Computing Resource Allocation Scheme for DAG-Based IOTA Nodes - Université de Pau et des Pays de l'Adour Access content directly
Journal Articles Sensors Year : 2021

Computing Resource Allocation Scheme for DAG-Based IOTA Nodes


IOTA is a distributed ledger technology (DLT) platform proposed for the internet of things (IoT) systems in order to tackle the limitations of Blockchain in terms of latency, scalability, and transaction cost. The main concepts used in IOTA to reach this objective are a directed acyclic graph (DAG) based ledger, called Tangle, used instead of the chain of blocks, and a new validation mechanism that, instead of relying on the miners as it is the case in Blockchain, relies on participating nodes that cooperate to validate the new transactions. Due to the different IoT capabilities, IOTA classifies these devices into full and light nodes. The light nodes are nodes with low computing resources which seek full nodes’ help to validate and attach its transaction to the Tangle. The light nodes are manually connected to the full nodes by using the full node IP address or the IOTA client load balancer. This task distribution method overcharges the active full nodes and, thus, reduces the platform’s performance. In this paper, we introduce an efficient mechanism to distribute the tasks fairly among full nodes and hence achieve load balancing. To do so, we consider the task allocation between the nodes by introducing an enhanced resource allocation scheme based on the weight least connection algorithm (WLC). To assess its performance, we investigate and test different implementation scenarios. The results show an improved balancing of data traffic among full nodes based on their weights and number of active connections.

Dates and versions

hal-03725650 , version 1 (18-07-2022)



Houssein Hellani, Layth Sliman, Abed Ellatif Samhat, Ernesto Expósito. Computing Resource Allocation Scheme for DAG-Based IOTA Nodes. Sensors, 2021, 21 (14), pp.4703. ⟨10.3390/s21144703⟩. ⟨hal-03725650⟩


56 View
0 Download



Gmail Mastodon Facebook X LinkedIn More