Cache-MCDM : A hybrid caching scheme in Mobile Named Data Networks based on multi-criteria decision making
Shahid Md. Asif Iqbal,Asaduzzaman
DOI: https://doi.org/10.1016/j.future.2024.01.016
IF: 7.307
2024-01-16
Future Generation Computer Systems
Abstract:Robust and reliable content retrieval in a Mobile Ad Hoc Network ( MANET ) is challenged by its intermittent connectivity and mobility characteristics. In-network caching and name-based content retrieval attributes of Named Data Networks ( NDN ) can counterattack the stated challenges of MANET s. The distance- or hop-based caching schemes, proposed mainly for infrastructure NDN , admit contents either close to the content source, client, or midway, which are trivial to realize since content mainly travels from the network core (content sources) to the edge (clients). In NDN -based MANET s, the mobility attribute causes source and consumer nodes to change their locations frequently. Therefore, the content can travel from the network core to the edge and vice-versa. Under such circumstances, the traditional hop-based schemes that cache at particular points, for instance, either near the source, client, or halfway, may vandalize the efficient content retrieval in MANET s. Thus, proposing a new strategy capable of caching at any of the stated caching points considering varying network contexts, for instance, location change of source and consumer, is inevitable. Such a strategy can scale well in terms of network resource usage and significantly improve PDR and content retrieval latency performances in NDN -based MANET . In this direction, we propose an AHP -based hybrid caching scheme, namely Cache-MCDM , that equips each router with a set of different hop-based caching options and employs AHP-TOPSIS method to pick the most suitable caching option. The proposed method models the suitable option selection as a Multi-Criteria Decision Making ( MCDM ) problem. Simulation reveals that Cache-MCDM is more effective and efficient than hop-based probabilistic caching ( HPC ), opportunistic caching ( OppProb ), probabilistic in-network caching ( ProbCache ), and SourceCacheProb ( SCP ), in terms of (i) PDR by approximately 3%, 15%, 10%, and 11%, respectively, (ii) latency by approximately 25%, 145%, 75%, and 77%, respectively, (iii) retransmission ratio by approximately 7%, 30%, 25%, and 25%, respectively, (iv) cache replacement rate by approximately 8%, 100%, 80%, and 25%, respectively.
computer science, theory & methods