Co-Operative and Hybrid Replacement Caching for Multi-Access Mobile Edge Computing

Abstract

Multi-Access Mobile Edge Computing (MEC) is proclaimed as a key technology for reducing service processing delays in 5G networks. Caching on MEC will decrease service latency and improve data access by allowing direct content delivery through the edge without fetching content from the remote server. Caching on MEC is also deemed as an effective approach guarantying more reachability due to proximity to end-users. This paper proposes a novel hybrid content caching replacement algorithm in MEC to increase its caching efficiency where future request references are predicted using a polynomial fit algorithm along with Lagrange interpolation. Additionally, a distributed cooperative caching algorithm to improve data access within MECs. Experimental results have shown that the proposed scheme obtains more cache hits and lesser average CPU utilization due to its selective caching approach when compared with existing traditional cache replacement algorithms. Keywords-Multi-access Mobile Edge Computing, selective caching, OPR algorithm, cooperative caching. I. Few researchers have proposed edge caching strategies for 5G that combines both computation caching and data caching. Markakis et al [8] proposed a proactive edge caching strategy that predicts and prefetch popular contents based on big data analysis. Sungwook K. [9] leveraged a

A Survey and Taxonomy on Task Offloading for Edge-Cloud Computing

“…One of the most effective ways to address the problem is caching data the offloaded task needed in advance. While it is difficult to design a good data caching strategy with a high access hit ratio [152], because user data access patterns are hard to be predicted due to the high diversity and mobility of users in edge-cloud computing.…”

A Novel Predictive-Collaborative-Replacement (PCR) Intelligent Caching Scheme for Multi-Access Edge Computing

“…Rules all = {rule 1 , rule 2 ..rule n } ∀ rule ant i ⊆ S con (16) The generated rules Rule all are then ranked to evaluate the strongest rules using the rule support (Rule support ) and rule confidence (Rule conf idence ) [40]. Rule conf idence of (con i → con j ) is the proportion of transactions in D including both con i and con j .…”

“…Therefore, these algorithms alone cannot adapt and adjust to the ever-changing user request patterns. Following the increasing popularity of machine learning and data analytics, there has been progress made on the prediction based caching algorithms [13], [14], [15], [16]. Most of these caching schemes use Recurrent Neural Network (RNN) algorithms like Long Short-Term Memory (LSTM) [17] which involves data preparation, feature extraction, model training and finally cache replacement using the trained model.…”

A centralized hybrid routing model for multicontroller SD‐WANs

“…In our previous work, an efficient resource provisioning scheme over MEC is proposed that avoids deadlock condition, 35 an efficient content-caching algorithm is introduced that uses MEC collaboration. To minimize traffic load among MEC instances a routing algorithm has been proposed 36 which makes use of node and link costs to find best-path over an SDN. We have devised STEN 37 for this purpose, which has also used to cater rapid-convergence in a network with high resource-variance, SDN-SIM.…”