Neural-Blockchain-Based Ultrareliable Caching for Edge-Enabled UAV Networks

Vishal Sharma, Ilsun You, Dushantha Nalin K. Jayakody, Daniel Gutierrez Reina, Kim Kwang Raymond Choo

Результат исследований: Материалы для журналаСтатья

8 Цитирования (Scopus)

Аннотация

Mobile edge computing (MEC) reduces the computational distance between the source and the servers by fortifying near-user site evaluations of data for expedited communications, using caching. Caching provides ephemeral storage of data on designated servers for low-latency transmissions. However, with the network following a hierarchical layout, even the near-user site evaluations can be impacted by the overheads associated with maintaining a perpetual connection and other factors (e.g., those relating to the reliability of the underpinning network). Prior solutions study reliability as a factor of throughput, delays, jitters, or delivery ratio. However, with modern networks supporting high data rates, a current research trend is in ultrareliability. The latter is defined in terms of availability, connectivity, and survivability. Thus, in this paper, we focus on the ultrareliable communication in MEC. Specifically, in our setting, we use drones as on-demand nodes for efficient caching. While some existing solutions use cache-enabled drones, they generally focus only on the positioning problem rather than factors relating to ultrareliable communications. We present a novel neural-blockchain-based drone-caching approach, designed to ensure ultrareliability and provide a flat architecture (via blockchain). This neural-model fortifies an efficient transport mechanism, since blockchain maintains high reliability amongst the peers involved in the communications. The findings from the evaluation demonstrate that the proposed approach scores well in the following metrics: the probability of connectivity reaches 0.99; energy consumption is decreased by 60.34%; the maximum failure rate is affected by 13.0%; survivability is greater than 0.90; reliability reaches 1.0 even for a large set of users.

Язык оригиналаАнглийский
Номер статьи8734799
Страницы (с-по)5723-5736
Число страниц14
ЖурналIEEE Transactions on Industrial Informatics
Том15
Номер выпуска10
DOI
СостояниеОпубликовано - 1 окт 2019

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Information Systems
  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint Подробные сведения о темах исследования «Neural-Blockchain-Based Ultrareliable Caching for Edge-Enabled UAV Networks». Вместе они формируют уникальный семантический отпечаток (fingerprint).

  • Цитировать