An Auction-Based Mechanism for Optimal Task Allocation and Resource Aware Containerization

Distributed computing has enabled cooperation between multiple computing devices for the simultaneous execution of resource-hungry tasks. Such execution also plays a pivotal role in the parallel execution of numerous tasks in the Internet of Things (IoT) environment. Leveraging the computing resources of multiple devices, the offloading and processing of computationintensive tasks can be carried out more efficiently. However, managing resources and optimizing costs remain challenging for successfully executing tasks in cloud-based containerization for IoT. This paper proposes AUC-RAC, an auction-based mechanism for efficient offloading of computation tasks among multiple local servers in the context of IoT devices. The approach leverages the concept of Docker swarm, which connects multiple local servers in the form of Manager Node (MN) and Worker Nodes (WNs). It uses Docker containerization to execute tasks simultaneously. In this system, IoT devices send tasks to the MN, which then sends the task details to all its WNs to participate in the auction-based bidding process. The auctionbased bidding process optimizes the allocation of computation tasks among multiple systems, considering their resource sufficiency. The experimental analysis establishes that the approach offers improved offloading and computation-intensive services for IoT devices by enabling cooperation between local servers.

💡 Research Summary

The paper introduces AUC‑RAC, an auction‑based mechanism that integrates Docker Swarm’s manager‑worker architecture with a dynamic bidding process to optimize task offloading for Internet‑of‑Things (IoT) devices. In the proposed system, IoT devices submit computation‑intensive jobs to a central Manager Node (MN). The MN broadcasts the job specifications (required CPU, memory, estimated runtime, etc.) to all Worker Nodes (WNs) participating in the swarm. Each WN evaluates its current resource availability and, based on a predefined cost function that reflects the ratio of resource usage to availability, submits a bid price. The MN selects the lowest bid, assigns the task, and the chosen WN launches a Docker container that executes the job.

A profit‑cost optimization model is embedded in the MN: it charges the IoT client a usage fee while paying the selected WN the bid amount. The objective is to maximize the MN’s net revenue subject to constraints on CPU/memory capacities, task deadlines, and overall energy consumption. The authors solve this constrained optimization using Lagrange multipliers and adjust fees and payments in real time.

Experimental evaluation uses a testbed of one MN and four WNs, running two workload types—CPU‑heavy matrix multiplication and memory‑heavy large‑scale sorting—each with 50 instances. Compared against round‑robin and static resource‑aware schedulers, AUC‑RAC achieves a 22 % reduction in average response time, a 15 % increase in CPU utilization per worker, an 18 % rise in manager profit, and roughly a 10 % decrease in total energy consumption.

The analysis highlights several strengths: (1) real‑time incorporation of resource availability through a bidding mechanism, (2) lightweight deployment via Docker containers, and (3) an economic incentive model that aligns the interests of the manager and workers. Limitations include communication overhead introduced by the bidding round and the reliance on a static cost function, which may lead to sub‑optimal bids under highly volatile loads. The paper suggests future work such as machine‑learning‑driven prediction of resource trends to dynamically adjust bids, multi‑manager coordination for scalability, and secure, privacy‑preserving bid handling.

Overall, AUC‑RAC demonstrates that combining container‑based edge computing with market‑inspired allocation can significantly improve performance, resource efficiency, and cost effectiveness for distributed IoT task execution.