve) resulting from automatic rate fallback (ARF) having difficulty distinguishing collision from channel noise, and scalable TCP performance over DCF that is able to curtail effective multiple access contention in the presence of many contending stations. The latter also mitigates the negative performance effect of ARF. In this paper, we present station-centric Markov chain models of WLAN cross-layer performance aimed at capturing complex interactions between ARF, DCF, and TCP. Our performance analyses may be viewed as multi-protocol extensions of Bianchi's IEEE 802.11 model that, despite significantly increased complexity, lead to tractable and accurate performance predictions due to modular coupling. Our results complement empirical and simulation-based findings, demonstrating the versatility and efficacy of station-centric Markov chain analysis for capturing cross-layer WLAN dynamics.

Widespread deployment of infrastructure WLANs has made Wi-Fi an integral part of today's Internet access technology. Despite its crucial role in affecting end-to-end performance, past research has focused on MAC protocol enhancement, analysis, and simulation-based performance evaluation without sufficient consideration for modeling inaccuracies stemming from interlayer dependencies, including physical layer diversity, that significantly impact performance. We take a fresh look at IEEE 802.11 WLANs and using experiment, simulation, and analysis demonstrate its surprisingly agile performance traits. Our findings are two-fold. First, contention-based MAC throughput degrades gracefully under congested conditions, enabled by physical layer channel diversity that reduces the effective level of MAC contention. In contrast, fairness degrades and jitter increases significantly at a critical offered load. This duality obviates the need for link layer flow control for throughput improvement. Second, TCP-over-WLAN achieves high throughput commensurate with that of wireline TCP under saturated conditions, challenging the widely held perception that TCP throughput fares poorly over WLANs when subject to heavy contention. We show that TCP-over-WLAN prowess is facilitated by the self-regulating actions of DCF and TCP feedback control that jointly drive the shared channel at an effective load of two to three wireless stations, even when the number of active stations is large. We show that the mitigating influence of TCP extends to unfairness and adverse impact of dynamic rate shifting under multiple access contention. We use experimentation and simulation in a complementary fashion, pointing out performance characteristics where they agree and differ.

An entire network can be managed as a single mobility entity when it moves as a unit. To support network mobility (NEMO), a mobile router has been introduced to manage the mobility of whole nodes inside the network. In this mobile network, multiple mobile router (MR)s and home agent (HA)s scenarios are considered to provide reliability and load sharing. In this paper, we present a neighbor MR authentication and registration mechanism in multihomed mobile networks. Also, using registered MRs, we propose a HA-based dynamic load sharing mechanism. Using measured latency from periodic binding update (BU) messages, the HA shares traffic load with an alternative tunnel. Our proposed mechanism requires no additional signaling messages except some options in the BU message.

In this paper, we first propose an analytical model of WLANs (wireless LANs) with an arbitrary backoff distribution and AIFS (arbitration inter-frame space). From the analysis, we show that the achievable bandwidth is determined by the mean of backoff distribution regardless of the shape of the backoff distribution. We compare the effectiveness of four parameters on channel access differentiation, namely, the mean of backoff distribution, CW/sub min/ (initial contention window), the number of backoff stages, and AIFS. Numerical results show that the mean of backoff distribution provides weighted fair channel access most accurately. Second, based on the proposed analytic frame work, we develop three schemes for the uplink/downlink bandwidth differentiation in order to achieve weighted fairness between uplink and downlink transmissions. Note that IEEE 802.11 is known to have unfairness between uplink and downlink accesses. Each scheme is characterized according to the corresponding channel access rule. The simulation results show that the proposed schemes achieve high system throughput while accurately differentiating bandwidth allocation.

Wide-spread deployment of infrastructure WLANs has made Wi-Fi an integral part of today's Internet access technology. Despite its crucial role in affecting end-to-end performance, past research has focused on MAC protocol enhancement, analysis and simulation-based performance evaluation without sufficient consideration for modeling inaccuracies stemming from inter-layer dependencies, including physical layer diversity, that significantly impact performance. We take a fresh look at IEEE 802.11 WLANs, and using a combination of experiment, simulation, and analysis demonstrate its surprisingly agile performance traits. Our main findings are two-fold. First, contention-based MAC throughput degrades gracefully under congested conditions, enabled by physical layer channel diversity that reduces the effective level of MAC contention. In contrast, fairness and jitter significantly degrade at a critical offered load. This duality obviates the need for link layer flow control for throughput improvement but necessitates traffic control for fairness and QoS. Second, TCP-over-WLAN achieves high throughput commensurate with that of wireline TCP under saturated conditions, challenging the widely held perception that TCP throughput fares poorly over WLANs when subject to heavy contention. We show that TCP-over-WLAN prowess is facilitated by the self-regulating actions of DCF and TCP congestion control that jointly drive the shared physical channel at an effective load of 2--3 wireless stations, even when the number of active stations is very large. Our results highlight subtle inter-layer dependencies including the mitigating influence of TCP-over-WLAN on dynamic rate shifting.

The IEEE 802.11 standard for wireless local area networks (WLANs) employs a medium access control (MAC), called distributed coordination function (DCF), which is based on carrier sense multiple access with collision avoidance (CSMA/CA). The collision avoidance mechanism utilizes the random backoff prior to each frame transmission attempt. The random nature of the backoff reduces the collision probability, but cannot completely eliminate collisions. It is known that the throughput performance of the 802.11 WLAN is significantly compromised as the number of stations increases. In this paper, we propose a novel distributed reservation-based MAC protocol, called early backoff announcement (EBA), which is backward compatible with the legacy DCF. Under EBA, a station announces its future backoff information in terms of the number of backoff slots via the MAC header of its frame being transmitted. All the stations receiving the information avoid collisions by excluding the same backoff duration when selecting their future backoff value. Through extensive simulations, EBA is found to achieve a significant increase in the throughput performance as well as a higher degree of fairness compared to the 802.11 DCF.

