Open-domain visual entity recognition aims to identify and link entities depicted in images to a vast and evolving set of real-world concepts, such as those found in Wikidata. Unlike conventional classification tasks with fixed label sets, it operates under open-set conditions, where most target entities are unseen during training and exhibit long-tail distributions. This makes the task inherently challenging due to limited supervision, high visual ambiguity, and the need for semantic disambiguation. We propose a Knowledge-guided Contrastive Learning (KnowCoL) framework that combines both images and text descriptions into a shared semantic space grounded by structured information from Wikidata. By abstracting visual and textual inputs to a conceptual level, the model leverages entity descriptions, type hierarchies, and relational context to support zero-shot entity recognition. We evaluate our approach on the OVEN benchmark, a large-scale open-domain visual recognition dataset with Wikidata IDs as the label space. Our experiments show that using visual, textual, and structured knowledge greatly improves accuracy, especially for rare and unseen entities. Our smallest model improves the accuracy on unseen entities by 10.5% compared to the state-of-the-art, despite being 35 times smaller.

Precise optical inspection in industrial applications is crucial for minimizing scrap rates and reducing the associated costs. Besides merely detecting if a product is anomalous or not, it is crucial to know the distinct type of defect, such as a bent, cut, or scratch. The ability to recognize the "exact" defect type enables automated treatments of the anomalies in modern production lines. Current methods are limited to solely detecting whether a product is defective or not without providing any insights on the defect type, nevertheless detecting and identifying multiple defects. We propose MultiADS, a zero-shot learning approach, able to perform Multi-type Anomaly Detection and Segmentation. The architecture of MultiADS comprises CLIP and extra linear layers to align the visual- and textual representation in a joint feature space. To the best of our knowledge, our proposal, is the first approach to perform a multi-type anomaly segmentation task in zero-shot learning. Contrary to the other baselines, our approach i) generates specific anomaly masks for each distinct defect type, ii) learns to distinguish defect types, and iii) simultaneously identifies multiple defect types present in an anomalous product. Additionally, our approach outperforms zero/few-shot learning SoTA methods on image-level and pixel-level anomaly detection and segmentation tasks on five commonly used datasets: MVTec-AD, Visa, MPDD, MAD and Real-IAD.

Knowledge graphs (KGs) have transformed data management within the manufacturing industry, offering effective means for integrating disparate data sources through shared and structured conceptual schemas. However, harnessing the power of KGs can be daunting for non-experts, as it often requires formulating complex SPARQL queries to retrieve specific information. With the advent of Large Language Models (LLMs), there is a growing potential to automatically translate natural language queries into the SPARQL format, thus bridging the gap between user-friendly interfaces and the sophisticated architecture of KGs. The challenge remains in adequately informing LLMs about the relevant context and structure of domain-specific KGs, e.g., in manufacturing, to improve the accuracy of generated queries. In this paper, we evaluate multiple strategies that use LLMs as mediators to facilitate information retrieval from KGs. We focus on the manufacturing domain, particularly on the Bosch Line Information System KG and the I40 Core Information Model. In our evaluation, we compare various approaches for feeding relevant context from the KG to the LLM and analyze their proficiency in transforming real-world questions into SPARQL queries. Our findings show that LLMs can significantly improve their performance on generating correct and complete queries when provided only the adequate context of the KG schema. Such context-aware prompting techniques help LLMs to focus on the relevant parts of the ontology and reduce the risk of hallucination. We anticipate that the proposed techniques help LLMs to democratize access to complex data repositories and empower informed decision-making in manufacturing settings.

Existing mobile network infrastructures, for example the Evolved Packet Core (EPC), are based on mission specific, specialized network nodes/entities that are interconnect via standardized interfaces in a static manner, each providing a specific service/function. One common practice adopted by many Mobile Network Operators (MNOs) in order to cater to increasing traffic demand is the over provisioning of network and processing resources. However, this approach results in increased CAPEX and OPEX and in view of future load forecasts this approach is no longer considered feasible or economical. In this paper we address this specific issue and propose the concept of softEPC, which leverages the recent advances in cloud technology concept such as Infrastructure as a Service (IaaS) and virtualization techniques in the realm of mobile networks. A softEPC is considered as a virtual network of Evolved Packet Core (EPC) functions over a physical transport network topology. This will enable the on-demand and load-aware dynamic instantiation of network functions and services at appropriate locations in response to the actual traffic demand. The main objective of this approach is to increase the utilization of network resources by flexibly and dynamically placing the network functions/services where most appropriate to provide optimum service and where resources are available to increase the number of services provided to mobile users. The benefits and gains of this approach will be demonstrated by means of simulation results.

