Browse the ITC publications of previous conferences below or visit our grouppage at bibsonomy.org
2021
da Silva Coelho, Wesley and Benhamiche, Amal and Perrot, Nancy and Secci, Stefano
A Math-Heuristic for Network Slice Design
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{coe21ITC33,
title = { A Math-Heuristic for Network Slice Design },
year = { 2021 },
address = { Avignon, France },
author = { da Silva Coelho, Wesley and Benhamiche, Amal and Perrot, Nancy and Secci, Stefano },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: In this paper, we address the Network Slice Design problem, which arises from blueprinting end-to-end communication networks using fifth-generation (5G) radio access technology. With regard to new sharing policies and radio-access integration, it shows peculiar requirements with respect to conventional function placement and routing problems. To address the underlying optimization problem, we propose an open-access framework based on a math-heuristic that encompasses control-plane and data-plane separation and novel mapping and decomposition dimensions influencing the placement and interconnection of slices. Our framework also incorporates flexible functional splitting, with possibly different splitting for different slices while taking into consideration dependency factors such as varying network latency and data volume throughout the virtual access networks. Numerical results are then presented to assess the efficiency of our approach.
Tlaiss, Ziad
Anomaly root cause diagnosis from active and passive measurement analysis
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{tla21ITC33,
title = { Anomaly root cause diagnosis from active and passive measurement analysis },
year = { 2021 },
address = { Avignon, France },
author = { Tlaiss, Ziad },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-3 }
}
Abstract: Diagnosis demands a deep analysis of data to identify the root cause of an anomaly and still mostly relies on human experts. The increase of Internet traffic combined with the arrival of the encrypted protocol QUIC which invalidates many troubleshooting methods, urges to automate this process. To this effect, both domain familiarity and analysis skills are required. In this work we present our methods and strategies to detect the root cause of anomalies from active and passive network measurement and we share our plan towards an automatic root cause diagnosis. We focus on four root causes : transmission, congestion, application limited and packet delay variation, and present the building blocks of classification methods.
Helm, Max and Stubbe, Henning and Scholz, Dominik and Jaeger, Benedikt and Gallenmüller, Sebastian and Deric, Nemanja and Goshi, Endri and Harkous, Hasanin and Zhou, Zikai and Kellerer, Wolfgang and Carle, Georg
Application of Network Calculus Models on Programmable Device Behavior
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{hel21ITC33,
title = { Application of Network Calculus Models on Programmable Device Behavior },
year = { 2021 },
address = { Avignon, France },
author = { Helm, Max and Stubbe, Henning and Scholz, Dominik and Jaeger, Benedikt and Gallenmüller, Sebastian and Deric, Nemanja and Goshi, Endri and Harkous, Hasanin and Zhou, Zikai and Kellerer, Wolfgang and Carle, Georg },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: Critical applications, such as industrial control systems or remote medical applications, require highly reliable networks. A key enabler of such applications are networks that deliver the required strict performance guarantees. A prominent tool for deriving such guarantees for networks and the involved components is network calculus (NC). Device specifics may have a stark influence on model characteristics, making modeling in heterogeneous environments work-intensive. OpenFlow and P4 are two approaches that emerged from the Software-Defined Networking (SDN) community making networks more flexible and, consequentially, even harder to model.In this work, we demonstrate a novel approach that uses NC to model such SDN-based devices despite their increased complexity. Abstracting away from overall device behavior, we initially model only the fundamental building blocks of SDN devices that define network device behavior. NC provides a framework to compose different NC models into a single model, which we use to combine the building blocks into a model that describes a network device program built from these building blocks. This approach allows for modeling a maximal number of devices with a minimal amount of measurements. We apply our approach to two different SDN devices, the Zodiac FX and the NetFPGA SUME. A comparison between the prediction of our composed models and real measurements reveals a prediction error below 1 %, thereby proving the validity of our approach.
