In this work we provide an optimal assignment of edge IoT nodes to act as hash calculators for blockchain mining using Pareto dominance sorting. IoT nodes being resource constrained are not suited to execute intensive mining operations individually. But, collectively they can be used small parallel computation cores for hash calculation within a specific nonce range. We propose an assignment algorithm that selects a set of IoT nodes which are expected to consume the least amount of resources by assigning the mining tasks over those nodes. Assuming that a prior empirical relation is known for each node that can predict the expected amount of mining resources like energy and time of mining, before each mining epoch, the proposed algorithm calculates the expected resource consumption values for all the nodes. It then sorts the nodes into different Pareto fronts based on dominated sorting and further, for each front it calculates a cumulative inter-Pareto distance for each node. We show that by selecting the nodes as per the sorted inter-Pareto distance, an optimal selection is achieved that reduces the overall composite resource consumption for block mining. Through detailed experimental analysis, we have corroborated our claims, where we find that proposed algorithm achieves a benefit of 47.71% with respect to energy consumption and a benefit of 19.27% with respect to improvement in mining time over other selection algorithms.