Mobile carriers are currently facing the tremendous challenge of coping with the rapid data traffic growth stemming from smart phones and data dongles. Recent studies have shown that caching of high volume content within carrier networks has the potential of substantial cost savings in network resources. This paper analyzes the effectiveness of caching in mobile carrier networks, being enabled by distributed mobility gateways and content caches, and estimates how much cost savings an operator can expect for different cache deployment scenarios. The study considers various backhaul network characteristics and several content replication strategies.

In this paper, we consider the Selected IP Traffic Offload (SIPTO) approach to handle increased data traffic of both local and macro-cellular networks. We devise different ap- proaches based on operator defined offload policies on a per des- tination domain name basis, which offer operators fine-grained control of whether a new IP connection should be offloaded or provided via the core network. Two of our solutions are based on Network Address Translation (NAT) named simple-NATing and twice-NATing, while a third one employs simple tunneling. These solutions support all kinds of UEs including those that support a single Packet Data Protocol (PDP) context/Packet Data Network (PDN) connection. In a forth solution, we consider the case where a User Equipment (UE) supports multiple PDP contexts/PDN connections, with at least one dedicated for SIPTO traffic. A qualitative analysis and a simulation study are presented.

Key incentives for mobile operators to offer femtocell services are the possibilities to cost-efficiently enhance indoor coverage and to offload traffic from the macro-cellular network. While femtocells were originally conceived for standalone deployments in user homes, there is an increasing interest in exploiting their advantages also in an enterprise context. However, with the current architecture as specified by 3GPP, such deployments face technical and economical challenges. This paper proposes a solution for localizing data and control plane traffic inside "Enterprise Femtocell Networks", which significantly offloads both the mobile core network and the backhaul to the core network.

Wireless communication technologies such as microwave radios are used to provide high-speed mobile backhaul connectivity for radio access networks in cases in which wire-based alternatives, e.g. cable or fiber, are not readily available and cannot be deployed in an economic or timely manner. Current mobile backhauls are predominantly deployed in tree or ring topologies, which simplify traffic management. Yet, with the increasing demand on backhaul capacity and the immense cost pressure on mobile backhaul solutions, meshed wireless mobile backhauls have been identified as a promising evolution. While traffic management in wireless mesh networks have been studied extensively in the literature, so far there is no quantitative analysis comparing the different topology options, i.e. mesh, ring and tree, regarding network performance and deployment cost. This paper fills this gap by studying the minimum cost problem of connecting a set of base station/gateway sites using different topologies while supporting both time- and space-varying traffic demands. Furthermore, we consider the additional constraint of resilience to single link failures. The evaluation results show that meshed wireless backhaul topologies are a cost-effective alternative to trees and rings, in particular in the face of spatial and temporal fluctuation of traffic demand and protection against link failures.

In the System Architecture Evolution (SAE) study of the nextgeneration mobile network in 3GPP, Serving Gateways (SGWs) and Mobility Management Entities (MMEs) are grouped to form a number of service and pool areas, respectively. While this concept of SGW service areas in the Evolved Packet Core is interesting to limit the administrative scope of SGWs and also provides a means to optimize the routing, the use of fixed/hard area boundaries can result in frequent unnecessary SGW relocations and can severely impact the Quality of Experience (QoE) of users. To avoid the drawback of fix/hard service (or pool) area boundaries, this paper proposes a scheme whereby every SGW can have a flexibly configurable service area, which is defined by a set of LTE (Long Term Evolution) cells or Tracking Areas (TAs). The service area of a SGW defines the LTE area (e.g., cells or TAs) that the SGW can serve. The working of the proposed mechanism is validated via computer simulations and encouraging results are obtained.

Available doorbell interphone systems are designed under the assumption that residents would be at home to enable the communication between them and their visitors. However, people spend a large fraction of their time away from home, thus undermining the basic assumptions of existing doorbell solutions. Recent developments in 3GPP Home Node B (HNB or Femtocell) and home gateway technologies, along with the growing proliferation of smart-phones, can offer interesting opportunities for the design and development of innovative doorbell solutions. Along this line, the paper describes a 3GPP-enhanced eDoorbell application prototype that relies on IMS to enable video-based communication between visitors and (possibly remote available) home residents. In our prototype, the visibility of context information, such as current location of residents, their diary, and their personal preferences, provides solid basis for definition and enactment of customizable management policies that determine the best suited home resident whom to route the notification of a visit to. While the paper presents no experimental results, it aims at assisting organizations such as the Femto Forum or the ETSI TISPAN to identify the requirements for standards and the different methods used to implement HNB-based or Home Gateway-based services, respectively.