Bronzino, Francesco and Maheshwari, Sumit and Seskar, Ivan and Raychaudhuri, Dipankar
Application-Aware End-to-End Virtualization Using a Named-Object Based Network Architecture
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{bro21ITC33,
title = { Application-Aware End-to-End Virtualization Using a Named-Object Based Network Architecture },
year = { 2021 },
address = { Avignon, France },
author = { Bronzino, Francesco and Maheshwari, Sumit and Seskar, Ivan and Raychaudhuri, Dipankar },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: Network virtualization enables applications and users to access network resources in isolation on top of a shared infrastructure. Through their mechanisms, virtual networks support performance and security guarantees that would otherwise not be achievable if relying solely on existing network protocols. Unfortunately, due to the large variety of network protocols and architectures that populate today’s Internet, existing virtualization techniques have become disparate and different depending on the segment of the network they are deployed on. Thus, as no consolidation protocol exists, no solution is available to modern services and applications to coordinate the resources they employ to function. Due to the inefficiencies associated with overlay implementations they have to revert to, it becomes challenging for these services to meet strict application requirements, e.g., low latency.In this paper, we revisit years of identity based communications to propose an approach to integrate network virtualization techniques around a single framework: the named-object abstraction. The presented approach uses unique virtual network identifiers to describe and implement custom topologies and routing metrics and achieve desired QoS requirements. This technique enables the support of an integrated network virtualization with end-to-end application aware routing on top of the network infrastructure. A proof-of-concept implementation running on the ORBIT testbed confirms that the named-object architecture can achieve low VN processing and control overhead. The proposed solution achieves an average latency performance improvement of 60% in comparison to a baseline implementation without compute and network cross layer optimizations.
Salem, Tareq Si and Neglia, Giovanni and Carra, Damiano
AÇAI: Ascent Similarity Caching with Approximate Indexes
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{sal21ITC33,
title = { AÇAI: Ascent Similarity Caching with Approximate Indexes },
year = { 2021 },
address = { Avignon, France },
author = { Salem, Tareq Si and Neglia, Giovanni and Carra, Damiano },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: Similarity search is a key operation in multimedia retrieval systems and recommender systems, and it will play an important role also for future machine learning and augmented reality applications. When these systems need to serve large objects with tight delay constraints, edge servers close to the end-user can operate as similarity caches to speed up the retrieval. In this paper we present AÇAI, a new similarity caching policy which improves on the state of the art by using (i) an (approximate) index for the whole catalog to decide which objects to serve locally and which to retrieve from the remote server, and (ii) a mirror ascent algorithm to update the set of local objects with strong guarantees even when the request process does not exhibit any statistical regularity.
Biswas, Rajorshi and Wu, Jie
Efficient Switch Migration for Controller Load Balancing in Software Defined Networking : (Invited Paper)
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{bis21ITC33,
title = { Efficient Switch Migration for Controller Load Balancing in Software Defined Networking : (Invited Paper) },
year = { 2021 },
address = { Avignon, France },
author = { Biswas, Rajorshi and Wu, Jie },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: The number of multi-controller datacenters is in- creasing with the increasing size of software-defined networking (SDN) datacenters. The performance of an SDN datacenter depends largely on the delay of response from the controller. The delay of response depends on the controller load and the distance from the SDN switch. The load of a controller depends on the number of requests it receives from the switches it controls. Therefore, a good switch-controller assignment is very important for load balancing the controller and the performance of an SDN datacenter. In this paper, we consider multiple controllers and formulate problems for initial and incremental load balancing. The initial assignment process is executed at the beginning of the network deployment. After initial deployment, the incremental assignment process is executed periodically. The incremental process migrates some of the switches to another controller to improve the performance of the network. We propose greedy and clustering-based solutions for initial switch-controller assignment. We also propose a greedy solution for incremental assignment. Our proposed solutions are evaluated using both synthetic and real datasets, and the parameters are driven by experiments at a data center.