Abstract Recently, the IEEE 802.11 working group has announced a new distributed access mechanism called Enhanced DCF (EDCF) to provide service differentiation among traffic classes defined as access category. With the increasing demand for supporting Quality of Service (QoS) in IEEE 802.11 wireless LANs, the EDCF is now attracting many researchers' attention due to its practical worth as a standard mechanism. In this paper, we focus on the analytical approach to evaluate the performance of the EDCF. An analytical model is presented to estimate the throughput of the EDCF in saturation (asymptotic) conditions by substantially revising and extending the analytical model developed by Bianchi for the performance analysis of the DCF. Extensive simulation studies based on the NS‐2 simulator have been carried out for the validation of the analysis, and they show that it estimates the throughput of the EDCF accurately. By utilizing the analytical model, we evaluate the performance of the EDCF. Specifically, we concentrate on discovering the characteristics of the EDCF parameters, such as CWmin, CWmax and AIFS, in the way that they influence on the performance of the EDCF. Copyright © 2004 John Wiley & Sons, Ltd.

An ad hoc network is a multi-hop wireless network formed by a collection of mobile nodes without the intervention of fixed infrastructure. Limited bandwidth and a high degree of mobility require that routing protocols for ad hoc networks be robust, simple, and energy-conserving. This paper proposes a new ad hoc multicast routing protocol called neighbor-supporting multicast protocol (NSMP). NSMP adopts a mesh structure to enhance resilience against mobility. NSMP utilizes node locality to reduce the overhead of route failure recovery and mesh maintenance. NSMP also attempts to improve route efficiency and reduce data transmissions. Our simulation results show that NSMP delivers packets efficiently while substantially reducing control overhead in various environments.

In an ad hoc network, each host assumes the role of a router and relays packets toward final destinations. This paper studies efficient routing mechanisms for multicast and broadcast in ad hoc wireless networks. Because a packet is broadcast to all neighboring nodes, the optimality criteria of wireless network routing is different from that of wired network routing. In this paper, we point out that the number of packet forwarding is the more important cost factor than the number of links in the ad hoc network. After we show constructing minimum cost multicast tree is hard, we propose two new flooding methods, self pruning and dominant pruning. Both methods utilize neighbor information to reduce redundant transmissions. Performance analysis shows that both methods perform significantly better than blind flooding. Especially, dominant pruning performs close to the practically achievable best performance limit.

A multirate multicast switch, which can provide a single uniform switching function for vastly different classes of traffic, is crucial for the successful deployment of integrated broadband networks. We analyze the call blocking probability of heterogeneous circuit switched traffic in a multirate multicast switch using the arrival modulation technique. For the analysis, we introduce two simple traffic models: fan-out heterogeneous traffic and bandwidth heterogeneous traffic. The analysis of two simplified traffic models shows that calls of different characteristics interact in complex yet subtle ways. Our results on fan-out heterogeneous traffic show that the performance gap between large fan-out traffic and small fan-out traffic increases when two types of traffic are mixed. On the other hand, in bandwidth heterogeneous traffic, the performance gap between traffic of different bandwidth decreases. A further investigation indicates that the two types of interactions coexist in general heterogeneous traffic. We conclude the paper introducing several methods to improve the performance of multirate multicast traffic.

A multicast switching architecture called a duplex multicast switch is proposed, and several switch control algorithms are developed. A duplex multicast switch, with two point-to-point routing nets, has potential to provide significantly better performance than a simplex multicast switch by reducing the output loadings of routing nets. To fully realize its potential, two call distribution algorithms, cluster distribution and spread distribution, are developed. Cluster distribution is partitioned into partial search cluster distribution and exhaustive search cluster distribution based on search policies, and the performance of the three algorithms is analyzed by the arrival modulation technique. The results show that a spread distribution eliminates most slot contention blocking and achieves near-optimal performance.>

Multicast switching is emerging as a new switching technology that can provide efficient transport in a broadband network for video and other multipoint communication services. The authors develop and analyze call scheduling algorithms for a multicast switch. In particular, they examine two general classes of scheduling algorithms: call packing algorithms and call splitting algorithms. The performance improvement by the call packing algorithms examined is shown to be negligible. In contrast, the call splitting algorithms can provide significantly lower blocking by reducing the level of output port contention. However, excessive call splitting could degrade performance because of the additional load introduced to the input ports. The authors present a simple call splitting algorithm called greedy splitting which achieves near-optimal performance.>

Multicast traffic encounters higher blocking probability than point-to-point traffic because of simultaneous output port contentions. To ensure adequate performance for multicast traffic, the authors develop and analyze a class of call scheduling algorithms via call splitting. Call splitting algorithms reduce output contention by generating smaller subcalls from a multicast call. It is shown that slot contention blocking is the predominant factor of blocking a multicast call, which suggests that call splitting may be an efficient strategy for a multicast call. The authors devise a deterministic call splitting algorithm and show that excessive call splitting can degrade performance because of the additional load introduced to the input ports. The authors also investigate an adaptive splitting algorithm which achieves performance approach the optimum by avoiding excessive call splitting.>

Repeated studies have shown that a single switching technique, either circuit or packet switching, cannot optimally support a heterogeneous traffic mix composed of voice, video and data. Integrated networks support such heterogeneous traffic by combining circuit and packet switching in a single network. To manage the statistical variations of network traffic, we introduce a new, adaptive routing algorithm called hybrid, weighted routing. Simulations show that hybrid, weighted routing is preferable to other adaptive routing techniques for both packet switched networks and integrated networks.