The objective of autonomic networking is to enable the autonomous formation and parametrization of nodes and networks by letting protocols sense and adapt to the networking environment at run time. Besides its dynamic aspects, a core requirement of autonomic networking is to define a structured framework and execution environment that enables algorithms to operate in a continously changing environment. This paper presents the major design principles of the Autonomic Network Architecture (ANA) and reports on a first implementation. The guiding principle of ANA is to strive for flexibility and genericity at all levels of the architecture. In our approach we explicitly avoid to impose a "one-size-fits-all" architecture (where communication protocols and paradigms are fixed by the architecture). To this end, ANA introduces generic abstractions, for example "information dispatch points" instead of addressable endpoints, as well as communication primitives that support network heterogeneity, adaptability, and evolution. These core abstractions allow for the coexistance of multiple and diverse networking styles and protocols. With the public release of the ANA prototype, we aim at federating autonomics related networking projects, enabling different actors to share, compare, and build upon each other¿s work. The ANA runtime can host clean slate network designs as well as legacy Internet technology and serves as a platform for demonstrating autonomic communication principles.

From the beginning of data networking, dynamic routing has been a challenge. Due to the constantly increasing number of devices and the introduction of multi-hop wireless networks, dynamic routing will remain an important issue for any future network architecture. Using more elaborate routing metrics for such environments is regarded as the general solution but neglects the introduction of significant overhead required to exchange the information. In this paper, we present a routing scheme which is based on metric dependent neighborhood to calculate the forwarding graph. The proposed routing scheme supports aggregation of this metric related information while disseminating routing updates to retain scalability. Simulation results with an exemplary metric based on link stability information show the feasibility of this aggregation approach and the improvement with respect to node reachability and reliable communication in self-organizing wireless networks. Moreover, we implemented this routing scheme on an autonomic networking architecture (ANA).

The recent developments in the fixed and mobile telecommunication industry, namely the rapid penetration of broadband access and new charging models in the broadband market in conjunction with the success of wireless local area networks and the general trend of fixed and mobile convergence, are transforming the telecommunication landscape. One of the main challenges is to integrate the fixed and mobile technologies in a way that leverages the strengths of different technologies and improves the service experience for the end user. A crucial step for this transformation is to provide adequate support for users to seamlessly switch between network domains and technologies. The main contribution of this paper is a solution that allows a mobile terminal to change from the mobile radio access to the fixed/cable access provided over a local wireless network or vice versa without disrupting an active ongoing voice call. The implementation of the proposed solution was tested and evaluated in an experimental network with real voice calls.

TurfNet is an internetworking architecture that enables communication among autonomous and heterogeneous network domains. The architecture uses a global identity namespace and does not require global addressing or a common internetworking protocol. It integrates the new concept of dynamic network composition with other recent architectural concepts, such as decoupling locators from identifiers. This paper analyzes whether the TurfNet naming and interdomain routing mechanisms can scale to networks of the size of the global Internet. To the authors' knowledge, this is one of the first scalability analyses of any next-generation internetworking proposal based on today's Internet AS topology. It adapts existing research results on the topology of the Internet's autonomous system graph and related results that quantify typical traffic patterns to construct a realistic model of an Internet-like TurfNet.

The ANA (Autonomic Network Architecture) project aims at providing a framework to flexibly host, interconnect, and federate multiple heterogeneous networks in an autonomic way, i.e. without requiring active human intervention. The guiding design principle is to strive for a maximum degree of flexibility at all levels of the architecture in order to inherently support heterogeneity and evolution. This paper describes the core abstractions and concepts of ANA (as defined during the first year of the project) and introduces their basic operation and interaction. While not autonomic themselves, the core architectural principles of ANA will enable autonomicity by not imposing a "one-size-fits-all" approach where protocols and paradigms are fixed by the architecture. We indeed argue that only the capacity of the network to be polyfunctional and fully adaptable will justify the label ' autonomic'. A prototype ofANA is currently being developed and will be released in 2008: the goal is to demonstrate the feasibility of autonomic networking within the 4 years of the project.

In this article, we apply autonomic and distributed management principles to wireless access networks. Most interesting is the application of autonomic properties and behaviors including adaptive, aware, and automatic operation in a decentralized setting. In particular, we present a generic and autonomic management architecture for decentralized management of wireless access networks, such as GERAN/UTRAN, E-UTRAN, WiMAX or WLAN. For evaluation purposes, we apply this architecture to the management of a Wireless LAN network, and we evaluate the architecture and some of the autonomic management functions through simulations, a prototype implementation and the setup of a real-world testbed for experimentation with the proposed management approach.