Shelbourne, Charles and Linguaglossa, Leonardo and Zhang, Tianzhu and Lipani, Aldo
Inference of virtual network functions’ state via analysis of the CPU behavior
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{she21ITC33,
title = { Inference of virtual network functions’ state via analysis of the CPU behavior },
year = { 2021 },
address = { Avignon, France },
author = { Shelbourne, Charles and Linguaglossa, Leonardo and Zhang, Tianzhu and Lipani, Aldo },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: The on-going process of softwarization of IT networks promises to reduce the operational and management costs of network infrastructures by replacing hardware middleboxes with equivalent pieces of code executed on general-purpose servers. Alongside the benefits from the operator’s perspective, new strategies to provide the network’s resources to users are arising. Following the principle of "everything as a service", multiple tenants can access the required resources – typically CPUs, NICs, or RAM – according to a Service-Level Agreement. However, tenants’ applications may require a complex and expensive measurement infrastructure to continuously monitor the network function’s state. Although the application’s specific behavior is unknown (and often opaque to the infrastructure owner), the software nature of (virtual) network functions (VNFs) may be the key to infer the behavior of the high-level functions by accessing low-level information, which is still under the control of the operating system and therefore of the infrastructure owner. As such, in the scenario of software VNFs executed on COTS servers, the underlying CPU’s behavior can be used as the sole predictor for the high-level VNF state without explicit in-network measurements: in this paper, we develop a novel methodology to infer high-level characteristics such as throughput or packet loss using CPU data instead of network measurements. Our methodology consists of (i) experimentally analyzing the behavior of a CPU that executes a VNF under different loads, (ii) extracting a correlation between the CPU footprint and the high-level application state, and (iii) use this knowledge to detect the previously mentioned network metrics. Our code and datasets are publicly available.
Song, Jiayi and Guérin, Roch and Sariowan, Henry
Minimizing network bandwidth under latency constraints: The single node case
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{son21ITC33,
title = { Minimizing network bandwidth under latency constraints: The single node case },
year = { 2021 },
address = { Avignon, France },
author = { Song, Jiayi and Guérin, Roch and Sariowan, Henry },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: Much of today’s traffic flows between datacenters over private networks. The operators of those networks have access to detailed traffic profiles with performance goals that need to be met as efficiently as possible, e.g., realizing latency guarantees with minimal network bandwidth. Of particular interest is the extent to which traffic (re)shaping can be of benefit. The paper focuses on the most basic network configuration, namely, a single link network, with extensions to more general, multi-node networks discussed in a companion paper. The main results are in the form of optimal solutions for different types of schedulers of varying complexity. They demonstrate how judicious traffic shaping can help lower complexity schedulers perform nearly as well as more complex ones.
da Silva Coelho, Wesley
Modeling and optimization of 5G network design
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{coel21ITC33,
title = { Modeling and optimization of 5G network design },
year = { 2021 },
address = { Avignon, France },
author = { da Silva Coelho, Wesley },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-3 }
}
Abstract: Every second, a large amount of digital data is transported through infinite types of devices via cellular networks worldwide, and expectations are at a greatly accelerated growth with increasingly large requests. In few years, these networks could thereby reach their maximum capacities in terms of data transmission. To face these challenges, Network Slicing has been presented as a novel virtualized infrastructure for the new generation cellular network system. This technology now not only covers application-level abstraction but also physical and switching layers virtualization, with different radio access and link communication technologies. Hence, each service provider is to be able to deploy its communication services on top of logical networks, named Network Slices, specifically tailored to its technical requirements. The Device-to-Device communication mode is another approach presented as a promising alternative to traditional communication in cellular networks. This technology allows to reuse radio resources and to decrease the end-to-end latency of local communications. Consequently, the optimization of physical resources in cellular networks becomes crucial to better deploy virtual networks. The overall objective of our research is therefore to define and study the concept of device-to-device communication and network slice design in 5G systems, and propose mathematical models and innovative algorithms to solve the underlying optimization problems.
Lassila, Pasi and Aalto, Samuli
Near-optimal uplink scheduling for age-energy tradeoff in wireless systems
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{las21ITC33,
title = { Near-optimal uplink scheduling for age-energy tradeoff in wireless systems },
year = { 2021 },
address = { Avignon, France },
author = { Lassila, Pasi and Aalto, Samuli },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: We study the optimal age-energy tradeoff arising in the context uplink scheduling of multiple heterogeneous devices that are transmitting time-sensitive update packets through the base station in a cellular network. In the model, new updates arrive stochastically at the devices and they are subject to losses when attempting to transmit. Also, associated with the update transmission are related energy costs. The problem is to develop policies for minimizing the weighted sum of the total mean age and energy costs. The Whittle index approach is applied to obtain a near-optimal policy. In particular, we prove the indexability of the problem and explicitly derive the associated Whittle indices. Our numerical experiments demonstrate the superiority of the resulting Energy-aware Whittle index policy compared with other heuristics.
