Abstract
Edge computing architectures are designed for providing in-network computational and resource allocation support for the connected users. Such architectures provide ease of computations and latency-less resource allocations for improving the efficiency of service disseminations. The problem of multi-access to the resources increases the latency in distributing the computing features for concurrent users. In this article, access to appropriate analogous computing (A3C) is proposed for addressing this issue. The proposed computing model performs the parallel request and migration processing for providing multiple accesses to the available edge resources. The multiple processing instances are identified based on the service dissemination and lag to the migrating resource factors. The instances are differentiated, and supportive migration access is analyzed for the two factors using transfer learning. This learning helps to improve the recommendations for distributing the resources for additive access. The proposed computing model improves the ratio of service dissemination under controlled latency and service lag. The stagnancy in distributing the service computations using the edge devices helps provide reliable access to edge users. The experimental analysis shows that the proposed A3C improves service dissemination by 11.52%, computation distribution by 8.6%, and access rate by 10.03%, whereas it reduces the service lag and latency by 22.2% and 11.07%, respectively, for the different edge devices.