DHT-based distributed ALE engine in RFID Middleware

International audienceFollowing the "Internet of Things" concept, each object is associated with a unique identifier which will allow to retrieve information about it in large databases. In the process of managing a large amount of objects, and consequently a large amount of events from readers, without overloading the network, these events have to be filtered and aggregated. This is the aim of the Application Level Events (ALE) standard from EPCGlobal, which receives events from readers and sends a useful and well constructed report to the business application. The ALE may be connected to several hundreds of readers. As the number of readers may increase with the increase of the company, a bottleneck may appear with all readers events sent to the ALE. A solution for scalability is to distribute the ALE. In this paper, we propose an efficient way to solve this problem based on a Distributed Hash table (DHT). One role of the ALE is to insulate business application from technical concern so in our solution, we present a mechanism to distribute the ALE using Chord, a well-known peer-to-peer lookup system, and being transparent for business applications. This solution is compliant with the EPCglobal existing standard, scalable, robust and transparent for other layers of the middleware. We show that our solution generates only 10% overhead than in a nominal case while offering a better robustness and scalability when numbers of tags and readers increase significantly