Ronteix–Jacquet, Flavien
On Radio Access Network Uplink Latency and Jitter : Measurements and Analysis
In 2021 33rd International Teletraffic Congress (ITC-33). 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{ron21ITC33,
title = { On Radio Access Network Uplink Latency and Jitter : Measurements and Analysis },
year = { 2021 },
author = { Ronteix–Jacquet, Flavien },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-3 }
}
Abstract: This work studies the latency experienced by a TCP connection in an operational LTE RAN. After exhibiting the well-known downlink bufferbloat phenomenon, our experiments shed some light on the less studied RAN uplink jitter. We explain this uplink jitter by the uplink grant-based access method. These results are reproduced in a lab environment based on the OpenAirInterface software RAN, and demonstrate the importance of RAN configuration and limitations in the current LTE standard. We conclude on open issues in the 5G grant allocation process and the current grant-free access methods.
Cheirmpos, Georgios and Karaliopoulos, Merkouris and Koutsopoulos, Iordanis
Optimizing shared data plans for mobile data access
In 2021 33rd International Teletraffic Congress (ITC-33). 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{che21ITC33,
title = { Optimizing shared data plans for mobile data access },
year = { 2021 },
author = { Cheirmpos, Georgios and Karaliopoulos, Merkouris and Koutsopoulos, Iordanis },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: The demand for mobile data has been steadily increasing over the last decade, forming an ever increasing portion of the overall Internet traffic. A considerable portion of this demand is served through capped cellular data plans that charge a fixed fee for data consumption respecting the cap and a typically higher penalty rate for additional consumption. Although shared data plans have been identified as a way to better utilize capacity that is paid for but not used, they are largely restricted to closed groups (e.g., family members) or the devices of a single user.In this paper we advocate the extension of shared data plans towards more open groups of users. We take the viewpoint of a platform that seeks to recommend optimal data plans to users subscribing to it and address the two main algorithmic tasks it faces: the partitioning of users to subscription groups and the selection of data plans that maximize their cost savings. We devise three algorithms that leverage clustering techniques. One of them addresses simultaneously the two tasks, whereas the other two decompose the problem and solve the two tasks sequentially. Our evaluation results suggest that the savings in subscription charges with shared plans are significant, ranging from 20% up to 80% of what users would pay with the cost-optimal individual data plans. They also highlight properties of the three algorithms and trade-offs they present involving the achieved cost savings, the intensity of under utilization and their sensitivity to deviations from the predicted users’ data consumption.
Simon, Mathieu and Spallina, Alessandro and Dubocquet, Loïc and Araldo, Andrea
Parsimonious Edge Computing to Reduce Microservice Resource Usage
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{sim21ITC33,
title = { Parsimonious Edge Computing to Reduce Microservice Resource Usage },
year = { 2021 },
address = { Avignon, France },
author = { Simon, Mathieu and Spallina, Alessandro and Dubocquet, Loïc and Araldo, Andrea },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-3 }
}
Abstract: Cloud Computing (CC) is the most prevalent paradigm under which services are provided over the Internet. The most relevant feature for its success is its capability to promptly scale service based on user demand. When scaling, the main objective is to maximize as much as possible service performance. Moreover, resources in the Cloud are usually so abundant, that they can be assumed infinite from the service point of view: an application provider can have as many servers it wills, as long it pays for it.This model has some limitations. First, energy efficiency is not among the first criteria for scaling decisions, which has raised concerns about the environmental effects of today’s "wild" computations in the Cloud. Moreover, it is not viable for Edge Computing (EC), a paradigm in which computational resources are distributed up to the very edge of the network, i.e., co-located with base stations or access points. In edge nodes, resources are limited, which imposes different parsimonious scaling strategies to be adopted.In this work, we design a scaling strategy aimed to instantiate, parsimoniously, a number of microservices sufficient to guarantee a certain Quality of Service (QoS) target. We implement such a strategy in a Kubernetes/Docker environment. The strategy is based on a simple Proportional-Integrative-Derivative (PID) controller. In this paper we describe the system design and a preliminary performance evaluation.