The Internet is currently evolving beyond what its architecture can support. Often, the mechanisms that allow the Internet to adapt to increasingly conflicting sets of new requirements break some of its basic design principles and can thus severely interfere with end-toend communication. This paper recognizes that increased autonomy of network regions is a key requirement for future internetworking. It outlines a new internetworking architecture that enables interoperation among a set of autonomous, heterogeneous network domains. The architecture is based on a global identity space and does not require global addressing or a shared internetworking protocol. It integrates the new concept of dynamic network composition with other recent architectural concepts, such as decoupling locators from identifiers. 1.

This paper proposes a novel mechanism to enhance data protection in communications across untrusted networks. The approach is based on the principle of network address hopping, whereby a data stream is spread across multiple end-to-end connections. The aim is to obscure the data exchange between two peers by shuffling the communication pattern. The hopping pattern - a shared secret between the communication participants - defines the sequence for the address hopping and determines the data spreading. Besides a description of the basic operation of the network address hopping mechanism, the paper evaluates the level of protection it can offer for end-to-end communications. This theoretical analysis is accompanied by a quantitative evaluation of the processing overhead of our prototype implementation on commodity end hosts.

The Internet is currently evolving beyond what its architecture can support. Often, the mechanisms that allow the Internet to adapt to increasingly conflicting sets of new requirements break some of its basic design principles and can thus severely interfere with end-to-end communication. This paper recognizes that increased autonomy of network regions is a key requirement for future internetworking. It outlines a new internetworking architecture that enables interoperation among a set of autonomous, heterogeneous network domains. The architecture is based on a global identity space and does not require global addressing or a shared internetworking protocol. It integrates the new concept of dynamic network composition with other recent architectural concepts, such as decoupling locators from identifiers.

With the increasing heterogeneity of Internet enabled devices and Internet access technologies, media content may often need to be adapted to best meet the user or application needs. However, server-side adaptation has shown its limitations, leading to the development of network-side adaptation techniques. Overlay networks have emerged as a possible solution to enable data processing between users and the content server, as it allows to transparently include media processing nodes into the end-to-end media path. Nevertheless, to deal with multimedia streams in an efficient way, current proposals lack application awareness when k comes to routing decisions. Typically overlay networks only consider IP header information when routing packets. In this paper we introduce a new concept that shall enable media aware routing via the use of overlay networks. The proposed routing approach considers service and application information to improve routing decisions based on the user/application requirements. In our proposal, service-specific overlay networks (SSONs) are created for every service, which allows customisation of the network resources. SSONs include overlay nodes with adaptation capability. The data routing in the SSON is performed using the media aware overlay routing logic. This paper presents the concept, along with a proposed architecture designed to provide a media aware routing service. The advantage of the proposed functionality is illustrated based on a mobile user scenario.

It is becoming increasingly apparent that the IETF mobile IPv6 protocol will have a role to play in future mobile and wireless networks. However, it is extremely likely that even after a real world deployment of IPv6, a considerable number of 'legacy' IPv4 applications and devices will remain in regular use for many years to come. Recently, much work has been focussed on developing protocols and tools to ease the transition from IPv4 to IPv6, yet despite mobile systems being one of the primary motivations for IPv6 deployment, none of these protocols and tools are designed to operate in a mobile environment. This paper documents the motivation, design, implementation and experimental evaluation of mobile 4-in-6, a mobile aware IPv4/IPv6 transitioning technology based upon the mobile IPv6 standard, that provides transparent support for IPv4 only applications in IPv6 only environments.

We believe that the 'killer application' of active networking will be the provision of services which are not known to either the hardware manufacturer or the vendor at the time of network deployment. This would enable the rapid deployment of new services by third parties, and an unprecedented level of customisation by network administrators and end users. This vision, however, requires a degree of flexibility not yet seen in any of the currently proposed active router architectures. The lack of flexibility found in current active router designs has inspired us to develop a novel component-based active router architecture enabling extensible network programmability based on a progressive service composition framework. This document presents this architecture, LARA++, discusses its rudimentary concepts, and illustrates by example how its component architecture can benefit network programmability.

This document introduces a novel access control architecture for publicly accessible wireless overlay networks. The architecture is designed to address the problems of ubiquitous Internet service provisioning within the city of Lancaster. The proposed access control mechanism is based on the concepts of secure user authentication, packet marking, and network-level packet filtering. The novelty of the architecture lies in its use of microcellular layer three networks to acquire fine grained access control in a link independent manner. The paper describes the concepts behind the access control architecture and demonstrates to what extent it addresses the security, performance and extensibility concerns of public access packet switched wireless networks.