Boosting computation reuse efficiency in ICN-based edge computing via improved forwarding algorithms

Abstract

With the advancement of the Internet of Things (IoT) and the changing needs of edge computing applications within the TCP/IP architecture, several challenges have emerged. One solution to these challenges is to integrate edge computing with information-centric networks (ICN). In ICN-based edge computing, there is a high level of similarity in request computing due to the proximity of users, which is leveraged to improve the efficiency of computation reuse. Computation reuse occurs through naming, caching, and forwarding. Computation reuse through forwarding means that similar requests are directed to the same compute node (CN). In many past works, forwarding algorithms for computation reuse have been used with high overhead for resource discovery or did not consider the important criterion of assessing the capacity of CNs. In this paper, we propose two forwarding algorithms, named TLCF) Trade-Off Between Load Balancing and Computation Reuse Forwarding) and AFCT (Adaptive Forwarding Considering Capacity Threshold), that measures criteria for selecting the best CN, the trade-off between computation reuse and load balancing, while also considering capacity. These two aspects lead to a reduction in completion time. Computation reuse inherently disrupts load balancing. The evaluation was conducted using the ndnSIM simulation. Through simulations, we have shown that our method significantly reduces completion time compared to the default method, achieving an improvement of approximately 22%. These findings highlight the efficiency and potential of our proposed method in optimizing edge computing performance. © The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2025.