Ozcan, Yigit and Rosenberg, Catherine
Performance Evaluation of Full Duplex Communications in 5G Networks: The Impact of Interference and Traffic Asymmetry
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{ozc21ITC33,
title = { Performance Evaluation of Full Duplex Communications in 5G Networks: The Impact of Interference and Traffic Asymmetry },
year = { 2021 },
address = { Avignon, France },
author = { Ozcan, Yigit and Rosenberg, Catherine },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: Full-duplex communications (FDC) is one of the enabling technologies for 5G systems and its true impact on performance needs to be evaluated by taking into account the new sources of interference inherent to this technology. We focus on the realistic scenario where FDC is only enabled at the base-stations (BS). The coexistence of uplink and downlink transmissions calls for a joint uplink and downlink scheduling problem and to take all interference into account, a multi-cell system needs to be considered. We also argue that traffic asymmetry (TA), i.e., the fact that the downlink traffic is much larger than uplink traffic, cannot be ignored. Specifically, we formulate a system-wide joint scheduling problem that incorporates all sources of interference and TA, and show how to solve it. The main engineering insights are that in a realistic setting, i) FDC does not double capacity as often mentioned; ii) TA has the largest (negative) impact on the FDC performance; iii) imperfect self interference cancellation has the second largest impact and iv) inter-base station interference, i.e., the interference caused by co-channel BSs, has the third. FDC cannot double the performance of a TDD (time division duplex) system. The gain is rarely above 50% even though it is higher in rural or heterogeneous networks.
Alam, Md Ibrahim and Anshelevich, Elliot and Kar, Koushik and Yuksel, Murat
Proportional Pricing for Efficient Traffic Equilibrium at Internet Exchange Points
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{ala21ITC33,
title = { Proportional Pricing for Efficient Traffic Equilibrium at Internet Exchange Points },
year = { 2021 },
address = { Avignon, France },
author = { Alam, Md Ibrahim and Anshelevich, Elliot and Kar, Koushik and Yuksel, Murat },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: We analyze traffic exchange between Internet Service Providers (ISPs) at an Internet Exchange Point (IXP) as a non-cooperative game with ISPs as self-interested agents. Each ISP has the choice of exchanging traffic either using the shared IXP facilities, or outside the IXP – through their transit providers or private peering. We analyze the efficiency (social cost optimality) of the traffic exchange equilibrium at the IXP taking into consideration the congestion cost experienced by the ISPs at the IXP, under a proportional pricing model where the per-unit price charged to ISPs is proportional to the aggregate level of congestion at the IXP. We obtain worst case bounds on the efficiency at traffic exchange equilibrium under two different models of the congestion cost (delay) functions. Simulations conducted using data for actual IXPs obtained from PeeringDB demonstrate that the theoretical bounds derived for social cost optimality at equilibrium (measured as the Price of Anarchy) are fairly tight, and correctly capture the performance trends against the variation of key model parameters. Further, the results show that for a certain range of the proportionality constant, proportional pricing not only results in significantly better efficiency compared to zero pricing, but also attains nearoptimal social cost and near-optimal IXP revenue simultaneously.
Huang, Jiatai and Huang, Longbo
Robust Wireless Scheduling under Arbitrary Channel Dynamics and Feedback Delay (Invited Paper)
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{hua21ITC33,
title = { Robust Wireless Scheduling under Arbitrary Channel Dynamics and Feedback Delay (Invited Paper) },
year = { 2021 },
address = { Avignon, France },
author = { Huang, Jiatai and Huang, Longbo },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-8 }
}
Abstract: Designing efficient scheduling algorithms is crucial to the development of modern wireless networks. In this paper, we study a wireless network model consisting of one central base-station and K mobile users. Each time the base-station can simultaneously transmit to 1 ≤ M ≤ K users. The channel states change over time adversarially, and the feedback of transmission outcome can experience arbitrary delays. The objective of the base-station is to search for a policy to maximize the overall transmission success rate. We propose a scheduling algorithm named Banker-OMD-Scheduling for this setting, based on a recent banker online mirror descent technique [1]. We show that Banker-OMD-Scheduling guarantees that the total regret over a finite time horizon T is $Ołeft( \sqrt MK łeft( \sqrt T + \sqrt Dłog D i̊ght) \g̊ht)$ where D is the total feedback delay.
