Resource Efficient Real-time Reliability Model for Multi-Agent IoT Systems

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F22%3A00353949" target="_blank" >RIV/68407700:21230/22:00353949 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/ACCESS.2021.3138931" target="_blank" >https://doi.org/10.1109/ACCESS.2021.3138931</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ACCESS.2021.3138931" target="_blank" >10.1109/ACCESS.2021.3138931</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Resource Efficient Real-time Reliability Model for Multi-Agent IoT Systems

Popis výsledku v původním jazyce

Rising demand for the integration of IoT services in cloud computing has brought cloud- native principles (e.g., system automation, loosely coupled services, etc.) to the edge of the network. A fundamental element of the cloud-native approach is a microservice — an instance, which is usually virtualised, serving for only a defined purpose. Virtualised microservices are mostly used in cloud-based IoT systems, which can be covered with multi-agent system (MAS) paradigm. The maintenance of a MAS requires a constant monitoring to track the system state and to satisfy the service level agreement (SLA) (or any other) requirements. This paper presents a resource efficient reliability model for MAS IoT systems for monitoring purposes. First, we provide a thorough mathematical analysis to describe the generic system model and its elements. This model takes into account SLA requirements and it is characterised with linear time complexity, simplicity of computations and input metrics (e.g., combination of SLA for CPU and its workload), where all of the input data are aligned to the same range. Then, we evaluate time complexity and provide a measurement to demonstrate the reliability of the model application. The results show that the proposed model is efficient for large number of the IoT agents. It is approximately 98 times faster in AQ:4 computation with 1,000 systems and 51 times faster in computations with 1,000 systems.

Název v anglickém jazyce

Resource Efficient Real-time Reliability Model for Multi-Agent IoT Systems

Popis výsledku anglicky

Rising demand for the integration of IoT services in cloud computing has brought cloud- native principles (e.g., system automation, loosely coupled services, etc.) to the edge of the network. A fundamental element of the cloud-native approach is a microservice — an instance, which is usually virtualised, serving for only a defined purpose. Virtualised microservices are mostly used in cloud-based IoT systems, which can be covered with multi-agent system (MAS) paradigm. The maintenance of a MAS requires a constant monitoring to track the system state and to satisfy the service level agreement (SLA) (or any other) requirements. This paper presents a resource efficient reliability model for MAS IoT systems for monitoring purposes. First, we provide a thorough mathematical analysis to describe the generic system model and its elements. This model takes into account SLA requirements and it is characterised with linear time complexity, simplicity of computations and input metrics (e.g., combination of SLA for CPU and its workload), where all of the input data are aligned to the same range. Then, we evaluate time complexity and provide a measurement to demonstrate the reliability of the model application. The results show that the proposed model is efficient for large number of the IoT agents. It is approximately 98 times faster in AQ:4 computation with 1,000 systems and 51 times faster in computations with 1,000 systems.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

20203 - Telecommunications

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

IEEE Access

ISSN

2169-3536

e-ISSN

2169-3536

Svazek periodika

10

Číslo periodika v rámci svazku

December

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

13

Strana od-do

2578-2590

Kód UT WoS článku

000739980100001

EID výsledku v databázi Scopus

2-s2.0-85122323953