Bloom filter based distributed and cached service registries for highly dynamic V2X environments

Abstract

Owing to the recent application of a Service-Oriented Architecture (SOA) to the Internet of Things (IoT), traditional service provisioning processes have been changed to consider the physical environment and changes generated by IoT services. However, in highly dynamic IoT environments such as Vehicle-to-Everything (V2X), it is difficult for IoT services to provide their capabilities to users as a whole because the changes in the environment occur too quickly and unpredictably. Therefore, it is necessary to discover services as quickly as possible that can deliver their physical effects. For a fast and effective service discovery in a highly dynamic environment, we propose the application of a service discovery from a cached service registry on a mobile entity and a redundant sharing method for the synchronization of registries. To overcome the network overhead caused by redundantly shared data, we adopt a Bloom Filter applied distributed service registry and extend it to the adaptation of highly dynamic environments. Experiments were conducted to evaluate our approach by simulating a V2X environment. The results show that the cached service registry can guarantee a higher chance of success of a service discovery for highly dynamic environments, and the network overhead is significantly reduced.