Datar, Mandar and Altman, Eitan
Strategic Resource Management in 5G Network Slicing
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{dat21ITC33,
title = { Strategic Resource Management in 5G Network Slicing },
year = { 2021 },
address = { Avignon, France },
author = { Datar, Mandar and Altman, Eitan },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: Network Slicing is one of the essential concepts that has been introduced in 5G networks design to support demand expressed by next generation services. Network slicing will also bring new business opportunities for service providers (SPs) and virtual network operators, allowing them to run their virtual, independent business operations on shared physical infrastructure. We consider a marketplace where service providers (SPs) i.e., slice tenants lease the resources from an infrastructure provider (InP) through a network slicing mechanism. They compete to offer a certain communication service to end-users. We show that the competition between SPs can be model using the multi-resource Tullock contest (TC) framework, where SPs exert effort by expending costly resource to attract users. We study the competition between the SPs under a static and dynamic resource sharing scheme. In a dynamic resource sharing scheme, SPs are pre-assigned with fixed shares (budgets) of infrastructure, and they are allowed to redistribute their shares and customise their allocation to maximise their profit. The decision problem of SPs is analysed using non-cooperative game theory, and it is shown that the resultant game admits a unique Nash Equilibrium (NE). Furthermore, a distributed reinforcement algorithm is proposed that allows each SP to reach the game’s unique Nash equilibrium. Finally, simulations results are conducted to analyse the interaction between market players and the economic efficacy of the network sharing mechanism.
Baiocchi, Andrea and Turcanu, Ion and Vinel, Alexey
To Buffer or Not To Buffer: IEEE 802.11p/bd Performance Under Different Buffering Strategies
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{bai21ITC33,
title = { To Buffer or Not To Buffer: IEEE 802.11p/bd Performance Under Different Buffering Strategies },
year = { 2021 },
address = { Avignon, France },
author = { Baiocchi, Andrea and Turcanu, Ion and Vinel, Alexey },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: A fundamental paradigm of the Internet of Things (IoT) consists of agents that communicate updates to each other to perform joint actions, e.g., cooperative awareness in transportation systems, swarms of Unmanned Aerial Vehicles (UAVs), fleet of robots, automated assembly lines and logistics. A common feature of update messaging is emphasis on reliable throughput and freshness of collected data. We develop an analytical model that yields accurate predictions of all relevant metrics, both in terms of moments and probability distributions, for the case of one-hop broadcast update messages exchanged by using a CSMA-based wireless network. The model is validated against simulations and then applied to compare two update message scheduling approaches: providing a minimal buffer resource or providing no buffer. Surprisingly, we prove that having no buffer improves Age of Information (AoI) performance as well as message delivery rate, in spite of dropped packets. This is essentially due to much smaller congestion and hence collision probability in the wireless channel. From a system point of view this suggests a simple design of message handling, with no need of buffering and overwriting older messages. From a modeling point of view, the result supports the definition of simpler models that need not keep into account buffer state.
Zerwas, Johannes and Kellerer, Wolfgang and Blenk, Andreas
What You Need to Know About Optical Circuit Reconfigurations in Datacenter Networks
In 2021 33rd International Teletraffic Congress (ITC-33). Avignon, France 2021
[BibTeX]
[Abstract]
[Download]
[BibSonomy]
@inproceedings{zer21ITC33,
title = { What You Need to Know About Optical Circuit Reconfigurations in Datacenter Networks },
year = { 2021 },
address = { Avignon, France },
author = { Zerwas, Johannes and Kellerer, Wolfgang and Blenk, Andreas },
booktitle = { 2021 33rd International Teletraffic Congress (ITC-33) },
month = { Aug },
pages = { 1-9 }
}
Abstract: Increasing demand for flexibility in datacenter networks has led researchers to propose various designs of adaptable topologies using optical circuit switching. However, reconfigurations interrupt data transmissions that can diminish the benefit of adapting physical networks. The state-of-the-art lacks detailed models of end-to-end reconfiguration characteristics of involved networking components like end-host NICs and switches. The measurements of five commercially available programmable NICs and switches demonstrate that end-to-end reconfiguration delays indeed exhibit variability across devices and settings. The results suggest, however, that their behavior can be modeled, which opens new opportunities for scheduling algorithms.