RepMChord: A novel replication approach for mobile Chord with reduced traffic overhead

Abstract

Structured peer-to-peer systems based on distributed hash table (DHT) have known a great popularity and performance since their appearance. They have experienced multiple improvements to increase the efficiency, like replication mechanism with different used methods and different objectives like increasing data availability or fitting to churn. Besides their benefits, these methods suffer from the excessive generated overhead in maintenance process. On the other side, we have the interest deployment of DHT overlay on mobile ad hoc network, which benefits from the infrastructure-less architecture, but presents some shortcomings because of the limited bandwidth and energy batteries, what require a reduced overhead. Therefore, the aim of this work is to improve the lookup efficiency of DHT-based Chord on mobile ad hoc network underlay. For that, we propose a novel replication mechanism based on data structure to determine the replica nodes, while avoiding the excessive generated overhead in maintenance process to cope with the problem of limited energy batteries. To evaluate the proposed method, we present an extensive simulation study that compares the work to another efficient replication method and to mobile basic Chord. The results show the efficiency of our approach in decreasing the lookup path, the maintenance overhead, and the energy consumption.

ZRR-P2P: Zone-Based Mechanism for Data Replication and Research Optimization in Unstructured P2P Systems

“…Replication methods applied on P2P networks are used to achieve some purposes as enhancing data availability or reducing data searching latency [8,19]. In the literature, we find different papers that are interested by this mechanism.…”

“…In structured P2P, we can notice some replication works like in [8,[20][21][22]. In which, authors replicate resources in specific placements according to a structured protocol like Chord [23].…”

“…The authors of each paper propose a formula to define the identifiers of nodes where to replicate data. Moreover, they improve the lookup routing to reach the closest replica node [8,21,22]. The formula proposed by Ghodsi et al [21] defines identifiers of nodes that are dispersed in the network.…”

“…As another case of mobile networks, we can consider MANETs since they share the same characteristics as P2P networks such as decentralization and dynamic topology [7]. Researches are interested by improving P2P systems over MANET in terms of different characteristics as lookup performance by maintaining data placement on P2P topology [8], routing lookup queries to relevant peers that are more suitable to answer the query [9] and data dissemination based on clustering to reduce traffic overhead [10].…”

Data replication mechanisms in the peer‐to‐peer networks

“…Furthermore, Cherbal et al have proposed a scheme to define the placement of replicas using a simple data structure and predecessor list. Then, a new lookup routing is defined to provide a facility for preparing the closet node by the requester.…”